Critical reflections on how research design and the ...

2 downloads 0 Views 35MB Size Report
Feb 16, 2015 - 2 “Secondary school” refers to the last six years of formal schooling in ...... promote social connections within the school itself (Blum & Libbey,.
Critical reflections on how research design and the attributes of a music program can affect investigations of the psychosocial wellbeing benefits of musical participation in mainstream schools

Alexander Hew Dale Crooke Submitted in total fulfilment of the requirements of the degree of Doctor of Philosophy

February 2015

Faculty of VCA and MCM University of Melbourne

Produced on Archival Quality Paper

Abstract This project explores the challenges of investigating the psychosocial wellbeing benefits of musical participation in mainstream schools. For a decade, Australian policy literature has claimed these benefits are to be expected outcomes of all students’ participation in school music programs (Australian Government, 2005). Despite these claims, there is little to no consistent evidence supporting a link between musical participation and psychosocial wellbeing in this context (Grimmett, Rickard, Gill, & Murphy, 2010; Rickard, Bambrick, & Gill, 2012). The reason for this inconsistency has been linked to both the research designs and methods used (Knox Anderson & Rickard, 2007), as well as the nature of musical participation investigated (Darrow, Novak, Swedberg, Horton, & Rice, 2009). Despite the identification of these limitations, researchers have continued to use designs that contain them. This can be attributed to a lack of critical engagement with approaches to research in this field, including assumptions about the efficacy of certain research methods, and the capacity for generic school music programs to promote wellbeing. This lack of critical engagement appears to account for the enduring inconsistency of findings in this area. This thesis aims to address this lack of engagement by critically appraising the research approaches used in two small studies that aimed to demonstrate the psychosocial wellbeing benefits of school-based music programs. This was achieved by undertaking two critical reflection analyses on the methods, designs, and contexts of each study, as well as the attributes of the music programs investigated. The first of these identified a number of important research challenges related to the research methods and designs used. Among other things, these findings challenge the assumption that self-report surveys are a valid way of collecting data from students. The second analysis identified a number of music program attributes that are likely to inhibit the reporting of positive results. For example, findings suggest music education programs are unsuited to promoting psychosocial wellbeing. Based on these findings, this dissertation makes a number of recommendations for the design of future studies in this area. It is argued that research following these recommendations is crucial for this field. This is both to develop a richer understanding of the relationship between music in schools and psychosocial wellbeing, and to produce reliable evidence that is better placed to inform relevant policy. It is further argued that without such evidence, policymakers may continue to make uninformed

ii

claims regarding the link between music in schools and psychosocial wellbeing. In turn, this has the potential to destabilise policy support for music in Australian schools. Finally, this thesis calls on researchers in this field, and others, to critically engage with the way that knowledge is created. It is maintained that such engagement is the responsibility of all researchers in the social sciences, and that only when this occurs can we claim the knowledge we generate is meaningful, and serving the communities we investigate.

iii

Declaration

This is to certify that:

i.

The thesis comprises only my original work towards the PhD except where indicated in the Preface,

ii.

Due acknowledgement has been made in the text to all other material used,

iii.

The thesis is fewer than 100 000 words in length, exclusive of tables, maps, bibliographies and appendices.

Signed:

Name:

Alexander Hew Dale Crooke

Date:

iv

Preface

This thesis contains some work that was carried out in collaboration with others. This includes certain data collection activities, and the co-authorship of a journal article. Details of these collaborations are stated here.

Data collection activities: Chapter 6

The primary PhD supervisor, Professor Katrina Skewes McFerran, administered Pre and post surveys for the Bereavement Group study. These data collection activities are described in Chapter 6 of this dissertation.

Journal article: Chapter 11

The recommendations of the PhD project have been written up as a journal article. This article was co-authored by both the PhD candidate and the primary PhD supervisor, Professor Katrina Skewes McFerran. The PhD candidate contributed 80% of the article content, analysis and authorship, while the primary supervisor contributed 20%. This article was published in a peer reviewed academic journal in December 2014. The PDF of this published article is included as an independent chapter (Chapter 11) in this dissertation.

Article information:

Crooke, A. H. D., & McFerran, K. S. (2014). Recommendations for the investigation and delivery of music programs aimed at achieving psychosocial wellbeing benefits in mainstream schools. Australian Journal of Music Education(1), 2749.

v

Acknowledgements It is said that undertaking a PhD is a journey. Many people have travelled with me on mine, contributing to its fulfilment in their own way. I would like to formally acknowledge their support in helping me complete my passage. To start, I express my sincerest gratitude to my primary supervisor, Katrina Skewes McFerran. By taking me on, you issued a ticket for what has been an incredible voyage of discovery. You have empowered me to grow by being both supportive, and pushing me to challenge myself. As inspirational as you are insightful, you have helped me navigate every stage of this journey with an enthusiasm that has been contagious. I also convey heartfelt thanks to my secondary supervisor, Paul Smyth. You also took me on with enthusiasm, and helped guide me through the complex world of policy. Both patient and practical, I have learnt a lot from you on this journey. To my National Music Therapy Research Unit (NaMTRU) colleagues at the University of Melbourne (Lucy [B], Grace, Carmen, Juyoung, Lucy [F], Cherry, Jason, Ben, Rebecca, Jennifer, Laura, Imogen, Kate, Melissa, Elizabeth, and John), your advice and support changed my idea of what it means to be part of a research community. To the scholars who visited NaMTRU during my candidature (Brynjulf Stige, Joke Bradt, Linda Finlay, and Hanne Mette Oschner Ridder [to name but a few]), your knowledge inspired me, and your feedback grounded me. I also thank my PhD advisory committee: Felicity Baker, Jeanette Tamplin, Nikki Rickard, and Denise Grocke. Your gentle, yet honest and realistic feedback guided me, and kept me on track. To my parents, I offer my deepest thanks. Without you I would neither have been in a position to start this journey, nor complete it. To my wider family (particularly my aunts, Nell and Pamela) your tireless support has meant the world to me. To Brenda: I cannot imagine having travelled this road without you. The ways you have pulled and pushed me through this PhD are innumerable, and cannot possibly be summed up here. Thank you for everything. To my examiners, thank you for taking the time to read this dissertation. By bearing witness to my journey you give its completion meaning. Finally, I thank the school community described in this PhD. Your participation and patience made this project possible. I hope my interpretation of the experiences you shared with me go someway to helping not only your school community, but that of your children, and future generations throughout Australia, and the world.

vi

“Critical does not mean destructive, but only willing to examine what we sometimes presuppose in our way of thinking, and that gets in the way of making a more livable world”.

(Judith Butler, 2012)

vii

Table of Contents

Abstract ............................................................................................................................. ii Declaration....................................................................................................................... iv Preface .............................................................................................................................. v Acknowledgements ......................................................................................................... vi Table of Contents .......................................................................................................... viii List of Tables ............................................................................................................... xviii List of Figures ................................................................................................................ xix Chapter 1: Introduction ................................................................................................. 1 Music, Schools, Policy and Research: A Brief Background ........................................ 1 The PhD “Journey” ....................................................................................................... 4 Life experiences informing this PhD. ....................................................................... 5 Beginning the PhD. .................................................................................................. 7 Three critical decisions. ............................................................................................ 8 Moments of dissonance. ......................................................................................... 10 Change in direction................................................................................................. 11 Current Study Aim and Thesis Structure. ................................................................... 13 Chapter 2: Music in Schools and Policy ..................................................................... 16 Setting the Scene ........................................................................................................ 17 Policy and music in Australian schools: a short history. ........................................ 17 Contemporary shifts towards advocacy for the Arts in education.......................... 18 Current support for music in Australian schools. ................................................... 19 Policy Developments Underpinning Support for the Arts in Education .................... 21 Social inclusion and the return of “the social”. ...................................................... 21 Education policy and the social role of schooling. ................................................. 23 Sustainability of Support for Music in Schools .......................................................... 25 Calls for a more informed understanding of school music benefits. ...................... 26 Chapter 3: Interrogating Policy Claims With Evidence ........................................... 28 Policy Promises for Music in Schools ........................................................................ 28 Intrinsic benefits. .................................................................................................... 30 Extrinsic benefits related to cognitive and academic development........................ 31 Extrinsic benefits related to subjective wellbeing. ................................................. 32 Models of musical participation advocated by policy. ........................................... 33 viii

Achieving intrinsic benefits. ............................................................................... 34 Achieving extrinsic benefits related to cognitive and academic development. .. 35 Achieving extrinsic benefits related to psychosocial wellbeing. ........................ 35 Evidence to Support Policy Claims for the Benefits of School Music....................... 36 Evidence for intrinsic benefits. ............................................................................... 36 Evidence for cognitive and academic benefits. ...................................................... 37 Evidence for subjective wellbeing benefits. ........................................................... 38 The Gap and its Perilous Potential ............................................................................. 41 Filling the Gap ............................................................................................................ 42 Chapter 4: Accounting for the Inconsistency in Current Evidence......................... 43 Conceptualising Psychosocial Wellbeing................................................................... 43 Objective versus subjective social wellbeing. ........................................................ 44 Social and psychological wellbeing: distinct or linked? ........................................ 44 Conceptualising young people’s psychosocial wellbeing. ..................................... 45 Reported Limitations of Investigating the Psychosocial Benefits of School Music .. 47 Study design and methodology. ............................................................................. 48 Quantitative versus qualitative methods. ............................................................ 48 Quantitative design and sampling. ..................................................................... 50 Reporters............................................................................................................. 51 Scales. ................................................................................................................. 53 Nature of musical participation. ............................................................................. 54 Duration of participation. ................................................................................... 54 Age of Participants. ............................................................................................ 55 Status of sample population................................................................................ 56 Student engagement and musical activities ........................................................ 57 Delivery approach............................................................................................... 59 Setting. ................................................................................................................ 62 Other factors. ...................................................................................................... 63 Conclusion .................................................................................................................. 64 Chapter 5: Project Methodology and Design ............................................................. 66 Project Rational .......................................................................................................... 66 Project Aim ................................................................................................................. 66 Research Questions..................................................................................................... 66 Methodology............................................................................................................... 66 ix

Values. .................................................................................................................... 67 Nature of knowledge. ............................................................................................. 69 Inquiry posture........................................................................................................ 71 Quality criteria. ....................................................................................................... 72 Use of the first-person. ........................................................................................... 72 Project Design............................................................................................................. 73 Stage 1: empirical studies. ...................................................................................... 75 Empirical study 1: Keys to Success.................................................................... 76 Empirical study 2: Bereavement Group. ............................................................ 76 Reflexive evaluation. .............................................................................................. 76 Stage 2: critical reflection analyses. ....................................................................... 77 Critical reflection analysis 1. .............................................................................. 79 Critical reflection analysis 2. .............................................................................. 80 Recommendations. ................................................................................................. 81 Discussion............................................................................................................... 82 Conclusion. ............................................................................................................. 82 Chapter 6: Design and Method of Empirical Studies ............................................... 83 Empirical Study 1: Keyes to Success ......................................................................... 83 Design overview. .................................................................................................... 83 The music program. ................................................................................................ 84 Quantitative method. .............................................................................................. 85 Participants. ........................................................................................................ 85 Materials. ............................................................................................................ 86 Procedure. ........................................................................................................... 89 Analysis. ............................................................................................................. 89 Qualitative method. ................................................................................................ 90 Participants. ........................................................................................................ 91 Materials. ............................................................................................................ 91 Procedure. ........................................................................................................... 93 Analysis. ............................................................................................................. 94 Observational reports.............................................................................................. 96 Empirical Study 2: Bereavement Group ..................................................................... 98 Design overview. .................................................................................................... 98 The music program. ................................................................................................ 98 x

Quantitative method. .............................................................................................. 99 Participants. ........................................................................................................ 99 Materials. .......................................................................................................... 100 Procedure. ......................................................................................................... 101 Analysis. ........................................................................................................... 102 Qualitative method. .............................................................................................. 102 Participants. ...................................................................................................... 103 Materials. .......................................................................................................... 103 Procedure. ......................................................................................................... 104 Analysis. ........................................................................................................... 105 Observational reports............................................................................................ 105 Chapter 7: Keys to Success Results .......................................................................... 106 Methodological Overview ........................................................................................ 107 Quantitative Results .................................................................................................. 107 Community domain. ............................................................................................. 109 School domain. ..................................................................................................... 111 Peer and individual domain. ................................................................................. 113 Emotional control domain. ................................................................................... 115 Social support domain. ......................................................................................... 116 Psychological distress domain. ............................................................................. 117 Psychological wellbeing domain. ......................................................................... 119 Psychological needs domain................................................................................. 120 Quantitative results summary. .............................................................................. 122 Student Qualitative Results ...................................................................................... 122 Step 1: creating category templates. ..................................................................... 123 Step 2: coding student interviews. ........................................................................ 123 Step 3: creating themes. ........................................................................................ 127 Psychosocial wellbeing outcomes. ....................................................................... 130 Social connection.............................................................................................. 130 Outcomes extended beyond KTS class. ........................................................... 131 School Engagement. ......................................................................................... 131 Engagement with learning. ............................................................................... 132 Affect. ............................................................................................................... 133 Prosocial behaviour. ......................................................................................... 134 xi

Self-efficacy...................................................................................................... 135 Absence of outcomes. ....................................................................................... 135 Perceptions of the program. .................................................................................. 137 We chose the music. ......................................................................................... 138 We did it. .......................................................................................................... 138 We did it together. ............................................................................................ 138 I had a part to play. ........................................................................................... 139 The role of music. ............................................................................................. 139 It was fun. ......................................................................................................... 140 Made class better. ............................................................................................. 140 It was just about having fun.............................................................................. 140 It was about music education. .......................................................................... 141 We had to do it. ................................................................................................ 142 Same as normal KTS. ....................................................................................... 142 Not just the music. ............................................................................................ 143 Perceptions of the survey...................................................................................... 145 Couldn't see the point. ...................................................................................... 145 Didn't really like it. ........................................................................................... 145 Lack of comprehension. ................................................................................... 145 It was pretty useful. .......................................................................................... 146 Questions not fully comprehended. .................................................................. 147 We pretty much took it seriously...................................................................... 147 Emergent themes. ................................................................................................. 148 Didn't grasp the concept. .................................................................................. 149 Responding to what I said. ............................................................................... 151 Contradicting or changing stories. .................................................................... 151 Off the track. ..................................................................................................... 151 Non-committal. ................................................................................................. 152 It's hard to explain. ........................................................................................... 152 Just suspicious. ................................................................................................. 153 Teacher Qualitative Results ...................................................................................... 155 Psychosocial wellbeing outcomes. ....................................................................... 157 Everyone became involved in the class. ........................................................... 157 Engaged in learning. ......................................................................................... 157 xii

They achieved the program goals. .................................................................... 157 Social connection.............................................................................................. 158 Self-efficacy...................................................................................................... 159 Prosocial behaviour. ......................................................................................... 159 Improved teacher-student relationship. ............................................................ 159 Perceptions of the program. .................................................................................. 160 It provided the environment. ............................................................................ 160 Embedding music in their daily lives. .............................................................. 160 Music as a vehicle to achieve outcomes. .......................................................... 161 They owned the music. ..................................................................................... 161 Facilitation. ....................................................................................................... 161 Outcomes are conditional. ................................................................................ 162 Even one success is worth it. ............................................................................ 163 Perceptions of the survey...................................................................................... 163 They didn't take it seriously. ............................................................................. 163 Couldn't engage with the topic. ........................................................................ 164 Results don't reflect the reality. ........................................................................ 164 They need to know who, what and why. .......................................................... 164 Perhaps another method would be better. ......................................................... 165 Qualitative results summary. ................................................................................ 165 Observational Reports .............................................................................................. 166 Survey sessions. .................................................................................................... 166 Interview sessions. ................................................................................................ 167 Chapter 8: Bereavement Group Results .................................................................. 168 Methodological Overview ........................................................................................ 168 Quantitative Results .................................................................................................. 169 School domain. ..................................................................................................... 169 Emotional control domain. ................................................................................... 172 Social support domain. ......................................................................................... 174 Psychological distress domain. ............................................................................. 175 Psychological wellbeing domain. ......................................................................... 176 Psychological needs domain................................................................................. 177 Quantitative results summary. .............................................................................. 178 Qualitative Results .................................................................................................... 178 xiii

Step 1: creating category templates. ..................................................................... 179 Step 2: coding interviews. .................................................................................... 179 Step 3: creating themes. ........................................................................................ 180 Psychosocial wellbeing outcomes. ....................................................................... 180 Social connection.............................................................................................. 180 Social support. .................................................................................................. 183 Letting it out. .................................................................................................... 184 Dealing with bereavement. ............................................................................... 185 I can remember the good things. ...................................................................... 186 Moving on. ....................................................................................................... 187 Affect. ............................................................................................................... 187 Conditional outcomes. ...................................................................................... 188 Perceptions of the program. .................................................................................. 189 How music helped. ........................................................................................... 190 Musicality not required..................................................................................... 190 Change in musical relationship. ....................................................................... 191 Gave us the place and space. ............................................................................ 192 Private group setting. ........................................................................................ 192 We also had fun. ............................................................................................... 193 Hard but worth it. ............................................................................................. 194 Emergent themes. ................................................................................................. 195 The survey was confusing. ............................................................................... 196 We missed class, but it was worth it................................................................. 196 Maybe next time. .............................................................................................. 197 I’d recommend it to anyone. ............................................................................. 198 Qualitative results summary. ................................................................................ 198 Observational Reports .............................................................................................. 199 Survey sessions. .................................................................................................... 199 Interview sessions. ................................................................................................ 199 Chapter 9: Critical Reflection Analysis 1 ................................................................. 200 Inconsistency in Reported Benefits .......................................................................... 201 Discrepancy between reporting methods.............................................................. 201 Benefits only reported via qualitative methods. ............................................... 201 Inconsistencies between qualitative and quantitative data sets. ....................... 202 xiv

Discrepancy between reporters............................................................................. 203 Between students. ............................................................................................. 203 Between students and teachers. ........................................................................ 203 The Challenges of Comprehension........................................................................... 204 Ability to understand the language in surveys. .................................................... 204 Stage of development. ...................................................................................... 205 CALD. .............................................................................................................. 206 Ability to understand interviewer’s questions. ..................................................... 207 Ability to respond to interviewer’s questions. ...................................................... 207 Incongruence between lay and researcher’s vocabulary. ..................................... 208 Unfamiliarity with academic concepts. ................................................................ 208 Engagement in the Research Process ....................................................................... 209 Disengagement from survey sessions. .................................................................. 209 “Playing up”. .................................................................................................... 209 Not taking it seriously. ..................................................................................... 210 Resistance to participation. ............................................................................... 210 Rushed, careless, and missing responses. ......................................................... 210 Disengagement from interview sessions. ............................................................. 211 Non-committal responses. ................................................................................ 211 Short answers.................................................................................................... 212 Not understanding the purpose of the research. ................................................... 212 The Challenges of Quantitative Reporting Methods ................................................ 213 Distribution of data. .............................................................................................. 213 Variability of reported data............................................................................... 213 Ceiling and floor effects. .................................................................................. 213 Nature of scales. ................................................................................................... 214 Purpose of scales. ............................................................................................. 214 Scale wording used to articulate a construct. ................................................... 215 Measuring the “right” constructs. ..................................................................... 216 Response burden. .................................................................................................. 217 Response hesitation. ............................................................................................. 218 The Challenges of Qualitative Reporting Methods .................................................. 219 Lack of data acquired. .......................................................................................... 219 Self-expression in adolescents. ......................................................................... 219 xv

The Challenges of Intervention Studies in Schools .................................................. 220 Sampling. .............................................................................................................. 220 Sample size. ...................................................................................................... 220 Accessing a control group. ............................................................................... 223 Randomisation. ................................................................................................. 223 Comparability of data between samples. .............................................................. 224 Missing data.......................................................................................................... 224 Missing responses within a data set. ................................................................. 225 Missing full baseline or follow-up.................................................................... 225 Confidentiality in classroom data collection. ....................................................... 225 Conclusion ................................................................................................................ 226 Chapter 10: Critical Reflection Analysis 2 ............................................................... 228 The Programs ............................................................................................................ 228 Comparison of Reported Outcomes ......................................................................... 229 Group Attributes ....................................................................................................... 231 Group size. ............................................................................................................ 231 At-risk status of participants. ................................................................................ 232 Participant engagement. ....................................................................................... 234 Participant understanding of program and its purpose. .................................... 234 Participant commitment to program. ................................................................ 235 Program Attributes ................................................................................................... 236 Setting. .................................................................................................................. 236 Program length. .................................................................................................... 238 Program purpose. .................................................................................................. 238 Wellbeing versus curriculum focus. ................................................................. 239 Targeted to the needs of a coherent group. ....................................................... 239 Facilitation. ........................................................................................................... 240 Facilitator. ......................................................................................................... 240 Participatory and democratic approach. ........................................................... 241 Fostering musical appropriation. .......................................................................... 242 Activities............................................................................................................... 243 Group activities. ............................................................................................... 243 Instrumental activities....................................................................................... 245 Creating a musical product. .............................................................................. 245 xvi

Fun activities. ................................................................................................... 246 Challenging activities. ...................................................................................... 247 Non-musical activities. ..................................................................................... 247 Musical skills not required. .................................................................................. 248 Conclusion ................................................................................................................ 248 Chapter 11: Recommendations Article .................................................................... 251 Chapter 12: Discussion............................................................................................... 275 The Challenges of Existing Research Approaches ................................................... 275 Recommendations for Future Research .................................................................... 277 Research method and design. ............................................................................... 277 Type of musical participation investigated........................................................... 278 Need for ongoing critical engagement with research. .......................................... 278 Policy Implications ................................................................................................... 280 Challenge to claims in current policy literature. .................................................. 280 Challenge to claims of psychosocial benefits. .................................................. 280 Challenge to recommendations for music education. ....................................... 281 Ongoing support for music in schools. ................................................................. 281 Recommendations for policy. ............................................................................... 283 Limit claims to musical benefits that are supported by evidence. .................... 283 Critical engagement. ......................................................................................... 284 Addressing policy goals. ...................................................................................... 284 Conclusion ................................................................................................................ 285 Chapter 13: Reflections .............................................................................................. 286 Aspirations ................................................................................................................ 286 Prior Experiences ...................................................................................................... 287 Beliefs ....................................................................................................................... 288 Conclusion ................................................................................................................ 290 References.................................................................................................................... 291 Appendix A: Theoretical Approach to Knowledge Accumulation .............................. 321 Appendix B: Keyes to Success Program Protocol........................................................ 325 Appendix C: Information Statement for Students ........................................................ 326 Appendix D: Information Statement for Families ........................................................ 327 Appendix E: Consent Form for Parents ........................................................................ 329 Appendix F: University of Melbourne Ethics Approval letter ..................................... 330 xvii

Appendix G: Department of Education and Early Childhood Development Ethics Approval letter .............................................................................................................. 331 Appendix H: Keys to Success Survey .......................................................................... 333 Appendix I: Keys to Success Scale Scoring Information ............................................. 346 Appendix J: Information Statement or Professionals ................................................... 352 Appendix K: Consent Form or Professionals ............................................................... 354 Appendix L: Bereavement Group Program Protocol ................................................... 355 Appendix M: Bereavement Group Scale Scoring Information .................................... 356 Appendix N: Bereavement Group Survey .................................................................... 357 Appendix O: Bereavement Group Survey Results ....................................................... 369

List of Tables Table 5.1 Methodological Elements of Project Aligning with Participatory and Critical Theory .............................................................................................................. 67 Table 5.2 Definition of Terms Used to Describe Overall PhD Design .......................... 75 Table 6.1 Scales Details for Keys to Success Study....................................................... 88 Table 6.2 Keys to Success Study Semi-structured Interview Schedules ........................ 93 Table 6.3 Scales Details for Bereavement Group Study .............................................. 101 Table 6.4 Bereavement Group Study Semi-structured Interview Schedule ................. 104 Table 7.1 Protective Factors: Mean Change and Difference of Mean Scores .............. 108 Table 7.2 Risk Factors: Mean Change and Difference of Mean Scores....................... 108 Table 7.3 Keys to Success Student Interviews: Codes, Themes and Categories ......... 129 Table 7.4 Raw Data and Codes Used to Form the “Engagement With Learning” Theme ........................................................................................................................ 137 Table 7.5 Raw Data and Codes Used to Form the “It Was Just About Having Fun” Theme ............................................................................................................ 144 Table 7.6 Raw Data and Codes Used to Form the “Couldn't See the Point” Theme ... 148 Table 7.7 Raw Data and Codes Used to Form the “Responding to What I Said” Theme ........................................................................................................................ 155 Table 7.8 Keys to Success Teacher Interview: Codes, Themes and Categories .......... 156 Table 8.1 Bereavement Group Student Interviews: Codes, Themes and Category ..... 182 Table 8.2 Raw Data and Codes Used to Form the “Social Support” Theme ............... 189 Table 8.3 Raw Data and Codes Used to Form the “Private Group Setting” Theme .... 195

xviii

Table 8.4 Raw Data and Codes Used to Form the “We Missed Class, But it was Worth it” Theme ....................................................................................................... 198 Table 9.1 CRA 1 Challenges and Sub-challenges ........................................................ 227 Table 10.1 Reported Psychosocial Wellbeing Themes with Rate of Endorsement ..... 231 Table 10.2 Music Program Elements that Affect Investigations of Psychosocial Wellbeing ....................................................................................................... 250

List of Figures Figure 5.1 Flow chart of overall project design, showing the two stages of analysis and the stage of reflexive evaluation. ..................................................................... 73 Figure 7.1 Mean scores at baseline and confidence intervals by group for Community domain............................................................................................................ 109 Figure 7.2 Mean change scores and confidence intervals by group for Community domain............................................................................................................ 110 Figure 7.3 Mean scores at baseline and confidence intervals by group for School domain............................................................................................................ 111 Figure 7.4 Mean change scores and confidence intervals by group for School domain ........................................................................................................................ 112 Figure 7.5 Mean scores at baseline and confidence intervals by group for Peer and individual domain .......................................................................................... 113 Figure 7.6 Mean change scores and confidence intervals by group for Peer and individual domain .......................................................................................... 114 Figure 7.7 Mean scores at baseline and confidence intervals by group for Emotional control domain ............................................................................................... 115 Figure 7.8 Mean change scores and confidence intervals by group for Emotional control domain............................................................................................................ 115 Figure 7.9 Mean scores at baseline and confidence intervals by group for Social support domain............................................................................................................ 116 Figure 7.10 Mean change scores and confidence intervals by group for Social support domain............................................................................................................ 117 Figure 7.11 Mean scores at baseline and confidence intervals by group for Psychological distress domain ....................................................................... 118

xix

Figure 7.12 Mean change scores and confidence intervals by group for Psychological distress domain............................................................................................... 118 Figure 7.13 Mean scores at baseline and confidence intervals by group for Psychological wellbeing domain ................................................................... 119 Figure 7.14 Mean change scores and confidence intervals by group for Psychological wellbeing domain ........................................................................................... 120 Figure 7.15 Mean scores at baseline and confidence intervals by group for Psychological needs domain .......................................................................... 121 Figure 7.16 Mean change scores and confidence intervals by group for Psychological needs domain ................................................................................................. 122 Figure 8.1 Mean scores at baseline and confidence intervals for School domain ........ 170 Figure 8.2 Raw baseline scores for Educational expectation scale .............................. 170 Figure 8.3 Mean change scores and confidence intervals for Academic failure, opportunities for prosocial involvement and rewards for prosocial behaviour ........................................................................................................................ 171 Figure 8.4 Mean change scores and confidence intervals for Low commitment to school and Educational expectation .......................................................................... 172 Figure 8.5 Mean baseline score with confidence intervals for Emotional control ....... 173 Figure 8.6 Mean change score with confidence intervals for Emotional control ......... 173 Figure 8.7 Raw baseline scores for Social support domain.......................................... 174 Figure 8.8 Mean change scores and confidence intervals for Social support domain.. 174 Figure 8.9 Mean baseline, and mean change score with confidence intervals for Psychological distress domain ....................................................................... 175 Figure 8.10 Mean baseline, and mean change score with confidence intervals for Psychological wellbeing domain ................................................................... 176 Figure 8.11 Mean scores at baseline and confidence intervals for Psychological needs domain............................................................................................................ 177 Figure 8.12 Mean change scores and confidence intervals for Psychological needs ... 178

xx

Chapter 1: Introduction This doctoral dissertation explores challenges related to investigating the subjective wellbeing benefits of musical participation in mainstream schools. Recent literature has linked student involvement in school-based music programs to a number of social and psychological wellbeing benefits. This project interrogates such literature by reflecting on two studies that aimed to demonstrate this link, and critically analysing the research methods and nature of musical participation involved in each. Findings of this analysis critique existing approaches to research used in this area, and the knowledge they are said to generate. This includes a challenge to the idea that established and widely used methods for collecting student self-report data, such as class-based surveys, are appropriate for conducting research in this context. Findings also question the assumption that all forms of school-based musical participation will achieve psychosocial wellbeing benefits: an assumption that appears to be embedded in both research and policy documentation in this field. The challenges of undertaking investigation in this area, identified through critical reflection analysis, are used to inform recommendations for the design of future studies, and future policy. Music, Schools, Policy and Research: A Brief Background Musical participation is broad term which describes an individual’s involvement in musical activity. This may include formal private instruction, informal membership in a band or choir, studying music theory, listening to recorded or live music performances, attending a music therapy session, or a range of other music-based programs, games, or activities (Pitts, 2005). Advocacy for the role of musical participation in mainstream schools has grown considerably in Australia over the last 15 years, and now provides a central theme among key documents in the fields of education and arts policy (Ewing, 2010). Embedded within this literature are explicit recommendations that students be provided with more opportunities to partake in musical activities. Much of this advocacy is based on arguments that such participation will lead to the acquisition of benefits in three broad areas (Parliament of Victoria, 2013). These include intrinsic benefits, such as creativity and cultural awareness, as well as extrinsic benefits, which relate to academic and cognitive development. The third group of benefits can also be considered extrinsic, and refer to a raft of subjective

1

experiences including social, emotional, and psychological wellbeing; grouped here under the umbrella term psychosocial wellbeing. Current policy recommendations for increased musical participation in schools indicate a level of political support historically unprecedented in this country (Stevens & Stefanakis, 2014). The nature of this support, however, is both promising and problematic. The fact that policy recommendations endorse a more centralised place for music in all schools provides great opportunity for campaigners and practitioners of school-based music programs. It also offers a promise to students that their time in schools will be enriched by the unique experiences and benefits that musical participation can afford. As someone who’s own school experience was enhanced by music, I know the lasting impact this can have, and am subsequently passionate that this opportunity is realised by as many students as possible. The problematic side of this political advocacy, however, is that certain claims upon which it is based appear to be somewhat uninformed and misguided. The claim that musical participation will increase students’ psychosocial wellbeing has become increasingly centralised in policy recommendations for supporting music in schools. This claim, however, is not supported by consistent evidence. While it is my belief that music can indeed promote student psychosocial wellbeing, as it did for me, research has yet to show conclusive evidence for this relationship. Further, for reasons that will be explained throughout the thesis, the models of musical participation that policy literature suggests will achieve increased psychosocial wellbeing – i.e. class-based music education – appear to be entirely unsuited for this task. This discrepancy between what is claimed in the policy literature and what has been demonstrated through research suggests that policy recommendations regarding the psychosocial wellbeing potential of school-based music are based on assumptions rather than evidence. This has serious implications for ongoing policy-based support in this area. The relationship between policy and school-based music provision has long been tenuous in this country, with several attempts to integrate music into mainstream education overruled by the government in the last 50 years (Ewing, 2010). As such, the idea that current advocacy is based on unsubstantiated foundations endangers the sustainability of this support. If recommendations are put in place but do not achieve the intended benefits, this will likely destabilise future political backing in this area. It is my opinion that in order to rectify this, it is necessary to challenge and deconstruct the 2

assumptions that underpin this advocacy before they compromise the future of school music in this country. To ensure that the advocacy and recommendations contained within policy literature are best placed to promulgate sustained support for music in schools, it is necessary to undertake further research regarding the link between psychosocial wellbeing and music in this context. This is needed both to provide accurate information to inform policymakers, but also to check to usefulness and legitimacy of current claims. For example, only through research can we interrogate claims made in key existing documents that “music education uniquely contributes to the emotional [and] social […] growth of all students” (Australian Government, 2005, p. 5). However, much like policy recommendations in this area, research that has investigated the benefits of music in mainstream schools also appears to be affected by a number of assumptions. As mentioned, findings regarding the psychosocial benefits of school-based music are inconclusive. Not only are results between and within studies often contradictory (McFerran, Roberts, & O'Grady, 2010), several authors report no evidence of this relationship at all (Rickard, Bambrick, & Gill, 2012; Schellenberg, 2004). In carefully reviewing this literature, it became evident that these inconsistent or inconclusive results were largely due to methodological challenges. These challenges are observable in two main areas. The first involves being able to capture the relationship between school-based musical participation and psychosocial wellbeing using certain methods. The second relates to whether the types of musical participation investigated are able to achieve psychosocial outcomes. For example, using quantitative self-report measures have been reported as particularly problematic (Knox-Anderson & Rickard, 2007), as has been the investigation of class-based music training programs (Rickard, Bambrick, et al., 2012). Despite these reported limitations, researchers have persisted in using the same study designs, and have continued to garner similarly inconsistent results. This drew me to the conclusion that, despite previously reported limitations, researchers were continuing to work on the assumption that certain research approaches are appropriate for this field of research. I thought then, and would argue now, that these approaches are not suitable (for reasons which will be articulated in detail throughout this dissertation). Further, it is the assumption that they are suitable which has promulgated the inconclusive body of evidence regarding a link between music in schools and psychosocial wellbeing.

3

Therefore, I believe that in order to address the assumptions underlying problematic claims in current policy literature, it is first necessary to address the presuppositions underpinning research in this area. Only when this is done will we generate empirical or evidence-based knowledge that is suitable to inform effective policy recommendations. I would not have come to this conviction had I not aimed to provide empirical evidence for the psychosocial wellbeing benefits of musical participation in mainstream schools myself. I too was influenced by assumptions similar to those above. It was only through a process of frustration and reflexive contemplation that occurred during my candidature that I was able to determine where and how my own presuppositions, and those which seem to be have been endemic in this field for over a decade, had affected my research. As such, this dissertation tells the story of, and communicates the knowledge generated from, a three-year journey of critical engagement with the field of research and policy surrounding music in mainstream schools, and its role in promoting student psychosocial wellbeing. Following theorists such as Linda Finlay, to tell this story “reflexivity will be used to demonstrate transparency [and as] a critical tool to deconstruct discourse” (2012, p. 319). Discourse here relates to the claims and presuppositions in policy literature, assumptions underpinning preferred research approaches in this area, and the effect of my own beliefs and motivations. This starts here with a transparent account of my own research journey, including why this PhD was undertaken and how it changed throughout my candidature to form the critically reflexive project it became. The aim of this account is to communicate the role that beliefs, intentions, and experiences have played in driving and re-shaping my PhD project. This, in turn, is a conscious decision to engage with how these factors can have an impact on the generation and communication of knowledge. Such engagement is important in any field, yet I believe it is particularly critical for negotiating the challenges in this area of research. It is also crucial for negotiating between both the potential and the risk presented by current policy documentation surrounding music in schools in Australia, and internationally. The PhD “Journey” Stanley (2014) suggests that writing a PhD is a “journey”. She describes how a rollercoaster of emotions can be experienced, oscillating between tragedy and triumph, elation and frustration, diligence and despair. Directions change, and motivations evolve. 4

This is certainly an apt description of my own experience. This journey, however, has been a valuable voyage of discovery through which my thinking has crystallised, enabling me to reach conclusions I feel are both informative and inherently valuable for both research and policy surrounding music in schools. Life experiences informing this PhD. Although my candidature is firmly rooted in the fields of music therapy and social policy (administratively located in both university departments), I am neither a music therapist nor have I been actively engaged in policy development. My choice to undertake study in these areas was influenced by four sets of experiences that had played dominant roles throughout my life, and which came together to inform my decision to pursue this PhD candidature. First and foremost, this decision was powerfully effected by my ongoing relationship with music and the experiences it precipitated. Starting early in primary 1 school, this relationship grew through learning guitar and performing in small community concerts. By secondary 2 school I was playing guitar, bass (and sometimes singing) in rock, punk and jazz bands. Later I worked as a DJ, playing hip-hop and dance music throughout the pubs and nightclubs of Melbourne. In my spare time I have run club nights, co-founded a record store, and I continue to experiment in music production. Throughout these experiences I have become acutely aware that by engaging with music through listening, performing, writing and producing, I was able to forge friendships, connect with groups, and develop a sense of meaning and wellbeing in my life. I could also see the positive effect music had on other individuals, and whole communities. The decision to undertake this PhD was also influenced by experiences which arose throughout my academic career. I started this career with a Bachelor of Social Science, majoring in both sociology and behavioural science, and later completed an honours degree in sociology and anthropology. Throughout these tertiary degrees, I found myself constantly gravitating towards music. My undergraduate assignments had focused on the role of music in society whenever it was possible. As the first

1

In the Australian context, “primary school” is akin to “elementary school” in

countries such as the US, and refers to the first seven years of formal schooling. 2

“Secondary school” refers to the last six years of formal schooling in Australia. 5

opportunity to design and conduct my own research, my honours degree project investigated the ethos of independent music production. Through these study opportunities I deepened my appreciation for music by learning more about its social role and its potential for affecting wellbeing at an academic level. I became increasingly captivated by the research and theory evolving in these areas, which in turn cultivated an interest in the field of music therapy. After graduation I entered the field of professional research, 3 and soon moved to a research position at a Centre for Adolescent Health located at the Royal Children’s Hospital in Melbourne. My experiences in this position influenced my thinking in three important ways. First, I developed a deep passion for supporting the wellbeing of young people. Second, I recognised the central role research plays in informing the development of effective health and wellbeing interventions for this population, and came to realise I enjoyed undertaking it. The third effect on my thinking was a growing scepticism of the preferred research methods and approaches that were both widely used and accepted for generating knowledge in this area. Among several more diverse roles, my time was often spent piloting and administering large, hour-long surveys to collect population and mental health data from students. This regularly entailed travelling to multiple schools in a week to administer surveys in classroom settings. As an “on the ground” research assistant, I experienced firsthand the challenges this approach involved for capturing meaningful data. However, while this gave me insight into this area, I was not in a position to act on it. The fourth set of experiences that informed my decision to undertake this PhD derived from both my undergraduate study, as well as my time as a professional researcher. While studying sociology I gained an understanding of the relationship between theory, research, policy, and social wellbeing. I became aware of the enormous potential that social theory and research have to influence our society, particularly in their ability to inform policy. Working at the hospital deepened my understanding of this relationship in the more localised area of young people’s health and wellbeing. Research projects were strongly geared towards influencing policy development in this field, or gaining government funding for interventions. I saw that by designing studies

3

Following the distinction made by Beck and Young (2005), “professional

research” is used here to denote research undertaken outside university settings. 6

and disseminating findings with certain policies in mind, research can be better positioned to actively influence society, and the lives of people within it. Beginning the PhD. My decision to cease professional research and pursue a PhD stemmed from two driving motivations, both of which were born out of the experiences described above. The first was a determination to consolidate the link between my musical and professional selves by combining my love and passion for music with my interest in research and the social sciences. The second was a reaction to the dominance of positivist epistemology and methodology in my professional workplace. I wanted to address my growing scepticism of the methodologies associated with this epistemological stance by engaging in more humanistic, or qualitative research practices. The music therapy research unit at the University of Melbourne provided an optimal context for integrating both of these goals, particularly given the emphasis on adolescent focused research within the unit. I proposed that by applying my sociological background and professional research experience to the field of music therapy, I could bridge the gap between the theory and expertise of this field and the larger social issues of our time. At the time of applying for candidature, I was heavily engaged with the work of community music therapy theorists such as Brynjulf Stige (2010) and Gary Ansdell (2004). I felt their theories regarding the use of music in community settings could be used to articulate the benefits of a community-based music program being run by acquaintances met through the hospital, and which was in need of ongoing funding. I decided that by evaluating this program using a community music therapy model, I could demonstrate the social wellbeing benefits of community-based music programs to the fields of population health, sociology and policy. I thought, these fields have so many important social and community issues they are trying to address, yet they are seemingly unaware of the work being done in music therapy, and the relevance and transferability of this work to their own agendas. 4 For me, this was a way of starting an interdisciplinary dialogue that would both enable these more traditional disciplines to benefit from the theory and practice of music therapy, and facilitate greater recognition,

4

Italics are used to denote my own personal reflections on a situation or event. 7

support, and integration of music therapy into other fields. This is a two-way exchange I thought, everyone wins, I just need to start the conversation. Wanting to ensure I maximised my contribution to the field of music therapy, as well as the music program I intended to investigate, I also decided, I need to investigate this in a way that is directed to relevant areas of policy. Then I can convince policymakers to support the use of music as a tool for social wellbeing, and get them to provide funding for music programs in the community. I sought additional academic supervision from the school of social and political science, and through consultation decided to focus on the suite of policies and goals associated with social inclusion. This was the paradigmatic approach to social policy adopted by the Australian government at the time, and one which embodied many of the social and subjective themes relevant to community music therapy, such as belonging, connectedness, and opportunities for social participation (Saunders, 2011). At this point, I was clear about the goal of the PhD, and how it was to be achieved: I would use qualitative methods to explore the potential for a communitybased music program to facilitate social inclusion in a group of young people. I would investigate and frame outcomes using the indicators of social inclusion set as priorities by government, and articulate the processes in which these were achieved using music therapy theory and terms. I perceived, if I can demonstrate the potential of music therapy-informed programs to meet social wellbeing objectives currently prioritised in government agendas – using policy language, concepts and indicators – surely government cannot refuse supporting these programs. Upon reflection, however, this initial perception was built upon a number of presumptions. Most significantly, this included the assumption that the link between music and social wellbeing was so obvious that it could be demonstrated regardless of the research approach used. Such presumptions, however, were to be challenged throughout various stages of my candidature. Three critical decisions. During the initial stages of my candidature both circumstance and deeper engagement with literature pushed me to make a number of critical decisions that changed my initial focus and design. The first of these involved a growing awareness of claims that policymakers are more receptive to quantitative data (Torrance, 2011). I had originally aimed to use a participatory or constructivist research design, a decision 8

driven by my scepticism of positivist and post-positivist methodologies that had developed during my professional roles. However, adopting the pragmatic approach to research advocated by Biesta (2010), I decided that if I was to achieve my goal of influencing policy, then collecting quantitative evidence was required. The second critical decision involved locating a music program and participant group for my study. The community-based music program I wanted to investigate had since ceased to operate. Given I am neither a music therapist nor community music program facilitator of any kind, this meant my options were limited. In discussion with my primary supervisor, she offered me data from two school-based music programs (conducted as part of a larger school music initiative she was spearheading) that I could use for my analysis, if I helped collect it. After reading education literature I quickly saw the importance and potential of investigating this area. Not only are schools in themselves an important community group and setting, significant debates were being undertaken in the area of education policy regarding the role of music and the Arts (Ewing, 2010), to which I felt both my skills and music therapy theory could contribute. I also remembered the positive role music had played in my life during primary and secondary school, and it was easy to become enthusiastic about contributing to research that advocated for greater policy support for music programs in this context. Further, I had worked in and was passionate about the area of student wellbeing, which also meant I was able to quickly engage with the area. Lastly, there were several more practical benefits to this approach: students present an accessible participant population; I was accustomed to collecting data in schools; this data would be quantitative which is what I had decided was needed to inform policy; and, I could simply be added to an existing ethics application. The third decision that marked a departure from my original design was the choice of constructs used to investigate social wellbeing. I had originally intended to explore the classic sociological concepts of social wellbeing, such as alienation (Seeman, 1983) or anomie (Bell, 1957). Instead, I was to be collecting data using tools that focused on constructs such as school engagement and psychological wellbeing. However, social wellbeing measures were included, and the school engagement measures were consistent with descriptions of student wellbeing in policy literature. Theorists I was reading at the time also claimed psychological and social wellbeing are inextricably linked, and had proposed the combined concept of “psychosocial wellbeing” as more useful (Keyes, 2005; Larson, 1996). Furthermore, these measures seemed 9

consistent with government indicators of social inclusion, the policy concept to which I aimed to appeal. Further, my conviction that the (psycho)social wellbeing benefits of musical participation were so obvious led to a confidence that they would become apparent regardless of the specific constructs or tools used to capture them. Moments of dissonance. Once engaged in carrying out my updated research design, I gradually became aware of several discordant themes arising in both the literature in this area, and the data collection activities I was undertaking. The first of these was a pervasive sense of doubt that the students we were surveying were either taking their participation in data collection seriously, or understood the questions we were asking them. My reservations about using surveys in schools from my professional research days resurfaced. I wondered, can we really trust this data? These kids aren’t engaged with this at all, and discussed this at length with my primary supervisor. However, I resolved to persevere. Drawing again from a pragmatic mindset I decided that I did not have to subscribe to this type of research, I was merely using it to collect the data needed to argue the psychosocial wellbeing benefits of music to policymakers – the ends justified the means. A sense of dissonance also emerged in my reading of the literature. Researchers who had also aimed to provide quantifiable evidence for the psychosocial wellbeing benefits of student musical participation had struggled. While there was some support for this link, much of it was either inconsistent (Knox-Anderson & Rickard, 2007), contradictory (Kim et al., 2006), or not in the group settings I was interested in. Other studies had found no evidence for these benefits at all (Rickard, Bambrick, et al., 2012). After a thorough review of the field, it became apparent this was a complicated area of study. While inconsistency of findings is normal in most fields, this area appeared to be characterised by its ambiguity. The only consistent message I could deduce from this literature was that there existed a vast array of challenges inherent to this area of research. Further, while there appeared to be some recurring patterns in these challenges – i.e. that they revolved around the methodological approach used, or the type of music participation investigated – there seemed to be little to no in-depth discussion or ongoing consideration of the impact of these challenges on the ability to provide consistent findings. Despite these observations, I again continued on using the rationale: perhaps they weren’t doing it right. Surely I can get the evidence needed. This existing ambiguity only makes my study more necessary, and the results more important. 10

The most fundamental source of dissonance, however, came directly from my initial analysis of the data. The majority of data was collected within the first six months of candidature, and it was not telling the story I expected. In fact, preliminary analysis of scale results showed no benefits at all. This feels ridiculous, I thought, I’m trying to fit a square peg in a round hole; not only do I doubt the validity of the methods I’m using, they are giving me nothing! Am I just wasting time? Sidestepping these initial reservations, I quickly moved to reconcile the issue by including qualitative data in my project. Interviews had already been conducted with one participant group (for program evaluation purposes) so I sought permission to conduct follow-up interviews for the second program I was investigating. Great, I thought, I can make this a mixed-methods study, use it as a pilot, and collect more data from a third program. Surely those results will provide some clarity. This updated research design became incredibly complex. It included multiple analyses on five data sets from three programs across two phases. Presenting this design at postgraduate seminars was met with silence, closely followed by a volley of questions from confused colleagues about what it was I was trying to achieve, and the necessity of such a complicated design. Frustrated and well into my candidature, I started to have serious doubts: what am I doing? This is all wrong, I either need to start again, or make some radical changes. Again, however, I pushed these thoughts aside with the rationalisation; it’s complex, I know that, but it makes sense to me and it’s not my fault they don't understand. Somewhat begrudgingly, I carried on. Change in direction. After several months of trying to further clarify the argument, direction and design of my study, my university advisory panel indicated that this had not been achieved. Afterwards I had a forthright discussion with my primary supervisor, which proved to be a major turning point. I realised I was not going to reconcile the issues of my study design, or my own reservations (at least not in a way that was going to be satisfactory for me). And, as my supervisor suggested, perhaps that was the point. The very fact that this project wasn't working appeared to represent the knowledge it generated. Throughout my candidature, much of the supervision sessions with my primary supervisor were spent reflecting on our experiences of and thoughts about investigating music in mainstream schools (and research in this setting more generally). This 11

involved critiquing the assumption that using surveys to conduct research in schools is a reliable way of collecting data. We also questioned the expectation that musical participation of any kind will lead to wellbeing benefits. Applying this critique to my own project, I realised that it was these assumptions that had crept into my own work, and which were the source of my frustration. This is the real story of my PhD, I thought, this is where my learning has occurred, and this is what I need to communicate to others. Taking a day to process this revelation, I realised, I don't need to collect more data; all the information I need is in the data I have. The fact this data didn't provide evidence of the psychosocial wellbeing benefits of school musical participation was the result. It became clear the best way I could contribute to this area was not to “prove” the “power of music” to policy makers. What I needed to do was articulate the challenges and inherent contradictions embedded within attempts to provide this proof. It was time to address the “elephant in the room” – the one that was staring back at me from the literature, my experience of conducting research in schools, and the very data I had collected for this project. I came to the conclusion: the reason no one has found consistent benefits in this area is because the way they have been investigating it doesn't work. The fact scholars had not properly acknowledged or critically engaged with this possibility, but rather continued to use established research approaches to reach the same level of inconsistency in their results, suggested to me, these researchers are just assuming that their approaches are valid. If no one challenges these assumptions and investigates whether or not they really are appropriate for this field, the field is not going to get anywhere. Further, I thought, the fact that policymakers are already making recommendations in this area is a problem. Not only do they not have the evidence to make them, but also, they seem to be based on unfounded assumptions about the potential of certain school-based music activities. If they keep making these claims its not going to do anyone any good. It became clear that a critical analysis which identified the challenges of investigating the psychosocial benefits of school-based musical participation was crucially important from both scholarly and policy perspectives. First, by identifying the challenges of research in this field, I could make recommendations for future studies. This would not only serve to further scholarly knowledge, but also hope to inform better quality research capable of reaching a consensus on if, how, and when music can 12

promote psychosocial wellbeing in schools. From a policy perspective, I felt that such a critical analysis would be useful on two levels. First, it would inform policymakers (and schools) of the shortcomings of the recommendations in existing policy documents, and the potential counter productivity of implementing them. Second, if this analysis does lead to better research, then the subsequent evidence will mean policymakers are better placed to make new recommendations. I also felt well positioned to undertake such an analysis. I argue that my status as an “outsider” in this area gave me the necessary insight and objectivity to do so: I don't have any vested interest in music therapy as a discipline – I’m not a music therapist, or a music practitioner of any sort. What’s more, I’ve been out there, I’ve spent years collecting data from young people in school contexts. I’ve seen that kids lie, make things up, don't concentrate, and don't understand survey questions. Also, I’ve seen that, despite our best intentions, we are often good at ignoring or rationalising these things, convincing ourselves that we are using established techniques, validated tools, and that the truth is in the data. Further, I’ve been a student, I know I didn't take surveys seriously. I also know that while music was an important part of my life as a student, I didn't always “feel better” after music class. Lastly, it is critical approaches like these that my previous years at university had prepared me for. Following in the footsteps of Marx, Horkheimer, Adorno and Mills, when deciding to undertake the critical analysis presented in this dissertation, I thought, if my sociology background has taught me anything, it is to be critical. Current Study Aim and Thesis Structure. As a result of the journey just described, this thesis presents a critical, reflective, and reflexive analysis of empirical attempts to demonstrate a link between musical participation in mainstream schools and increased student psychosocial wellbeing. It aims to both explore and report the challenges that surround the investigation of these benefits. It then aims to used these challenges to inform recommendations for further research in this area, as well as policy development. To communicate this project in a way that best conveys both the originality, complexity and innovation it contains, a unique approach to dissertation structure and presentation has been taken. As such, this thesis is constituted by a captivating assortment of chapters. One of these has been published, according to the “thesis with publication” guidelines at the University of Melbourne, and the others have been 13

divided into useful groups. These groupings, and the order in which they are placed, have been carefully constructed with a view to enabling the reader being to best engage with the complex array of materials being provided. In the following three chapters, key literature relevant to this dissertation is introduced and discussed. The first of these covers the policy context surrounding the use of music in schools, giving a brief overview of the history of this area, the current situation, and how this has been informed by larger policy trends on global and national levels. Critical arguments regarding the sustainability of current policy support made by Australian authors are also presented. The second of these literature review chapters further explores some of these critical arguments by interrogating current policy claims against existing evidence, and highlighting the gaps and need for further study. The third chapter briefly offers an operational definition of psychosocial wellbeing, and then reviews existing published research that has investigated a link between students’ psychosocial wellbeing and musical participation. In doing this it identifies some common justifications for the inconsistency of findings in this area offered by the authors of this research, so as to further refine the focus and aim of the PhD project. Following this grouping of literature review chapters, two chapters are used to communicate the methodological elements of the project. The first of these begins by stating the aim and main research questions. It then gives an account of the philosophical and theoretical perspectives underpinning the project in order to locate the PhD within the theoretical landscape of social sciences, and explain my philosophical approach to the generation of knowledge. In this chapter I also present the overall structure of the research project, which includes an explanation of the different research activities that have contributed to this dissertation. This also involves a description of how critical reflection has been used as analysis technique. The second methodological chapter presents a more conventional account of the methods used in the two schoolbased empirical studies. The next grouping presents the subsequent results from these two empirical studies, each in a respective chapter. Rather than an attempt to provide empirical evidence for the psychosocial impact of musical participation in schools, the purpose of this presentation is to communicate information that is used to inform the two critical reflection analyses that follow. These critical reflection analyses are each given a chapter, and represent the next grouping. The first uses information from both studies (presented in the previous 14

chapters) to identify several challenges of investigating the psychosocial wellbeing benefits of musical participation that relate to the methodology. The second compares reported psychosocial wellbeing benefits to the attributes of the music programs investigated in each study. This is done to identify how different porgram elements relate to the reporting of benefits. Both chapters discuss results in relation to relevant literature. Following the presentation of the critical reflection analysis, the challenges identified in each are used to make a number of recommendations for the design of future research projects. Included in these recommendations are suggestions for the attributes of a music program that should be pursued, and those that should be avoided, when investigating this area. These recommendations have been published, and therefore this chapter is presented as the published journal article. Given the article focuses largely on recommendations for the design of future research, it is followed by a more general discussion chapter that also discusses the policy implications of the project findings, and recommendations for policymakers. Finally, I undertake a reflexive conclusion in which I reflect on the PhD candidature, and how it has affected the experiences, aspirations, and expectations that inspired this journey, and which have been described in this reflexive introduction. I also discuss how the project has influenced my own beliefs and assumptions, and reflect on the key points that I myself have learned through this candidature.

15

Chapter 2: Music in Schools and Policy Policy support for music in mainstream education has grown considerably in Australia over the last decade (ACARA, 2011; Australian Government, 2005), with current policy documentation recommending that all students be given the opportunity to participate in music at school (Parliament of Victoria, 2013). While a promising development, there is recognition from the Arts sector that the practical application of this support is yet to be realised (Ewing, 2010; Stevens & Stefanakis, 2014), and that there are no guarantees it will be (Gattenhof, 2009). Academics note that to further support and develop advocacy in this area, more research is needed to better understand how and when music can make a contribution in educational contexts (Stevens & Stefanakis, 2014). To fully understand the importance of this research, and help direct its development, it is necessary to consider the context of current policy support, including both the history of policy in this area, and the wider influences that have shaped it. This enables the development of research which can pinpoint identified weaknesses, while also capitalising on the strengths of existing policy. An approach to investigation that is informed in this way would likely be more effective in producing the kind of knowledge needed to further support the practical manifestation and maintenance of advocacy and support for music in schools. This chapter locates current policy support for school-based musical participation in the context of historical and contemporary debates in this area, as well as developments in wider fields of policy. It does this by providing a historical account of the relationship between policy and school-based arts participation in Australia, as well as explaining how the recent rise in support for such participation in this country can be understood as part of as a global push for the Arts in education. The link between this push and developments in areas of social policy and educational policy at both local and international levels is also described to offer an account of the factors that have underpinned and driven contemporary support for the Arts in mainstream education. Having established these contextual factors, key arguments are then presented which highlight weaknesses in current policy support, and which can provide signposts for directing the kind of research needed to address these weaknesses, and support the maintenance and application of advocacy in this area.

16

Setting the Scene Policy and music in Australian schools: a short history. Deliberations regarding the place of music in Australian schooling are reported to be as old as the country’s schooling system itself (Stevens, 1978). Dating back to 1872, the year when Australia’s first curriculum was established in Victoria, debates raged in England and Australia as to whether education models should focus on a narrow agenda of vocational-based subjects, or include a broader range of activities (O'Toole, 2010). “Poorly drafted and lacking in detail” (Bessant, 1984, p. 5), this first piece of legislation regarding compulsory education in Australia focused on literacy, numeracy and geography (Government of Victoria, 1872), thus excluding all artistic endeavours from a government supported education. Despite the Universal Declaration of Human rights (General Assembly of the United Nations, 1948), which was said to champion the place of the Arts in the curriculums of many Western nations (Ewing, 2010), the Arts became increasingly marginalised in most Western education systems throughout the 20th century. This has largely been attributed to the fact government funding for education intensified its focus on the “core” disciplines of maths, sciences and literacy (Ewing, 2010; O'Toole, 2009, 2010; Stevens, 1978). The largely token role that art played in public schooling throughout Australia was further side-lined by the global shift to neoliberal models of education during the 1970’s and 1980’s (Ewing, 2010), which were seen to value only those subjects which directly contributed to national economic prosperity (Cranston, Kimber, Mulford, Reid, & Keating, 2010). Notwithstanding declining support from government, the 1980’s saw a number of valiant efforts to bolster the status of the Arts in our schools through policy. At a national conference, Ruth Buxton (1981), then Principal Education Officer of the Education Department of South Australia, called for national policy reform which placed music at the core of the curriculum for all Australian children. Likewise, a research report undertaken in collaboration with the Education Department of Tasmania argued for the potential to motivate and engage students in education through a pedagogy of arts-based programs (Parsons, Schaffner, Little, & Felton, 1984). Towards the end of this decade, the National Affiliation of Arts Educators was formed in order to rally for the position of arts education as one of the seven key learning areas in the Australian curriculum. This goal was realised in 1992, yet largely abandoned shortly

17

after, again due largely to the government’s increased focus on school as site for vocational training (Ewing, 2010). In 1996, a report was released that argued for the inclusion of arts practices in education based on their ability to augment a student’s employability through the development of key vocational skills and competencies (Bryce, Harvey-Beavis, Livermore, & O’Toole, 1996). However, for the most part, the 1990’s saw a period in which a “death” of engaged literature relating to the Arts in education was experienced (Grierson, 2007). As Ewing puts it: [The] understanding that art needed to be at the centre of any society had thus been greatly diminished over two centuries ago. This devaluing continues to be reflected in the lack of a strong presence of government support for the Arts and of support for the role of the Arts in the formal and compulsory Australian curricula. (2010, p. 20) The above literature tells a story in which music has largely been relegated to the sidelines of Australian mainstream education by policies and legislation that have prioritised vocational training. Contemporary shifts towards advocacy for the Arts in education. While the history of policy support for the Arts in Australian education throughout the 19th and 20th Centuries tells a demotivating story, so far the 21st Century has seen a push from government for the inclusion of such activities in schools. This push can be seen as part of a global call for the need to reconnect with the positive impacts of the Arts in education, which gained traction around the turn of the century. Such calls were evident in UNESCO’s (1999) International appeal for the promotion of the Arts in education, and various subsequent publications such as Bamford’s (2006) The wow factor: Global research compendium on the impact of the arts in education, and Barhi’s (2006) Educating Through Art in Secondary Education. These documents, and others (see Ewing, 2010; UNESCO, 2006), outline policy calls at the international level for the need to include the Arts in education on the grounds of everything from quality education provision, to economic stability, wellbeing, and world peace. In Australia, government recognition of the above became evident throughout the first decade of this century. This started at the state level, with reports such as ARTSsmart, commissioned by the South Australian government (Department of

18

Education & Children’s Services, 2002), which espoused the role of the Arts in the intellectual and cultural development of children. This reached the national consciousness in 2005 with a research report undertaken for the Australian Government’s Australian Council for the Arts (ACA; Hunter, 2005) that argued the importance of including the Arts in mainstream education. It reported a range of positive education and wellbeing outcomes for young people and their communities, and recommended the implementation and support of artistic activities in all Australian schools. This call was reinforced in 2007 with a statement jointly released by Education and the Arts ministers from both state and federal tiers, as well as the Ministerial Council for Education, Employment, Training and Youth Affairs (MCEETYA) and the Cultural Minsters Council (CMC). This statement articulated that “all children and young people should have a high quality arts education in every phase of learning” (MCEETYA & CMC, 2007, p. 5), based on the claim that: An education rich in creative arts […] is vital to students’ success as individuals and as members of society, emphasising not only creativity and innovation, but also the values of broad cultural understanding and social harmony that the arts can engender. (CMC & MCEETYA, 2007, p. 3) In 2009, the Australian Commonwealth Minister for Arts, Peter Garret, issued a statement suggesting that the Arts have an important role to play in meeting the demand on students to gain creativity, innovation and interpretation skills, as well as develop a level of cultural understanding necessary for many careers in today’s economy (Garrett, 2009). In the same year, the Victorian Department of Education and Early Childhood Development (DEECD, 2009) released a report further reinforcing the importance for all children to access arts activities through their schooling for the unique development opportunities they provide. A subsequent review by the Australia Council for the Arts (2010) showed this position was also supported by public opinion, reporting increasingly positive attitudes towards the Arts and their role in education among Australians citizens. Current support for music in Australian schools. While the above documents show support for the Arts in general, the Australian Government’s Department of Education, Science and Training’s (DEST) landmark

19

report, National Review of School Music Education (NRSME), focused specifically on the role of music in schools (Australian Government, 2005). It concluded that music is not just beneficial for students, but “essential”. It went on to claim that, while music activities in mainstream education have numerous benefits, due to issues of access equity and program quality, too many students are deprived of such benefits. As such, they recommended the need to promote the provision of music in schools, and suggest a number of policy developments which would support the equitable provision of quality music programs for all Australian students. As a result of the NRSME report, and the other more general arts-based reports above, a number of strategies have been put in place aimed at providing equitable access to opportunities for musical participation in Australian schools. These include collaborations between government departments and several national arts and education bodies to provide resources for schools to implement music programs (see www.musiceducation.edu.au; and www.moremusictoolkit.org.au). More notably, steps have been taken at the curriculum level to incorporate music into the daily activities of students. This, however, has been a complex area given the recent release of the Review of Funding for Schooling – Final Report (Gonski, 2011), and the subsequent negotiations between state and federal governments to form a consolidated schooling system. Despite this complexity, it is telling that moves to include music in the curriculum are visible at both state and federal levels. Not only do current Victorian curriculum documents contain suggestions regarding the inclusion of music (VCAA, 2012), but the Parliament of Victoria (2013) have recently released the outcomes of their Inquiry into the extent, benefits and potential of music education in Victorian schools, aimed at reviewing current funding and resource provision strategies for music in schools. The “Inquiry” report states “that music is a uniquely important art form and that music education has wide-ranging benefits for students, schools and the wider community” (Parliament of Victoria, 2013, p. xvi). It goes on to advocate greater policy support for music activities in schools, both in the curriculum and as extra-curricula programs. At the federal level, the Australian Curriculum, Assessment and Reporting Authority (ACARA) has released a number of reports to serve as guides for writing a national curriculum, should one be implemented. The Shape of the Australian Curriculum: The Arts, as the name suggests, is a section dedicated solely to how and why each of the five major art forms should be included. It states that including the Arts 20

in education contributes “to the development of a vibrant, modern and inclusive Australian society” (ACARA, 2011, p. 3), continuing that: [The] Australian Curriculum for the Arts will be based on the assumption that all young Australians are entitled to engage with the five Arts subjects and should be given an opportunity to experience the special knowledge and skills base of each. (ACARA, 2011, p. 4) The above policy documents illustrate the current level of support for music in schools is strong. While scholars note a significant gap between this policy rhetoric and the reality of school music provision remains (Crooke & McFerran, 2015; McFerran & Crooke, 2014; Stevens & Stefanakis, 2014) – an issue discussed in further detail later in this chapter – from a historical perspective this support can be considered both remarkable and promising. Policy Developments Underpinning Support for the Arts in Education The remarkable shift from 150 plus years of marginalisation in Australian schools to current policy stances that advocate the use of government and public resources to support the Arts in mainstream education can be attributed to larger policy developments. These developments include paradigmatic shifts over recent decades at the level of social policy – described as the philosophical framework which informs the overall aims of a given government (Saunders, 2011) – as well as more localised areas such education policy. Social inclusion and the return of “the social”. Since its beginnings in France during the 1970’s (Humpage, 2006), the social inclusion social policy paradigm has become increasingly influential throughout the Western world (Taket et al., 2009) at both federal (Burchardt, Le Grand, & Piachaud, 1999) and international levels (Saraceno, 2002). For some, social inclusion is a reaction to the failings of the neoliberal paradigm to both articulate and address disadvantage (Saunders, 2011; Smyth, 2010; Wyn, 2009b). For others it also signals a return to classical sociological and economic theories of Weber, Durkheim and Marx regarding post-agrarian societies, the rise of individualism, and alienation in modern and postmodern societies (Levitas, 1996; Wilson, 2006). While social inclusion has been critiqued for its ability to achieve the level of social justice and inclusivity it promises (for discussion see Humpage, 2006; Labonte, 21

2004), for this discussion it is important in that, at a paradigmatic level, it espouses the need for governments to take a multidimensional approach to disadvantage (Kostenko, Scutella, & Wilkins, 2010; Saunders, 2011). This includes attending to the “social” elements of the human condition, rather than focusing solely on economic aspects (Crisp, 2010; Huxley & Thornicroft, 2003; Morrison, 2010). As a result, many Western governments, including Australia, have adopted policy models that depart significantly from the neoliberal ideals of economic rationalism and “the market”, thus adopting more holistic policies that also include a focus on social and subjective elements of wellbeing (Jarman, 2001; Saunders, 2013). While one of the main criticisms of social inclusion is its failure to commit to these subjective and social dimensions – with scholars arguing that governments adopting this model have retained material and objective approaches to practical implementation and measurement (Morrison, 2010) – these factors largely define it from previous social policy paradigms (Saunders, 2011; Taket et al., 2009). Arguably, it is also the presence of these dimensions in key social policy documents, including those published by the Australian Social Inclusion Board (ASIB, 2009, 2012), that have enabled the Arts to emerge as playing an important role in government agendas. Policy authors have been quick to acknowledge the potential for arts participation to align with social inclusion policy goals underpinned by these subjective and social dimensions. This includes claims that engagement in arts activities can achieve social inclusion outcomes through empowering individuals, fostering social connections, healing communities, and providing contexts for addressing social issues (Barraket, 2005; Hall & Thomson, 2007). Specific links have also been drawn between the Arts and social inclusion outcomes for young people, including the potential for participation in the Arts to foster social capital, as well as emotional and social wellbeing (Hampshire & Matthijsse, 2010). Similar claims have been made regarding the ability for arts participation to foster social inclusion goals in educational contexts, including engagement (Barraket, 2005; Kinder & Harland, 2004; O'Connor, 2008), social and emotional wellbeing (Karkou & Glasman, 2004), and inclusive education (Wetz, 2004). For this reason, in places such as the UK, the provision of school-based arts programs has become an important strategy for achieving social inclusion in schools in recent decades (Lynch & Allan, 2006).

22

Some scholars have also made specific reference to the role that music can play in promoting social inclusion in educational settings. This includes claims that musical participation is effective in building a sense of community (Iadeluca & Sangiorgio, 2008), respect for diversity (Odena, 2014), and developing general social skills (Blandford & Duarte, 2004) among students. As such, music participation has been recommended as an important strategy for social inclusion policies which focus on young people in disadvantaged (Garrett, 2010) and educational contexts (Odena, 2007). Education policy and the social role of schooling. While the paradigmatic shift to social inclusion at the social policy level can be seen as enabling the Arts to surface as important for broader government goals, debates in more localised areas have also influenced the current government advocacy for music in schools. The most obvious of these is the field of education policy. Neoliberal education models, which have dominated Western society for several decades, are now recognised both locally (Teese & Polesel, 2003; Vickers, 2005; Wyn, 2009b, 2009c) and globally (Ball, 2006; Fox Piven, 2004; Laberee, 1997; Moses, 2004) as both out-dated and counter-productive for today’s generation of students. This counter-productivity is articulated in a number of ways, yet revolves primarily around the claim neoliberal education policies focus too strongly on an economic model of schooling that (among other things) prioritises vocational training. As such, scholars (Henderson, 2005; MacIntyre, 2008; Vickers, 2005) and government departments (MCEETYA, 1999, 2008) have argued that today’s schools should place equal emphasis on vocational training and fostering the social wellbeing and development of students. This includes supporting student engagement, connectedness and resilience (Holdsworth, Cahill, & Smith, 2003); the acquisition of life skills, pro social values and behaviours; promoting civic (Seddon, 2008) and intercultural understanding; as well as social participation (Jones, 2010). Others suggest schools should promote social connections within the school itself (Blum & Libbey, 2004), including student relationships with peers (Libbey, 2004), staff, and whole school communities to develop protective factors for adolescent wellbeing (Glover, Burns, Butler, & Patton, 1998) related to educational outcomes and health-risk behaviours (Resnick, 2000). Therefore, mediating alienation from schools (Nutbeam, Smith, Moore, & Bauman, 1993), supporting school-based social cohesion (Springer, Parcel, Baumler, & Ross, 2006) and satisfaction (Samdal, Nutbeam, Wold, & Kannas, 23

1998), as well as promoting student connectedness (Bond et al., 2007; Bond, Carlin, Thomas, Rubin, & Patton, 2001; Resnick et al., 1997), belonging (Osterman, 2000), engagement (Bond et al., 2004; Doll & Hess, 2001), and attachment (Marcus & Sanders-Reio, 2001) are now considered critical in education contexts. The above literature marks a critical point in our history, as it places significant importance upon the role of schools in addressing students’ social and subjective wellbeing needs. This signifies a departure from the neoliberal view of schools as a training site for young people to participate in the economy. While such positions do not discount the importance of preparing young people for work (Ewing, 2010; MCEETYA, 2008; Wyn, 2009a, 2009b), their significance lies in the fact that, by taking a holistic approach, education should also aim to take a social role (AIHW, 2010; MCEETYA, 2008). Parallels between the above developments in the field of education and the remergence of “the social” in social policy are discernable. Some suggest endorsement of the social role of education has been pushed by the proliferation of the social inclusion paradigm (Saunders, 2011). Others have drawn direct links between the two (Barraket, 2005; Crisp, 2010; Gorard, Taylor, & Fitz, 2001; Hyde, 2010; Karkou & Glasman, 2004; Lareau & Horvat, 1999; Odena, 2007; Wetz, 2004; Wyn, 2009b), including a number of Australian government documents which list the social goals of education as key indicators of social inclusion (ASIB, 2009, 2010, 2012; Ryan & Sartbayeva, 2011; Scutella & Wilkins, 2010). The effect that both social inclusion and debates regarding the social role of schooling have had on education policy in this country is observable in statements contained within some of the most important Australian education policy documents in recent history. For example, the Melbourne Declaration on Educational Goals for Young Australians, stated that: Schools play a vital role in promoting the intellectual, physical, social, emotional, moral, spiritual and aesthetic development and wellbeing of young Australians, and in ensuring the nation’s ongoing economic prosperity and social cohesion. (MCEETYA, 2008, pp. 04-05) This statement indicates a clear departure from the neoliberal model of education to one that promotes school activities which achieve not only economic and employment goals, but also wellbeing and other forms of social development outcomes. 24

Further – given claims school-based musical participation can promote social (Hallam, 2010), intellectual (Bresler, 2005), aesthetic (Winner & Hetland, 2000), spiritual (Wills, 2011), wellbeing (Vaughan, Harris, & Caldwell, 2011), and economic (Garrett, 2009) student outcomes – it is this shift in emphasis at both social policy and education policy levels that can be seen as the driving force behind the recent policy-based advocacy for the place of music in Australian schools. Sustainability of Support for Music in Schools As evidenced in this chapter thus far, the level of current policy support for music and the Arts in Australian schools is both strong and significant, particularly when viewed from a historical perspective. Further, the fact this support can be seen as driven by a greater global push for the Arts in education; influenced, or even underpinned, by paradigmatic shifts in social policy on national and global levels; and related to global and local developments in the field of education and education policy, suggest that the foundations for this support are solid. When coupled with an increasing proliferation of research focused on providing empirical evidence for the benefits of music in schools – evidenced in part by the exponential growth of postgraduate research studies in this area (Stevens & Stefanakis, 2014) – these factors suggest we are closer than ever to securing a meaningful place for quality music programs in Australian schools through government led legislation and resource support. While the promise of this situation is palpable, observations from several academics intimate that any celebration at this point would be premature. A more discerning view of current support reveals that the need to critically assess our current situation is as crucial as ever. Proponents of this need include Sandra Gattenhof. Appearing at conferences across Australia in 2009, she noted that while reading the National Education and the Arts Statement (MCEETYA & CMC, 2007) and the Melbourne Declaration on Educational Goals for Young Australians (MCEETYA, 2008), “one could presume that the battle [to include the Arts in schools] has been won and put to bed”, in reality, the “battle” had “just begun” (Gattenhof, 2009, p. 3). Evidence this battle is ongoing includes the fact that 10 years after the publication of some of the most influential policy documents to support music in Australian education (Australian Government, 2005; Hunter, 2005), music remains unmandated (Stevens & Stefanakis, 2014) and un-resourced by policy (Crooke & McFerran, 2015; McFerran & Crooke, 2014). Thus it remains widely lacking in our 25

schools (Parliament of Victoria, 2013). For Ewing (2010), this shows that, despite policy rhetoric which supports the role of art in education, in practice little action has been taken to implement arts participation in schools: “The seemingly contradictory gap between rhetoric and reality about the importance of arts [in education] continues” (Ewing, 2010, p. 28). Such claims suggest our history may be a more telling indication of the future of the Arts in Australian schools than recent policy documentation. This is supported by Gattenhof’s (2009) claim of a 10-year cycle in which the Arts, after enjoying brief support from government, are routinely stricken from the list of priorities for mainstream education in Australia. The last turn of this cycle was characterised by the inclusion of the Arts in the 1992 national curriculum; a move which “failed because the issues of training a teacher workforce and resourcing schools appropriately were not addressed” (Gattenhof, 2009, p. 15), and because the government resorted to neoliberal priorities which excluded the Arts (Ewing, 2010). Gattenhof’s cycle can also be seen as part of a larger political cycle in which initiatives, or even paradigms such as social inclusion (Saunders, 2011), are abandoned when seen to fall short of their of promised benefits. These observations suggest that the sustainability of current support in this area is vulnerable. Despite the historical significance of this support, the practical realisation of policy recommendations is far from being a certainty. Calls for a more informed understanding of school music benefits. Despite the legacy of arts and policy in our schools, policymakers’ failure to follow through on recommendations, and the fact that music programs remain largely illusive in many Australian schools, there is reason to be cautiously optimistic. Gattenhof’s 10-year cycle is more than a decade past its due, current support for music in schools is both unique in its breadth and depth, and the most recent documents (Parliament of Victoria, 2013) suggest this support is growing. Indeed, rather than adopting a defeatist attitude, some authors have rallied, voicing the need to continue research in this area to support the realisation of practical policy support. For Ewing, this includes the need to capitalise on the current window of opportunity: With a national arts curriculum imminent in Australia, this is an important moment to build on the paucity of the provision of quality arts education in the

26

past and develop a future coherent body of research to deepen our understanding of learning and the Arts. (Ewing, 2010, p. 29) For Stevens and Stefanakis, this involves the need to augment the legitimacy of claims upon which current recommendations are based in order to strengthen their influence: Despite […] recommendations from successive government reviews and inclusion of the arts in the new Australian Curriculum, music still remains in a highly vulnerable position. While there are many voices advocating better provision […] some arguments rely on subjective opinion and therefore represent unsubstantiated assertion. In order for the benefits of music education to be taken more seriously by governments and education authorities, there is a need for more persuasive arguments that are irrefutable. (Stevens & Stefanakis, 2014, p. 12) In both cases we can see a consensus that if support is to continue, and be converted into sustainable action, there is a need to ensure the implementation of any policies are based on a sound understanding of the potential of the Arts in education. As such, it appears critical at this precarious stage of political advocacy that the numerous benefits attributed to musical participation in schools are realistic, achievable, understood, and perhaps most importantly, demonstrable (Stevens & Stefanakis, 2014). To assess the legitimacy of current claims for music in schools it is necessary that we first undertake a preliminary review of the benefits currently pledged by policy, the models of participation that are said to achieve them, and whether these are supported by existing research. This exercise enables the identification of gaps in existing evidence and provides a focus for research needed to further strengthen support in this area. Indeed, strengthened support is needed to guarantee its sustainability, the probability it will be implemented, the likelihood any implementation will live up to expectations.

27

Chapter 3: Interrogating Policy Claims With Evidence Advocacy for school-based musical participation in Australian policy literature is underpinned by several claims regarding the benefits that students, and society, will gain from this participation. While these claims have been reasonably consistent in their presentation throughout nearly a decade of government funded reports (Australian Government, 2005; Parliament of Victoria, 2013), some academics have questioned their legitimacy and called for more rigorous evidence to support them (Stevens & Stefanakis, 2014). This call to justify policy claims with research is made on several levels. This includes the more general argument that further insight is needed into the potential of music in schools (Ewing, 2010). More pointed arguments include the need for research which helps build an understanding of the particular processes which achieve this potential, and can therefore inform strategic policy approaches to the implementation of music in schools (Crooke & McFerran, 2015; Gill & Rickard, 2012; McFerran & Crooke, 2014). Yet, perhaps one of the most fundamental reasons for interrogating policy claims in this area is to determine their credibility (Stevens & Stefanakis, 2014); for if the claims upon which current advocacy is based are not valid, it seems unrealistic that any resultant policy implementation will achieve expected outcomes. To assess this credibility necessitates a cross-examination of both the benefits expected from musical participation, and the mechanisms or models of participation expected to achieve them. This chapter aims to undertake a broad level interrogation of policy advocacy for school-based musical participation in this country. To do this, it first articulates the main claims that have been used to justify recommendations for more government support in this area, as they have been articulated in the most influential policy documentation released over the last decade. Having established the nature of these claims, it then uses existing research to scrutinise, and determine the degree to which they are supported by evidence. In doing this, it aims to identify any elements of policybased advocacy which appear either unrealistic, or in need of further understanding before they are used to inform policy implementation or legislation. Policy Promises for Music in Schools Government recommendations for the inclusion of music in mainstream schools are based on claims that musical participation will have numerous benefits for students, 28

and society more generally (Parliament of Victoria, 2013). These benefits are primarily aligned with three broad categories found in the majority of academic (Gill & Rickard, 2012), policy (Australian Government, 2005) and education (Taylor, 2008) literature in this area. These include intrinsic benefits, extrinsic benefits related to cognition and academic achievement, and extrinsic benefits related to subjective wellbeing. The intrinsic benefits of musical participation refer to what are considered uniquely “musical” benefits (Ewing, 2010). These include musical competency, but also cover the pleasure and personal growth that can result from practical engagement in music (Gill & Rickard, 2012; Taylor, 2008). Intrinsic benefits have also been linked to the ability for music to encourage engagement with, and develop of empathy for, cultural, social, and environmental issues (Jones, 2010); provide opportunities for selfexpression (McCarthy, Ondaatje, Zakaras, & Brooks, 2004); as well as promote creativity, and a range of other thinking skills (Deasy, 2002; Fiske, 1999) valued in todays knowledge economy (Lebler, 2007; Wyszomirski, 2004). The extrinsic benefits of music are considered the “non-musical” (Gill & Rickard, 2012) or “extra-musical” (Stevens & Stefanakis, 2014) outcomes of musical participation. These are best known in terms of the cognitive (Stewart, 2007) and academic benefits (Vaughan et al., 2011). Cognitive benefits relate to claims that musical training can improve IQ (Schellenberg, 2004), memory (Roden, Kreutz, & Bongard, 2012), and spatial learning (Črnčec, Wilson, & Prior, 2006). Academic benefits refer to assertions that musical participation can increase performance in numerous areas of learning, including numeracy and literacy (Babo, 2004; Gill & Rickard, 2012). Extrinsic benefits of music are also articulated in terms of subjective wellbeing outcomes, often termed under the heading of psychosocial wellbeing (Gill & Rickard, 2012). Such benefits are said to include improved self-esteem (Costa-Giomi, 2004), anger management (Currie, 2004), social, emotional and psychological functioning (Teachout, 2005), engagement with school and the community (Ewing, 2010), and several other mental health outcomes (Karkou & Glasman, 2004). While these intrinsic and extrinsic categories are sometimes noted to be problematic as discrete groupings (Taylor, 2008), these three categories have largely provided the rallying cry for the inclusion of music in schools both in Australia (Australian Government, 2005) and internationally (Jorgensen, 1995). For this reason, these categories provide a useful framework for exploring the outcomes of school music 29

provision expected by and articulated in key policy documentation in Australia. Further, given research literature has used these three categories to conceptualise and investigate the benefits of musical participation in schools (Gill & Rickard, 2012; Stewart, 2007), they provide a workable framework for interrogating claims in policy literature against available research evidence. Intrinsic benefits. The National Review of School Music Education (NRSME; Australian Government, 2005) partly identifies the rationale behind its advocacy for music with the claim that “Music in schools contributes to [the] transmission of cultural heritage and values; and, students’ creativity, identity and capacity for self-expression and satisfaction” (Australian Government, 2005, p. v). This aligns with the literature that describes the intrinsic benefits of music and arts participation (Fiske, 1999). The National Education and the Arts Statement’s claim that “All children and young people should have a high quality arts education in every phase of learning” (MCEETYA & CMC, 2007, p. 5) is based largely on the argument such education is important for facilitating creativity, communication, risk taking, adaptation, innovation and imagination. The statement also stresses the importance of creating active citizens with sufficient cultural understanding. Again this indicates a rationale built on the intrinsic benefits of music related to learning, thinking and knowing the world, and recognition for the economic benefits that may occur at both individual and national levels. The Shape of the Australian Curriculum: The Arts, or “The Shape Paper”, states: The Australian Curriculum for the Arts will be based on the assumption that all young Australians are entitled to engage with the five Arts subjects and should be given an opportunity to experience the special knowledge and skills base of each. (ACARA, 2011, p. 4) This skill and knowledge base is said to include the acquisition of aesthetic knowledge, critical and creative thinking, as well as an understanding of artistic processes and the different historical and cultural contexts of artistic artefacts. The stated purpose of such acquisition and understanding is to learn how they relate to industry and also culture. Such a curriculum is also said to enable self-expression. Again we can see an alignment with intrinsic benefits.

30

Peter Garrett, former Minister for the Environment, Heritage and the Arts, put forth a similar argument in a 2009 media release in which he voiced his support at the prospect of including the Arts in the national curriculum: Creativity, interpretation, innovation and cultural understanding are all soughtafter skills for new and emerging industries in the 21st century. Arts education provides students with the tools to develop these skills. (Garrett, 2009, para. 3) The Inquiry into the extent, benefits and potential of music education in Victorian schools is explicit in its focus on intrinsic benefits: Music should be taught in schools because of its intrinsic value [it] is an art form that can communicate the depth of human emotions and transform the way in which we experience the world. Learning music at school provides children with the foundation to play and enjoy music in their adult life. (Parliament of Victoria, 2013, p. xvi) This “Inquiry” also makes reference to extrinsic benefits related to cultural, vocational and economic outcomes: Music has economic, cultural and social benefits for the wider community including, providing a pathway to employment in a range of occupations; increasing cultural vibrancy; and contributing to greater cross-cultural understanding and appreciation. (Parliament of Victoria, 2013, p. xvii) Again, the Inquiry articulates a rationale which is clearly based on the intrinsic benefits of music, both in terms of the musical outcomes themselves, and their impact on employment opportunities and the society as whole. Extrinsic benefits related to cognitive and academic development. The NRSME also claims that students who take part in music education are more likely to excel in other academic areas, and are more likely to experience cognitive growth and development (Australian Government, 2005). This is consistent with the idea of extrinsic benefits that are related to academic achievement and cognitive development (Gill & Rickard, 2012). Similarly, the National Education and the Arts Statement articulates that, “Over and above the obvious development of individual creativity and self-expression” (MCEETYA & CMC, 2007, p. 4), arts education has the potential to support

31

achievement in a number of academic areas, including science, technology, mathematics, and the Arts. It also talks briefly about cognitive processes. Again this signifies a rationale based on the extrinsic benefits of the Arts related to academic achievement and cognition. The Shape Paper also states that including music and other arts activities in schools will assist learning and capabilities in areas such as literacy and numeracy, as well as competence in using information and communication technologies (ICT; ACARA, 2011). Again this clearly aligns with the extrinsic benefits related to achievement and cognition. The academic and cognitive benefits of the Arts in education were also mentioned by Garret. Again voicing his view of the potential of arts in the curriculum, he refers to international evidence that he suggests “has found arts education is important to the development of young minds and positively influences learning in other areas” (Garrett, 2009, para. 4). The Parliament of Victoria’s Inquiry made a more reserved statement about the expected academic benefits of musical participation in schools: “music education can […] possibly contribute to learning in other subjects areas” (Parliament of Victoria, 2013, p. xvi). It also mentions cognitive benefits, yet these were limited to work with students with disabilities. Extrinsic benefits related to subjective wellbeing. The NRSME report also makes an argument based on the extrinsic benefits of music that are related to subjective, or psychosocial, wellbeing. It asserts that music in schools “contributes to the emotional, physical, [and] social […] growth of all students” (Australian Government, 2005, p. v), and goes on to claim that it has a significant impact on the health and wellbeing of students. The National Education and the Arts Statement also advocates the potential for arts education to promote social harmony and social cohesion by helping individuals become successful in themselves, and as members of society. This is through building “confidence and motivation” and promoting the “wellbeing and life skills of children and young people” (MCEETYA & CMC, 2007, p. 4). Here we can also see a rationale based on extrinsic benefits related to psychosocial wellbeing. Similarly, the Shape Paper’s justification for including the Arts in the curriculum also includes the claim that Arts education will lead to the acquisition of 32

personal and social skills, thus facilitating social competence, emotional development, self-esteem and pro social behaviours (ACARA, 2011). It also claims that such an education will lead to social and community connections. While these benefits are not articulated as clearly as the others, we can see an alignment here with the extrinsic benefits of arts participation related to psychosocial wellbeing. In his media release, Garret makes the claim that “Arts education can also help address social exclusion” (2009, para. 5). While he does not specify which dimensions of social exclusion are likely to be addressed, given he had already mentioned the academic, cognitive, economic, creative, artistic and cultural benefits – and the fact that social inclusion/exclusion is unique in its focus on “the social” (see Chapter 2) – it seems reasonable to assume he is referring to the social and subjective elements, which in turn relate to the extrinsic psychosocial category of musical benefits. The Parliament of Victoria also used psychosocial-related extrinsic benefits to validate their calls to increase school-based musical participation. While they cited intrinsic benefits as their main argument, the most notable “additional benefits” cited in the Inquiry referred to this psychosocial category: “student engagement and wellbeing [and] personal and social development” (Parliament of Victoria, 2013, p. xvi). The ability for musical participation to increase social connectedness, social and emotional capabilities, social and emotional wellbeing, and self-esteem, are also mentioned as findings of the Inquiry. As evidenced in the above documents, current policy recommendations for schools to provide music and the Arts is based on presuppositions that students’ participation in these activities will lead to a transfer of benefits in all three categories. Another important element to note in this support is the type of musical participation which is both attributed to and recommended to achieve these intrinsic and extrinsic benefits. Models of musical participation advocated by policy. The models of musical participation credited to achieve the transfer of benefits articulated by policy can also be identified in the policy documents cited above. The most obvious indication of this is in the titles of these documents. Both the National Review of School Music Education (Australian Government, 2005) and the Inquiry into the extent, benefits and potential of music education in Victorian schools (Parliament of Victoria, 2013) articulate a single focus on music education activities. Similarly, the 33

Shape of the Australian Curriculum: The Arts (ACARA, 2011) and the Arts in Australia’s National School Curriculum (Garrett, 2009) both indicate a focus on the Arts as curriculum-based activities. The only document to stray slightly is the National Education and the Arts Statement (MCEETYA & CMC, 2007), which advocates for education and, through, and rich in, the arts. It also articulates the Arts and education systems as two separate but parallel entities, and suggests a wider definition of the relationship between the two through the use of the term “school-based arts participation”. However, it goes on to claim that “All children and young people should have a high quality arts education” (MCEETYA & CMC, 2007, p. 5), and recommends the need for curriculum development, thus also suggesting a strong emphasis on education and curriculumbased activities. It is also possible to take a more detailed look at the government supported model of music education in a curriculum setting, including recommendations for the specific activities and processes recommended for delivering the above benefits to students. Acting on recommendations from the National Education and the Arts Statement (MCEETYA & CMC, 2007), the ACARA’s Shape Paper describes in detail how curriculum-based activities are intended to achieve arts-related outcomes. It is also possible to refer to state-based curricula documents such as the Victoria Curriculum and Assessment Authority’s (2012) document, Progression points examples for AusVELS The Arts - Music Level 8–10, which provides a list of the standards and progression points related to the inclusion of music in the curricula in the sate of Victoria. Yet given the similarity in these documents, and the fact the Shape Paper represents recent thinking at the national level, this section will refer only to the Shape Paper. Achieving intrinsic benefits. In describing the musical activities and processes students will participate in and be exposed to, as part of curriculum-based music programs, the Shape Paper outlines a model of musical education which they claim will lead to the acquisition of intrinsic benefits. This includes practical activities such as performance, composition and arrangement, which aim to promote musical competency. As part of this model, the ACARA also suggest that students should engage with musical appreciation activities, which will require them to learn about, listen to, and respond critically to the music of different cultures and traditions. It is also claimed that engaging in such activities will 34

encourage students to use their imagination, communicate ideas through music to classmates, and respond to the musical expressions of others. These activities, and their intrinsic benefits, are located within the context of music education and curriculumbased music classes. Achieving extrinsic benefits related to cognitive and academic development. The Shape Paper also states that literacy will be supported through all arts activities by the opportunities they provide to: Make meaning, express thoughts, emotions and ideas, interact with others and participate in a range of communication activities, using a wide variety of printed, audio, visual and digital materials [and also] interpret, analyse, respond to and construct increasingly complex works. (ACARA, 2011, p. 23) In specific relation to music, it suggests literacy will be fostered through building skills in performance and composition, through which students learn notation, pitch, tempo, rhythm and sound dynamics. Similarly: Numeracy will occur naturally as students are asked to learn, design, make, create, evaluate and complete activities that require significant numeracy skills. Students will develop their ability to interpret and use symbols and graphic texts, diagrams, charts, tables, graphs and time, from score reading and composition in music. (ACARA, 2011, p. 23) Again here, we can see a strong focus on music instruction and the development of musical skill in a curriculum setting. Achieving extrinsic benefits related to psychosocial wellbeing. While the Shape Paper maps out activities and processes linked to the acquisition of intrinsic and extrinsic benefits related to academic and cognitive development relatively clearly, how activities support the acquisition of psychosocial outcomes is not clear, if mentioned at all. Ultimately, given recommendations are based around a model of music education, or curriculum-based activities undertaken in mainstream classrooms, there appears to be an assumption that this model will automatically produce a transfer of subjective wellbeing benefits. This same assumption is implicit in each of the key policy documents referenced in this chapter.

35

Evidence to Support Policy Claims for the Benefits of School Music The policy documentation explored above provides insight into the nature of current government support and recommendations for the inclusion of music in schools. To summarise, it would appear that such recommendations and support are based on two main sets of assumptions: that musical participation in mainstream schools will facilitate the acquisition of intrinsic benefits, extrinsic benefits related to cognitive and academic development, and extrinsic benefits related to psychosocial wellbeing; and that musical participation in the form of music education, or curriculum-based music classes, will achieve all of these benefits. To test the legitimacy of these assumptions, it is possible to compare them against evidence from the growing body of research literature that has investigated the relationship between musical participation in educational contexts and all three categories of benefits. Such evidence comes from research undertaken in a number of fields, including music therapy (Carr & Wigram, 2009), psychology (Gill & Rickard, 2011; Welch et al., 2011), and education (Kelstrom, 1998). Evidence for intrinsic benefits. There is significant evidence that shows school-based musical participation does lead to the intrinsic benefits stated in policy literature (Fowler, 2002; Ivey, 2002). This includes evidence for outcomes such as the development of self-expression, social values, life and job skills, and other factors considered important for students to become active citizens (Jones, 2010; Robinson, 2004). Benefits have also been reported which relate to the type of skills, learning and thinking considered important for today’s knowledge economy (Jones, 2005). These include creativity; fluency; originality; elaboration; resistance to closure; preparedness to take risks in learning; self-expression; expression of thoughts and ideas; imagination; cooperation; problem solving; curiosity; creative and flexible thinking (Burton, Horowitz, & Abeles, 1999); as well as aesthetic knowledge (McCarthy et al., 2004). Further, intrinsic outcomes are invariably linked to the models of music participation advocated in policy. For example, music education has been reported as the primary driver of intrinsic outcomes such as increased musical competence (Shields, 2001) and development (Welch et al., 2011). This is consistent with evidence that Arts education more generally increases motivation in, as well as enjoyment and understanding of the artistic experience (Haynes & Chalk, 2004). Music education has

36

also been reported to develop civic engagement (Doyle, 2014), intercultural understanding (Jones, 2010), and address issues of multiculturalism in schools (Joseph, 2011). More specifically, some claim that increased respect for diversity is facilitated through music education and the development of musical skills (Odena, 2007). Others have been more specific still, declaring direct instruction significantly increases musical, especially rhythmical, literacy and comprehension (Lowe & Belcher, 2012). While the intrinsic benefits have long been central to this area, authors note such benefits of have often been overlooked as justifications for music in schools (Rickard, Bambrick, et al., 2012; Stewart, 2007; Taylor, 2008). Nevertheless these studies support claims in current policy literature that musical participation will lead to a range of intrinsic benefits, and that music education is a suitable model of participation to facilitate this. Evidence for cognitive and academic benefits. The claim that music education can play an important role in facilitating cognitive and academic development has arguably become the strongest argument for supporting music in schools (Taylor, 2008). Some have gone as far to label this a “new paradigm shift” within music education, with such benefits now cited as the primary reason for teaching music (Vitale, 2011). While others have criticised (Stewart, 2007), or at least questioned (Rickard, Bambrick, et al., 2012) this “shift”, its popularity is understandable given the amount of research to support it, and the level at which research has been able to articulate how music can achieve these outcomes. Evidence in this area has shown that musical participation can lead to increased academic achievement (Vaughan et al., 2011) progress (Bryce, Mendelovits, Beavis, McQueen, & Adams, 2004) and competency (Burton et al., 1999). Such competency has been shown in general areas of numeracy (Catterall, Capleau, & Iwanga, 1999) and literacy (Moreno et al., 2009), and in more specific areas such as learning alternate languages (Kang & Williamson, 2014; Kennedy & Scott, 2005) and performance on standardized achievement tests (Babo, 2004). Research has also shown several benefits in areas related to cognition (Črnčec et al., 2006; Schellenberg, 2005), such as intelligence (Schellenberg, 2004), and memory (Ho, Cheung, & Chan, 2003; Roden et al., 2012). Evidence in this field is also relatively clear about the type of musical participation that leads to benefits. For example music education has been linked to 37

cognitive development in areas of language, motor control and visual spatial abilities (Biasutti & Concina, 2013). Music instruction in particular has been shown effective in achieving increased academic outcomes (Doyle, 2014), IQ (Schellenberg, 2004, 2011b), spatiotemporal reasoning skills (Črnčec et al., 2006) and performance (Rauscher & Zupan, 2000), as well as verbal memory (Roden et al., 2012). Some have gone further to explain how musical skills acquired through instruction, which are similar in nature to skills in other areas, result in a transfer of benefits (Haynes & Chalk, 2004). This involves the augmentation of language, literacy, mathematic, intelligence, and other areas of development (Hallam, 2010). Some go on to suggest cognitive and academic development occur through processes in which different parts of the brain relative to the development of these skills are stimulated by musical activities such as listening (Kang & Williamson, 2014; Schellenberg, 2005), instruction, or skill development (Hargreaves & Aksentijevic, 2011; Ho et al., 2003). For example, processing rhythmic or temporal beat patterns are said to be instrumental in developing mathematical skills (Catterall, 2009). While the link between music and cognition and said to be well established (Hargreaves & Aksentijevic, 2011), this remains an area of knowledge under development. Some authors stress the need to further investigate this area (Mehr, Schachner, Katz, & Spelke, 2013), including the directionality of causation, as well as the impact of factors such as personality (Corrigall, Schellenberg, & Misura, 2013) and executive function (Schellenberg, 2011a). Others also caution a categorical conclusion that school-based music can and will always achieve increased academic outcomes, suggesting more work is also needed in this area to fully understand this phenomenon (Gill & Rickard, 2012). That being said, as evidenced by the above literature, academic and cognitive benefits are not only well supported by research evidence, the processes in which these benefits are achieved are becoming increasingly understood. Overall, this suggests that claims regarding the academic and cognitive benefits of school music contained within policy documents are at least reasonable, and that the models of participation credited for their acquisition are in line with current evidence. Evidence for subjective wellbeing benefits. The subjective or psychosocial wellbeing benefits of school-based musical participation have received much less attention in research than the categories above (Gill & Rickard, 2012). Nevertheless, it is an area that has received increased attention, 38

with the amount of studies aimed at demonstrating a link between psychosocial factors and music in educational contexts growing in recent decades. Evidence from such studies, however, has shown varying levels of support for this link. Some studies have shown positive results; for example, music instruction has been credited for increased self-esteem (Costa-Giomi, 2004; Hietolahti-Ansten & Kalliopuska, 1990), coping, social networks, pro-social behaviour (McKegg et al., 2012), self-efficacy, school engagement (Rusinek, 2008), and a sense of happiness (Devroop, 2012). Other school-based music programs have also been linked to reduced psychological symptoms (Choi, 2010; Vaughan et al., 2011), as well as enhanced anger management (Currie, 2004; Currie & Startup, 2012; Jones, Baker, & Day, 2004), affect, social functioning (Cheong-Clinch, 2009), social behaviours (Kirschner & Tomasello, 2010), self-confidence, self-efficacy, self-acceptance, group and social bonding, and the ability to cope with loss (McFerran & Teggelove, 2011) and bereavement (McFerran & Crooke, In press). While promising, this evidence is somewhat mediated by a similarly significant body of literature which has reported inconsistent findings in this area. This includes a number of studies that found mixed results for the impact of school-based music on psychosocial constructs (Baker & Jones, 2006) such as self perception, coping (McFerran et al., 2010), social skills, social behaviours, social competence (Gooding, 2011), social support, social acceptance, self worth, (Shields, 2001), and several other psychosocial constructs (Kim et al., 2006). Also included in this body of literature are several studies that have reported an absence of benefits (Rickard, Bambrick, et al., 2012). Such studies question links between school music and psychosocial constructs such as: social skills; social (Schellenberg, 2004), problem (Rickard, Appelman, et al., 2012), and other behaviours (Baker & Jones, 2006; Michel & Farrell, 1973); self-esteem (Darrow, Novak, Swedberg, Horton, & Rice, 2009); anger expression (Knox-Anderson & Rickard, 2007); anxiety; social relationships; communication disturbance (Rickson & Watkins, 2003); and a range of other psychosocial factors (Rickard, Bambrick, et al., 2012). One study has even reported increased aggression and reduced motivation in a sample of students participating in school-based music activities when compared to a non-arts control groups (Rickard, Bambrick, et al., 2012). Another has reported a lack of student engagement through musical activities, and gone on to refute results regarding the role of musical participation in increasing self-esteem reported elsewhere (Grimmett, Rickard, Gill, & Murphy, 2010). One article, which reported no link 39

between music lessons and emotional intelligence, went as far as concluding the nonmusical benefits of music instruction are limited to intellectual abilities only (Schellenberg, 2011b). Overall, it is apparent that research in this area demonstrates a significant level of inconsistency. This is observable not only in the latter literature, which contradicts and challenges positive research results, but also within the results of many individual studies which were either inconclusive or contradictory in themselves. This makes any conclusive assertion regarding the potential for music to achieve psychosocial outcomes in schools inherently problematic, and directly challenges associated claims within the policy literature. Such claims are further problematised by Gill and Rickard’s (2012) suggestion that much of the positive evidence in this area has resulted from investigations of musical participation based outside classroom or school contexts. This latter observation is particularly pertinent given the apparent assumption in policy literature that musical participation in the form of in-school music education, or curriculum-based music classes, will achieve psychosocial benefits. Indeed, as Gill and Rickard suggest, the majority of studies that did report psychosocial benefits (noted above) investigated participation in alternate forms or settings. Those that did focus solely on music instruction or education, investigated these activities outside classrooms settings (CostaGiomi, 2004), or outside schools altogether (Hietolahti-Ansten & Kalliopuska, 1990). Other studies reporting benefits investigated music therapy (Cheong-Clinch, 2009; Choi, 2010; Jones et al., 2004; McFerran & Teggelove, 2011; McFerran & Crooke, In press) or similar therapy programs (Currie, 2004; Currie & Startup, 2012; Smith, 2004), as well as programs not limited to music education or instruction activities (Kirschner & Tomasello, 2010; McKegg et al., 2012; Rusinek, 2008; Vaughan et al., 2011). The fact that positive evidence for psychosocial outcomes in this area is largely attributed to forms of musical participation that differ from those which policy literature advocates, further questions the legitimacy of such advocacy. This inconsistency supports Stevens and Stefanakis’ claim that advocacy in this area can be seen as “representing unsubstantiated assertion[s]” (2014, p. 12). While Stevens and Stefanakis make this assertion in regards to all policy-based advocacy, the research literature presented here shows this is particularly the case for claims related to the psychosocial benefits of school-based musical participation.

40

The Gap and its Perilous Potential As evidenced above, there is clear gap between claims in the policy literature for the psychosocial benefits of musical participation, and existing evidence in this area. Such a gap is of considerable concern both because these benefits have become increasingly central in justifications for the inclusion of music in schools offered by government (Parliament of Victoria, 2013), and because of the level of conviction with which they are touted in existing policy documents. At this level of conviction, any school or policymaker, who acting on these recommendations, takes steps to introduce class-based music education programs would understandably expect such benefits for students. That research evidence seriously questions the likelihood of this occurring has disastrous potential for the ongoing support of music in schools. As mentioned in the first chapter of this literature review, this support in its current form is delicate at best. If policymakers do implement legislation which allocates resources to the mandatory provision of music programs in all schools – including teacher training, provision of new facilities, and a mandated place in a curriculum (Parliament of Victoria, 2013) that is already overcrowded (Lowe & Belcher, 2012) – and these programs do not deliver the promised benefits, such legislation and resourcing will surely be retracted. This is of particular concern given the climate of budgetary cuts by the current government (Kirby, 2014), as such a sizable investment would likely be subjected to heavy evaluation, with any initiatives not producing the expected return swiftly cut. This will undoubtedly result in the next turn of Gattenhof’s (2009) 10-year cycle, in which government support for the Arts in schools will be set back for a decade or more. Further, even if legislation is not introduced, and schools decide to divert already scarce resourcing (Stevens & Stefanakis, 2014; West, 2012) toward music programs which fail to achieve the advertised benefits, support at the school leadership or administration level will likely falter. Given evidence that support from school leadership is one of the strongest enablers of school-based musical participation in the current school climate(Crooke & McFerran, 2015; McFerran & Crooke, 2014), this would have serious implications for the future of music in Australian schools. There appear two solutions to this dilemma. The easiest is to place less emphasis on claims that musical participation will result in psychosocial benefits for students, or remove them from policy literature altogether. Yet, arguably, this would deny students of one of the most important and unique benefits offered by this participation (Blandford & Duarte, 2004; Gill & Rickard, 2012). Further, as discussed in the previous 41

chapter, it was the re-recognition for the importance of the social aspects of human experience, in both social policy (Crisp, 2010; Morrison, 2010) and education (Wyn, 2009b), which can be seen as underpinning contemporary shifts towards advocacy for the Arts in education (Karkou & Glasman, 2004). To ignore “the social” in resulting policy initiatives could therefore be seen to disregard the policy shifts that enabled them. Filling the Gap Following recommendations from several authors, it seems critical more research be undertaken to fill the current gap in knowledge in this area, and inform decisions about the implementation of music in schools made by both policymakers and schools themselves (Gill & Rickard, 2012; Stevens & Stefanakis, 2014). However, while others claim the answer to filling this gap is more rigorous (Gill & Rickard, 2012; Knox-Anderson & Rickard, 2007), outcomes-based research (Kim et al., 2006; Stevens & Stefanakis, 2014), the staggering level of inconsistency in current evidence suggests otherwise. Indeed, given claims that policymakers are more receptive to outcomesbased evidence (Saunders, 2011; Torrance, 2011), such research is likely necessary. Yet, owing to the conflicting results from studies so far taking this approach, it would appear that before more of same research is undertaken, we must first aim to understand why previous studies have been so problematic – particularly since health and science research funding has been heavily reduced in the national budget (Kirby, 2014). Researchers in this area note several limitations to which they have attributed their lack of conclusive findings. While mentioned only briefly in limitation sections, a review of these factors across multiple studies can be seen to point the way to how and where we may identify limitations and challenges, not in current findings, but current approaches to investigating this area. This review is undertaken in the next chapter and aims to identify how and where to focus subsequent analyses of current research approaches, which in turn hope to identify research challenges, provide recommendations for future research, and inform policymakers and schools in this area.

42

Chapter 4: Accounting for the Inconsistency in Current Evidence This chapter undertakes a review of current research literature which has investigated the psychosocial benefits of musical participation in educational contexts. Specifically, it aims to identify how researchers who have published findings account for the inconclusive nature of evidence in this area. It does so by undertaking a synthesis of the factors that authors have reported as potential explanations for the inconsistency or absence of positive findings in their studies. It also compares these factors against research which has found positive findings, and literature from adjacent fields, were appropriate or possible. The main aim of this exercise is to identify broad areas in which this PhD project may undertake further analysis – where such analysis aims to identify the challenges of investigating the psychosocial wellbeing benefits of musical participation in mainstream schools, and thus provide recommendations for the design of further research and policy development. First however, this chapter presents a brief overview of the different conceptualisations and definitions of the term “psychosocial wellbeing”. This is done to locate the use of this term throughout the remainder of the dissertation within the fields of music research, education, and policy. Conceptualising Psychosocial Wellbeing Psychosocial wellbeing as a term and concept is complex and multidimensional (AIHW, 2012), and has been noted as a construct particularly problematic to both define and measure (AIHW, 2010; Hamilton & Redmond, 2010). Indeed, attempts to provide indicators by which to measure it have been ongoing (White, 2008), and draw upon several different definitions and markers (AIHW, 2012; Saunders, Naidoo, & Griffiths, 2008). The difficulty in theorising this concept is due in part to the World Health Organization’s (WHO) use of the term “wellbeing”, which is widely considered to be the origin of the term in contemporary discourse (Breslow, 1972; Kaplan & Anderson, 1988; Keyes, 1998). The WHO’s original 1948 definition of health as “a state of complete physical, mental, and social well-being [sic]” (cited by Bourne, 2010, pp. 1516) separated social and mental (or psychological) dimensions of wellbeing, and led to thinking about and measuring these as distinct constructs (Larson, 1993).

43

Objective versus subjective social wellbeing. Those who focused on the social dimension of wellbeing have been further divided on the question of whether it should be conceptualised objectively or subjectively (Gaspart, 1998). This debate, most notable in policy literature, saw the manifestation of two approaches (Bourne, 2010). The “Welfarists” (Keyes, 1998), took an economic welfare stance (Sen, 1980, 1988, 1998), and have used objective, externalised indicators, most notably gross national product and welfare mechanisms, to measure or predict social wellbeing at a national level. This model is popular among health economists and proponents of economic redistribution the world over (Smyth, 2010), and continues to influence policy in this country (ASIB, 2009, 2012; Kostenko et al., 2010; Saunders, 2011; Saunders et al., 2008; Scutella & Wilkins, 2010). Others have challenged a purely objective operationalisation of social wellbeing (Bourne, 2010), arguing the concept should accommodate subjective elements of human experience (Ringen, 1995; Veenhoven, 2002), and be measured using self-report methods (Tait, 2004). Those who have proposed “what” should be measured in this approach have focused on the internal responses to, or subjective evaluation of, ones’ social world (Keyes, 1998; Larson, 1996). This includes one’s satisfaction with their work (or school), family, and social networks (Hooghe & Vanhoutte, 2011); belief about the inherent “goodness” (Nussbaum, 2001) and potential of their society; and their perception of their place, and ability to participate, in society more generally (for further discussion, see Keyes, 1998). This model has also become increasingly influential in the Australian policy context, especially in relation to the wellbeing of young people (AIHW, 2010, 2011), and is arguably more consistent with the social inclusion policy paradigm described in Chapter 2 (Crisp, 2010; Morrison, 2010; Taket et al., 2009). Social and psychological wellbeing: distinct or linked? While some proponents of subjective social wellbeing originally aimed to measure it separately from its psychological (or mental) counterpart (Keyes, 1998), theorists have long argued for their integration (Kaplan & Anderson, 1988). Larson (1996) and Keyes (in his later work - 2002, 2005, 2006, 2007) have suggested these dimensions be combined under the heading “psychosocial” wellbeing, given that both can be seen as internalised constructs. Rather than diminishing the uniqueness of either, they argue the conceptualisation, measurement, and analysis of both dimensions can be

44

complementary. This is evident in tools used in Keyes’ later work, which, while still retaining social wellbeing subscales, combines these with psychological measures (Keyes, 2006, 2007; Shapiro & Keyes, 2008). This combined approach to subjective wellbeing has gained traction in many fields over recent decades, including policy (AIHW, 2012), psychology (Dobbs, Doctoroff, Fisher, & Arnold, 2006), education (Barclay & Doll, 2001), music education (Devroop, 2012), music therapy (Carr & Wigram, 2009), and music psychology (Boer & Abubakar, 2014). It has also been described using different terminology, including “psychosocial functioning” (Rickard, Appelman, et al., 2012), “social and emotional wellbeing” (ARACY & AIHW, 2012), “socio-emotional functioning” (McLoyd, Jayaratne, Ceballo, & Borquez, 1994), “socioemotional adjustment” (Demo & Acock, 1996), or simply “subjective wellbeing” (Tomyn, 2013). Despite their alternate phrasings, these different permutations all relate to the complementary operationalisation of social and psychological wellbeing, and are often used interchangeably (Coplan, Arbeau, & Armer, 2008; Turner & Roszell, 1994). For the remainder of this dissertation, the term “psychosocial wellbeing” is used to refer to this complementary conceptualisation of subjective wellbeing, and the different factors included under this umbrella term. This choice is both influenced by the work of Keyes (2002, 2007) and Larson (1996), and the fact much of the recent research investigating the subjective wellbeing of music in schools has used this term (Gill & Rickard, 2012; Knox-Anderson & Rickard, 2007; Rickard, Bambrick, et al., 2012). Conceptualising young people’s psychosocial wellbeing. To give a broad overview of factors considered important for the psychosocial wellbeing of young people in Australia, it is possible to draw on several government documents that have provided indicators for conceptualising and measuring the construct in this population – done here to maintain a link to policy literature. Such documents, predominantly from the Australian Institute of Health and Welfare (AIHW), stress the necessity of taking a holistic (AIHW, 2012), or “whole child” (Hamilton & Redmond, 2010) approach, which emphasises the multidimensional nature of a young person’s welfare. They go on to state that psychosocial wellbeing: Refers to the way a person thinks and feels about themselves and others. It includes being able to adapt and deal with daily challenges (resilience and 45

coping skills) while leading a fulfilling life. Hence, there is an emphasis on the behavioural and emotional strengths of children, as well as how they respond to adversity. (AIHW, 2012, p. 8) This is further divided into two interdependent dimensions: the “individual dimension – consisting of internal (intrapersonal) and relational (social/interpersonal) characteristics; and the environmental dimension – consisting of influences from three spheres: family/home, early education settings/school, and community” (AIHW, 2012, p. 9). The individual dimension is said to include “feelings, behaviour, relationships, goals and personal strengths” (Hamilton & Redmond, 2010, p. 5), and: Encompasses a number of skills that children need to develop in order to succeed at school, and in life in general. These include the ability to identify and understand one’s feelings, accurately read and comprehend emotional states in others, manage strong emotions and their expression, regulate one’s behaviour, develop empathy for others, and establish and sustain relationships. (AIHW, 2009, p. 60) Also, while there is a push towards conceptualising this area through the lens of positive psychology (i.e. factors that prevent, rather than those that are symptoms of, negative wellbeing; AIHW, 2012), authors recognise this individual dimension may include “individual behaviours that are seen as socially problematic, such as disruptive behaviour at school, or drug use or other risky behaviour, [or other] problems, for example hyperactivity, low self-esteem, anxiety or depression” (Hamilton & Redmond, 2010, p. 6). The environmental dimension is said to relate to the perceptions a young person has about their place in community, and the support they receive in their daily life (AIHW, 2012; Hamilton & Redmond, 2010). This includes support they receive from their peers, families, school community (including teachers and classmates), and broader community or neighbourhood (Hamilton & Redmond, 2010). This may involve a young person’s feeling that they: belong to their community or school (AIHW, 2011); have a positive role model, have someone to go to when they have a problem (AIHW, 2010); have something to contribute to their community; or that their views and opinions are acknowledged, recognised or respected by those around them (Hamilton & Redmond, 2010). It also relates to school (AIHW, 2012) and community (AIHW, 2010) engagement and participation. Perceived and actual support or engagement in these 46

areas is said to augment the coping and resilience of a young person, their educational outcomes, and their quality of life more generally (AIHW, 2012). The multidimensional and subjective nature of psychosocial wellbeing articulated in these documents is consistent with the conceptualisation offered by theorists such as Keyes (2002, 2007) and Larson (1996) mentioned earlier in this chapter. It also aligns with the extrinsic benefits of a student’s musical participation described in the literature presented in Chapter 3. Additionally, it captures the subjective (Crisp, 2010; Morrison, 2010), social (Taket et al., 2009), and multidimensional (Levitas et al., 2007; Wilson, 2006) nature of social inclusion– said to be the defining characteristics of this paradigm (Saunders, 2011) – and the recent push for a social and holistic approach to education observable in both policy (MCEETYA, 2008) and academic spheres (Bond et al., 2007; Wyn, 2009a). Reported Limitations of Investigating the Psychosocial Benefits of School Music Studies investigating the psychosocial wellbeing outcomes of musical participation in mainstream schools report several factors to explain the inconsistency of findings that characterise this area of study (as outlined in Chapter 3). These factors can be grouped into two broad categories: study design and methodology, and the forms of musical participation under investigation. Importantly, literature reviewed here is limited to investigations of the psychosocial wellbeing benefits of music programs delivered in mainstream educational contexts. This has excluded any study which has not assessed at least one or more construct of psychosocial wellbeing, and the many studies that have investigated the benefits of musical participation in special education contexts – defined as education institutions catering to students with moderate to sever disabilities (Adamek & Darrow, 2010; McFerran & Stephenson, 2010; Nordoff & Robbins, 1971). This is based on the argument made in Chapter 3 regarding the need to further investigate the psychosocial benefits of musical participation. This choice is also based on the fact that policy in this area has mainly focused on mainstream contexts (ACARA, 2011; Australia Office for the Arts, 2011; MCEETYA & CMC, 2007), and that this context is underrepresented in the research literature regarding the benefits of music in schools (Carr & Wigram, 2009).

47

Study design and methodology. Reported limitations related to study design and methodology involve the overall methodological approach and study designs used. They also include methods of sampling, data collection, and choice of reporter. Quantitative versus qualitative methods. To begin at the broadest level, Knox-Anderson and Rickard (2007) intimate the inconclusive results of their study were potentially due to the shortcomings of quantitative methodologies. Investigating the impact of active classroom-based music programs on six-year-olds, they found no statistically significant improvements for selfesteem or anger expression. In their discussion, they suggest this non-significance may be due, in part, to the fact that the benefits of some music programs are too subtle to detect using standardised psychometric measures. They go on to recommend the use of qualitative methods when evaluating the benefits of school-based musical participation. While the only authors in this specific field of research to explicitly propose the shortcomings of a quantitative methodology (as an entire approach), Knox-Anderson and Rickard’s suggestion is consistent with claims from the field of arts education. Authors in this field confer that, given the subtlety, complexity and ambiguity of artistic processes (Ewing, 2010), standard quantitative methods are unsuited to capturing the benefits of arts-based activities (Eisner, 1999; O'Toole, 2010) for students (Hunter, 2005), especially in classrooms environments (Winner & Hetland, 2000). Supporting these arguments is the fact qualitative studies in this area have more regularly returned consistent results than their quantitative counterparts. Those using qualitative methodologies have reported school-based music therapy programs were effective in: improving self-esteem, peer relationships, confidence, social interaction, motivation, enthusiasm, communication, affect, and social functioning (Cheong-Clinch, 2009); facilitating the development of coping strategies, social and peer support systems, increased emotional expression, and a sense of belonging; as well as reducing psychological symptoms (Choi, 2010) and aggressive behaviours (Jones et al., 2004) in refugee students. Others report similar outcomes for mainstream students, including reduced anger, depression and misbehaviour (Currie, 2004), increased school engagement, self confidence, self-efficacy (McFerran & Teggelove, 2011; Rusinek, 2008), connectedness, community and cultural awareness, and positive learnt behaviours (McKegg et al., 2012).

48

Meanwhile, like Knox-Anderson and Rickard (2007), several studies using quantitative methodologies have struggled to provide consistent evidence in this area. Using standardised psychometric scales, Schellenberg (2004) reported music instruction had no positive impact on adaptive or maladaptive behaviours, while Rickard, Appelman, et al. (2012) found classroom-based music education had no significant effect on social skills. Others have reported mixed findings, where positive benefits are reported for some quantitative measures, or form or data collection, but not others (Baker & Jones, 2006; Michel & Farrell, 1973; Rickson & Watkins, 2003). Further lending to problematic nature of quantitative methods in this area is the fact several investigations reported non-significant (Darrow et al., 2009; McFerran et al., 2010; Michel & Farrell, 1973; Rickard, Bambrick, et al., 2012) or inconsistent (Montello & Coons, 1998; Shields, 2001) results using quantitative methods, yet found positive results using qualitative measures in the same study. This discrepancy between methods is also described in Grimmet, Rickard and Gill’s (2010) retrospective analysis of a multi-school music education program. While another study of the program had, via quantitative measures, reported self-esteem benefits (see study by Rickard, Appelman, et al., 2012, cited above), Grimmet, Rickard and Gill’s re-evaluation of the program, using qualitative methods, challenged the validity of these previous findings. 5 They countered that the music program was sometimes counterproductive for both the students and school communities involved. This literature largely supports suggestions from Knox-Anderson and Rickard (2007), and authors from the wider arts education field (Eisner, 1999; Ewing, 2010; Hunter, 2005; O'Toole, 2010; Winner & Hetland, 2000), that quantitative approaches are less suitable for capturing the psychosocial benefits of musical participation. However, concluding the quantitative/qualitative divide alone can explain the inconsistency of results in this area becomes problematic when considering the amount of studies that have reported positive benefits using quantitative methods. These include Costa-Giomi (2004), who using standardised self-esteem scales, reported significantly improved self-esteem for students participating in three years of piano lessons when compared to a non-musical control group. Devroop (2012) reported similarly significant

5

While Rickard, Appelman, et al.’s, (2012) study was published after Grimmet,

Rickard, and Gill’s (2010) paper, their research was completed first. 49

results on measures of self-esteem, optimism, happiness and perseverance for young people participating in a two-year school-based music program, also using a control. Hietolahti-Ansten and Kalliopuska (1990) also reported students with an active interest in music, defined by undertaking private instruction in violin and piano, had significantly higher self-esteem and empathy scores than a similar control group. Comparably positive results were also reported on quantitative measures of pro-social behaviour by Kirschner and Tomasello (2010), and aggressive behaviour, anger, depression, and self-esteem by Currie and Startup (2012). What’s more, Vaughan, Harris, and Caldwell (2011) suggest their findings that a specialised school music program increased the school engagement and social wellbeing for a number of students was supported equally by both qualitative and quantitative methods. Overall, this literature suggests that, while it may play a role, the suitability of either quantitative or qualitative methodologies cannot, in isolation, explain the inconsistency of findings in this area. Indeed, studies reporting limitations in this area have been much more focused in accounting for their inconclusive or varying results. Quantitative design and sampling. Given the majority of studies publishing inconclusive results have used quantitative methodology, many reported limitations have focused on the design and sampling methods used in these approaches. For example, Kim et al. (2006), who investigated the psychosocial benefits of a music mentor program, attribute their inconclusive results for participant-report measures to a lack of a pre-post design. While several studies have reported benefits using only correlation designs (Hietolahti-Ansten & Kalliopuska, 1990; Vaughan et al., 2011), Kim et al.’s conclusion aligns with claims from several authors that research in this area lacks rigorous (Gill & Rickard, 2012; Stevens & Stefanakis, 2014) or systematic designs (Ewing, 2010; McCarthy et al., 2004), and thus question the validity of subsequent findings. Another reported limitation concerning design rigour is sampling. Kim et al. (2006) also attributed their insignificant results to the absence of a control group, as did other studies that (unlike Kim et al.) did use pre-post designs. For example, while both McFerran et al. (2010) and Shields (2001) reported some positive results, they suggest the potential to attribute positive results to the musical element of their interventions was limited due to the lack of control groups. Other studies that did include control groups reported the fact that participants were not allocated to groups randomly led to 50

non-comparability between groups at baseline, and limited their ability to reach statistically significant results (Knox-Anderson & Rickard, 2007; Michel & Farrell, 1973). The importance of using effectively randomised control groups was also noted in studies that did find significant results using this approach. This includes Kirschner and Tomasello’s (2010) study that found significant evidence for links between musical participation and prosocial behaviours. Also, Costa-Giomi (2004) concluded their significant findings that musical instruction improves self-esteem, were “most likely” due to the fact randomisation enabled similar baseline scores between groups. Others reported that achieving randomisation in schools can in itself be a challenge. Rickard, Bambrick, et al. (2012) reported (for the first of two studies) school staff involvement in allocation procedures resulted in a “pseudo-random” design, in which no psychosocial improvements were reported. Citing similar complications, Currie and Startup (2012) reported an imperfect randomisation procedure limited their ability to generalise findings for improvement on measures of depression, self-esteem and anger. These reports suggest facilitating ideal research protocols when working within school is its own limitation. Authors also reported that full randomisation does not guarantee significant results. Rickson and Watkins (2003) and Rickard, Appelman et al. (2012) both suggested, despite effective randomisation, their non-significant results may be explained by disparate baseline scores between groups. Further, Rickard, Bambrick, et al.’s (2012) second study, which employed both full randomisation and statistical adjustments to ensure groups were consistent at baseline, also found no significant psychosocial benefits. Other studies using effective randomisation have also produced mixed results (Baker & Jones, 2006; Gooding, 2011), suggesting randomisation alone is not sufficient in elucidating consistent outcomes. These studies suggest that, while the absence of a rigorous design and sampling techniques are limitations in this area, even when present they do not guarantee that psychosocial benefits will be captured via quantitative methods. This indicates inconsistency in quantitative findings may be due to more nuanced factors. Reporters. That the inconsistency of research results in this area are due to factors that extend past the more overt elements of study design is supported by several authors. 51

This includes issues related to choice of different reporting methods, or the reporters themselves. The limitations of student self-report methods receive notable attention in the literature. This often relates to the ability for young people to respond to or interpret concepts associated with psychosocial wellbeing. Michel and Farrell (1973) explain the non-significance of their results partly with the observation that students struggled to comprehend and respond to items in the self-esteem scale used in their study. Kim et al. (2006) also reported the Year 7 students in their study had trouble conceptualising, articulating and reporting the psychosocial benefits of musical participation through self-nomination techniques. Knox-Anderson and Rickard (2007) also credit the notable amount of missing data in their study to the use of self-report surveys, particularly where students had low academic scores. These observations confer with literature which questions the validity of using self-report tools with young people – especially when investigating psychosocial wellbeing – due to varying developmental stages of comprehension and understanding (for discussion, see AIHW, 2012). Another stated limitation of self-report methods is the format of survey delivery. Knox-Anderson and Rickard (2007) suggest the use of “paper-pencil” surveys may have disengaged participants with lower academic abilities, thus accounting for the drop-out rates experienced in their study. McFerran et al. (2010) also cite this format disengaged students, likewise linking it to their non-significant scale results. While several studies used third-party-reporters – such as parents (Schellenberg, 2004), teachers (Rickard, Appelman, et al., 2012), caregivers (Rickard, Bambrick, et al., 2012), and case managers (Gooding, 2011) – without notable issues, this approach was reported a limitation by some. Rickson and Watkins (2003) noted the use of both caregivers and teachers to score a behaviour scale led to conflicting results, and limited the consistency of study outcomes. They suggest this reinforces Rapoport and Ismond’s (1996) ideas that consistency between scorers is rare, and that caregivers are better suited to rating behavioural measures than teachers. Michel and Farrell (1973) further state that teacher-report methods used to measure behaviour in their study were problematic. They conclude that either teacher reports are unreliable, or that several teachers need to undertake scoring of behavioural measures. Inconsistencies were also noted when using third-party reports to complement student self-report data. Kim et al. (2006) note therapist observations collected in their study contrasted with student responses, and subsequently suggest participants and 52

therapists may perceive subjective phenomena differently. Perceptual differences between third-party observers and participants when using scales was also noted by Baker and Jones (2006). They reported language and cultural barriers between refugee students and staff meant teachers had difficulty identifying culturally-specific internalised student behaviours, citing this as a possible explanation for non-significant results observed on teacher-report measures. This confers with Michel and Farrell’s (1973) suggestion that internalised constructs, such as self-esteem, are more appropriately captured by self-report than third-parties. Scales. Reported limitations also relate to the nature of scales used to collect data. However, limitations reported in this area sometimes contradict each other. For example, Kim et al. (2006) and McFerran et al. (2010) partly attribute the inconsistency of their results to the fact they did not use appropriate standardised measures. Yet, KnoxAnderson and Rickard (2007), who did use standardised tests, propose psychosocial benefits are too subtle to pick up via standardised psychometric measures; arguing instead “Ad-hoc measures specific to each classroom may therefore be more sensitive to the types of changes occurring following implementation of music programs as opposed to theory-driven, pre-determined measurements” (2007, pp. 13-14). Kim et al. (2006), McFerran et al. (2010), and Darrow et al. (2009) also suggested the shortcomings of their scales owed to the fact they did not accurately address constructs targeted by their respective music interventions. This in part concurs with Knox-Anderson and Rickard (2007) proposed need for ad-hoc measures relevant for individual studies. Further, Devroop (2012), who’s survey instrument was designed specifically for the constructs targeted in their study, returned positive results. While this ad-hoc approach arguably limits the reliability of results (i.e. Devroop’s scales were not validated), the value of targeting scales is reinforced by other studies which, using established scales relevant to their particular studies, also reported positive results (Costa-Giomi, 2004). Other reported limitations relating to scale suitability include Darrow et al.’s (2009) claim their insignificant results occurred not only because measures insufficiently targeted constructs, but also because these constructs were insufficiently addressed by the music intervention. Also, Rickard, Bambrick, et al. (2012) suggested that, while their study found no positive results for the psychosocial constructs 53

measured, it is possible psychosocial benefits did occur but they were not detected by their measures. This is supported by several studies which did not find benefits related to constructs measured by scales used, yet did find other benefits were reported via qualitative methods (Darrow et al., 2009; McFerran, 2010b; Shields, 2001). Nature of musical participation. Methodological elements are not the only extenuating factors cited as potential explanations for the significant variation in outcomes reported in this area. Several authors note the nature of the music activities themselves may have significantly impacted the reporting of outcomes; including how, where and for who an activity or program was facilitated, who facilitated it, and the length of students’ participation. Duration of participation. Offering yet another explanation for their inconclusive results, Knox-Anderson and Rickard (2007) propose the four-week music programs in their study were insufficient in length to achieve positive outcomes. Subsequently, they suggest programs need to last for 10 weeks or more to achieve psychosocial outcomes, especially ones related to anger and self-esteem. This is consistent with Michel and Farrell (1973), who found their seven-week music instruction program was also too short to effect significant changes in the self-esteem, and also go on to cite music program length a potential limitation for research in this area. These claims are somewhat supported by the outcomes of Rickson and Watkins’ (2003) study, which investigated the outcomes of an eight-week program, and found no significant change on a range of developmental behaviours. Duration of musical participation is also cited a limitation by several other authors who propose even 10-week programs are insufficient. Montello and Coons (1998), who’s 12-week program investigated (among other things) group cohesion in two sample groups, claim one sample only reached group cohesion in the last week of the program, suggesting the 12-week length was critical for this group. Other authors also explain their ambiguous or insignificant results with the claim that music programs lasting 16 (Darrow et al., 2009) or 20 (Baker & Jones, 2006) weeks are too short to achieve statistically significant results, thus arguing for longer programs. Other authors have taken more extreme stances. Rickard, Bambrick, and Gill (2012), who investigated a 6-month music program, suggest programs need to be even

54

longer to effect consistent and statistically significant changes. Such a position is supported by studies which have reported positive results when investigating programs of one (McKegg et al., 2012) or more years (Rusinek, 2008), including some lasting up to three (Costa-Giomi, 2004; Devroop, 2012) and even six years (Hietolahti-Ansten & Kalliopuska, 1990). Support for program length also comes from Vaughan et al.’s (2011) study, which reported that students who participated for periods between 12 to 18 months had better outcomes than those who participated for only six months. This backs the idea that, when it comes to achieving psychosocial benefits through musical interventions, longer periods of participation are more effective. However, again, this conclusion becomes problematic when considering benefits observed in longer studies have been reported to decline in the second (Rickard, Appelman, et al., 2012) and third (Costa-Giomi, 2004) years of a program. Also challenging this is the fact that studies have reported positive benefits from programs as short as eight (Kim et al., 2006) and five weeks (Gooding, 2011). Others have reported results from even shorter periods of participation, including one study that lasted only 20 minutes (Kirschner & Tomasello, 2010), and another that reported observable benefits immediately after the music program commenced (Jones et al., 2004). While there is evidence to support suggestions that longer programs increase the likelihood of achieving outcomes, the fact some programs showed a decline in benefits after a year, and some reported outcomes in 20 minutes or less, question this as an accurate conclusion. Again this suggests that while program length may be a limitation in research in this area, it alone cannot account for the level of inconsistency in outcomes reported in this area. Age of Participants. Another limitation related to the nature of a music program is the age of participants. Most notably, this research caveat is reported by Rickard, Bambrick, et al. (2012), who suggest the absence of outcomes reported in their study may be explained by the fact non-musical benefits of school-based musical participation “are more easily generated in younger participants” (2012, p. 12). In support of this idea they cite research that has shown the effect of music training on synaptic plasticity is age dependent (Pantev et al., 1998; Schlaug, 2001, as cited in Rickard, Bambrick, et al., 2012). They also cite studies that show musical participation has had positive effects on 55

academic skills in early primary school and kindergarten-aged children (Anvari et al., 2002; Costa-Giomi, 2004; Rauscher & Zupan, 2000; Schellenberg, 2004, as cited in Rickard, Appelman, et al., 2012). Closer inspection of these studies, however, makes extending this argument to the acquisition of psychosocial outcomes problematic. While Costa-Giomi (2004) did report self-esteem benefits for Year 4 students, positive effects for Schellenberg’s (2004) six-year-old participants were restricted to IQ outcomes, with no significant outcomes found for psychosocial outcomes. The only other studies to identify age as affecting the reporting of psychosocial outcomes are Rickson and Watkins (2003) – who briefly mention that the age difference between participants may have affected results – and those that stated age a factor when collecting data via student self-report measures (Kim et al., 2006; Knox-Anderson & Rickard, 2007; Michel & Farrell, 1973). Apart from these studies, this limitation gained little further attention in the literature. Studies in this area have investigated students from various age brackets, with no clear pattern observable for age in affecting reported benefits. Status of sample population. Another potential limitation identified in this area is whether participants are classified as “mainstream” or “at-risk”. The definition of “at-risk” used here borrows from Shields’ description as: “those enmeshed in debilitating social, emotional, physical, academic, and economic difficulties, whose individual configurations of assets and deficits may have diminished their likelihood of success in school and society” (2001, p. 275). Again, this excludes significantly disabled populations. “Mainstream” is used here to describe participants who have not specifically been identified as being at-risk, and who attend schools not classified as special education institutions, or participate in programs not targeted to moderately or severely disabled populations. The main argument related to at-risk populations is articulated by Rickard, Bambrick, et al. (2012), who argue a possible caveat for their insignificant results is the fact that children were not at-risk. Rickard, Appelman, et al. (2012) confer, suggesting their study’s insignificant findings are due largely to high baseline scores on psychosocial constructs, which they attribute to the fact participants were not at-risk. They go on to argue musical interventions only affect students with low baseline scores for the psychosocial construct assessed. As noted by these authors, this is consistent with previous claims from those such as Teachout (2005) and Shields (2001), both of 56

who suggest at-risk populations are more likely to experience psychosocial benefits from musical participation. This claim is also supported by the fact several studies which have targeted atrisk participant groups have reported consistent findings using both quantitative (Currie & Startup, 2012) and qualitative (Cheong-Clinch, 2009; Choi, 2010; Jones et al., 2004; McFerran & Teggelove, 2011) methods. This includes Rusinek’s (2008) study, which goes on to advocate the importance of music programs in addressing students at-risk of social exclusion and life failures that result from disengagement in academic, school, cultural or other spheres. Again, however, there is evidence that challenges the need to include at-risk students in programs to observe positive results. Montello and Coons (1998) reported one of their intervention groups exhibited more pronounced behavioural problems than the other. Not only was this reported to affect scale results for this sample, qualitative observations suggested it prevented group cohesion, and actually made some students feel unsafe in sessions. Authors go on to suggest musical participation may actually exacerbate disruptive behaviours for some at-risk students. Rickson and Watkins’ (2003) findings that students with Attention Deficit Hyperactivity Disorder can become over stimulated and disruptive after active music therapy sessions support this. The idea that programs with at-risk students are more likely to achieve consistent outcomes is also questioned when looking at the outcomes of studies that investigated programs with mainstream students not considered at-risk. While some of these studies found no (Grimmett et al., 2010; Rickard, Bambrick, et al., 2012; Schellenberg, 2004) or mixed results (Kim et al., 2006; Knox-Anderson & Rickard, 2007; Rickard, Appelman, et al., 2012), there are several that did find positive outcomes in mainstream samples (Devroop, 2012; Hietolahti-Ansten & Kalliopuska, 1990; Kirschner & Tomasello, 2010; McKegg et al., 2012; Vaughan et al., 2011). The idea that investigating participants not at-risk impedes positive or consistent results is further challenged by Costa-Giomi (2004). They suggest the effect of their intervention, which was originally targeted towards students considered at-risk because of low socioeconomic status, was found regardless of this status. Student engagement and musical activities Authors also note that students’ level of engagement in music programs has the potential to affect reported benefits. Rickard, Bambrick et al. (2012), found no positive 57

results for the impact of music education and training on psychosocial wellbeing. They partly explain their findings with the claim that the program was insufficiently engaging, as student motivation and engagement was reported to decline throughout the study. These authors suggest the cause of this declined engagement was the program’s focus on music theory, and the fact some students reported a preference for alternate activities. As they note, this supports Hallam’s (2010) claim that non-musical benefits of classbased music instruction are only achieved if sessions are enjoyable. Rickard, Appelman, et al.’s (2012) study also investigated music education, which again focusing on music instruction and theory. They reported modest benefits for self-esteem, which dissipated in the second year, and no other psychosocial outcomes. These results are again partly explained with the idea that the music program was not sufficiently engaging. Grimmet et al.’s (2010) retrospective analysis of that same program strengthens this notion. They refuted the presence of self-esteem benefits outright, linking this to the observation that the structured nature of music instruction programs was unsuitable for younger students, and the instruments in which instruction was offered were described as boring for students. Subsequently, authors reported the program was disengaging, and potentially counterproductive for psychosocial wellbeing, and the musical culture of the school communities more generally. The link between musical activities, engagement, and psychosocial wellbeing is also supported by a number of studies which investigated programs not focused on music instruction or theory. A music therapy program investigated by Jones, Baker, and Day (2004) was reported as beneficial for several psychosocial outcomes. Importantly, authors noted that choosing music activities that were appropriate for these students was a vital step in engaging them and achieving outcomes. This included activities such as collaborative rhythm percussion, which reportedly promoted musical expression, relaxation, communication, and built rapport and connection between group members. The choice of hip-hop music was also noted important, and said to engage participants. Other music therapy studies reported similar observations. Cheong-Clinch (2009) stated that the use of song sharing, lyrical composition, music listening, and singing activities in their music therapy program led to psychosocial wellbeing outcomes. The music therapy program investigated by Choi (2010) used similar activities. She credited the notable improvement in psychological symptoms observed to the “innate characteristics of music therapy activities” (2010, p. 402), which aided communication and support processes, and provided a safe way for participants to 58

express feelings within the group. Similarly, McFerran et al.’s (2010) positive findings were attributed to the appropriation of musical activities which created an atmosphere of fun and safety in which issues of grief and loss could be addressed. Montello and Coons (1998) likewise suggest program benefits observed in their study were linked to the fact students chose the type of music used, and therefore remained engaged in the program. They go on to state the need to ensure music activities suit the personality of participants. Rickson and Watkins (2003), also note the importance of tailoring music programs and activities for each population. Such observations are supported by music therapy theorists such as Wigram (2004). McFerran (2010a), who, in her book Adolescents, Music and Music Therapy, also stresses the need for music interventions to appropriate activities and musical styles relevant and appropriate to the young people involved in order to foster both engagement and wellbeing benefits. The idea programs that place little emphasis on education, training or discipline are more engaging and more likely to promote wellbeing is supported by assessments of similar therapy-based programs. Currie (2004) and Currie and Startup (2012), who both evaluated a drumming program aimed at addressing anger and aggression, indicate the program was successful in facilitating several psychosocial outcomes. For Currie (2004) this was because the collaborative percussion activities facilitated group cohesion, and that the enjoyment felt in sessions meant participants returned to sessions despite disengagement from school, and conflict with peers and family. Together, these studies make a strong case to support Rickard, Bambrick et al. (2012) and Grimmet et al.’s (2010) claim that the nature of musical activities can affect the way students engage with a music program or intervention. They also support the idea that both engagement levels and the musical activities themselves can influence the acquisition and reporting of psychosocial wellbeing benefits. Delivery approach. Closely related to the choice of musical activities and student engagement is the conclusion from some authors that programs not tailored to students and school communities are ineffective in promoting psychosocial wellbeing. Grimmet et al. (2010) state this case strongly in their qualitative review of a program that was delivered didactically using a pre-determined structure. Authors suggest a major factor that prevented psychosocial benefits was that those who designed and introduced the music program to the schools did not consult school staff or students before doing so. 59

Instead, they imposed a program that was conceived externally, with little regard for the school context, or consideration of student and staff needs. This lack of consideration was said to have several important ramifications: students were asked to learn instruments unappealing to them, and thus did not fully engage in the program; school staff did not fully engage either, meaning the program lacked quality and coordinated facilitation; and, as the program purpose was not properly communicated to schools as a whole, these communities did not fully appreciate or make the most of the musical opportunities presented to them. Authors made several recommendations for future programs, focusing on the need to consult with schools and students before delivering a program. They conclude that in order to deliver successful programs, it is critical to “establish a collaborative [approach] through the recognition and development of individuals’ strengths and interests” as well as “actively plan for student engagement by listening to and valuing students’ opinions” (Grimmett et al., 2010, p. 63). The argument that music interventions should take a collaborative and tailored approach to program delivery is consistent with music therapy theory. Concepts of collaboration (Stige, Ansdell, Elefant, & Pavlicevic, 2010), democracy, and the cofacilitation of musical activities – often referred to as a “participatory approach” – are noted as vital to group and community music therapy activities (Rolvsjord, 2010; Stige & Aarø, 2012). When working with young people, McFerran (2010a; 2005) specifically stresses the importance of consulting participants, and providing opportunities for their input into the activities and music styles used in interventions. This democratic, co-facilitated approach is described in several music therapy studies in this area. In McFerran et al.’s study, this was typified by the description of the therapist’s approach, in that they “maintained a stance of empathic positive regard rather than the more directive or suggestive approach common to groups run from a cognitive behavioural or educational standpoint” (2010, p. 550). Similarly, Rickson and Watkins (2003) describe the humanistic approach taken by the therapist in their study, and note that even though the program followed a psychotherapy model, it was altered to meet the needs of students, and provided opportunities for their input. In both cases, authors attribute reported benefits to this collaborative program delivery approach. Similar approaches are described in several other music therapy studies, each of which also reported notable psychosocial outcomes (Baker & Jones, 2005; Cheong-Clinch, 2009; Choi, 2010; Gooding, 2011; McFerran & Teggelove, 2011). 60

Other music therapy researchers expand on the importance of this model. Jones et al. (2004) note the need to be attentive not only to music choice, but also cultural and musical elements familiar and appropriate to specific clients. They claim failure to do this potentially limits the ability to achieve optimal participant outcomes. Authors also declare that even the use of common and established music therapy techniques and theory may at times be inappropriate and detrimental to program goals. This sentiment is supported by Montello and Coons (1998) who, also investigating a music therapy program, suggested facilitators did not sufficiently adapt the techniques used to participants. Although students were given a choice in musical styles, standardised models of music therapy were imposed upon participants without fully assessing their needs, and were therefore counterproductive for psychosocial wellbeing in some cases. These latter studies suggest it is the collaborative and democratic features of these programs, rather than their therapeutic focus, that are important for psychosocial outcomes. This distinction is supported by Rusinek’s (2008) study. While he investigated a classroom-based music instruction program, it involved a similar ethos to many music therapy programs above: it used an “inclusive pedagogy” (2008, p. 15), and focused on collaborative and democratic, rather than didactic, delivery methods. Having significant input into the program “allowed students to experience 'learner agency' [as well as] collaboration and peer-learning [and] responsibility, because of the social nature of the activities and the engagement of each student” (2008, p. 16) in decision making and program planning. Several benefits were attributed to this approach. Vaughan et al. (2011) and McKegg et al. (2012), who both reported on wholeschool education-focused programs, also suggested that while their programs followed protocols, they were tailored to the needs of individual groups and involved a choice of several activities. Furthermore, in both cases, those running the programs made a considered effort to consult with the school communities to make sure programs were delivered in a way that met the needs of each individual school community. Again, both studies reported notable psychosocial outcomes. Further supporting the idea that consultation and collaboration are necessary for programs which aim to promote psychosocial wellbeing comes from the investigation of several programs that were imposed upon schools, and which reported mixed (KnoxAnderson & Rickard, 2007) or no benefits (Michel & Farrell, 1973; Rickard, Bambrick, et al., 2012; Schellenberg, 2004). The idea that the failure to consult schools and tailor programs accordingly are significant barriers to the acquisition (and reporting) of 61

psychosocial benefits also accords with research investigating music program delivery in schools (Crooke & McFerran, 2015; McFerran & Crooke, 2014). However, the totality of these claims can be challenged by a number of music instruction programs that were imposed and delivered didactically, and for which positive benefits were reported (Costa-Giomi, 2004; Devroop, 2012; Hietolahti-Ansten & Kalliopuska, 1990). Yet, with the exception of Schellenberg’s (2004) study, these programs all share one factor differentiating them from all other didactic music instruction programs described above: each was delivered outside the classroom, either in a school band setting (Devroop, 2012), or as private instruction (Costa-Giomi, 2004; Hietolahti-Ansten & Kalliopuska, 1990). While authors of the latter two studies suggested reported psychosocial outcomes may related to the benefits of long-term music instruction (i.e. responsibility, self-expression, concentration, self-control and performance opportunities), rather than the instruction itself, some authors consider the private setting of these programs an important factor (Rickard, Bambrick, et al., 2012). Setting. The idea that the setting of music participation plays an important role in the acquisition of psychosocial benefits is one that has received notable attention in research literature. This is particularly evident for studies that have investigated the impact of classroom-based programs. For example, Shields (2001) claims the inconclusive results for their music intervention were due to the fact the classroom in which it was delivered provided an uncontrolled setting. This meant sessions were disruptive and facilitators often spent more time attending to behavioural issues than musical activities. As such Shields advocates for more controlled settings. Similar recommendations are reported by Montello and Coons (1998), who also stress the need to use more structured settings for group music programs, especially when behavioural issues are present among students. Rickson and Watkins (2003) also reported that group music sessions exacerbated behavioural problems, and that this effect did indeed vary across settings. Specifically, when delivered in a residential setting, student behaviour was reported to improve after a session, while in classroom settings, it was said to deteriorate. Rickard, Bambrick et al. (2012) also note the limitations of classroom settings, claiming this factor explains why their music training intervention was less effective than other music training programs, such as those investigated by Costa-Giomi (2004), 62

which were delivered privately. They confer with Pitts and Davidson (2000) that classroom settings prevent facilitators from effectively catering to varying ability levels, and giving students the individual attention received in private lessons. Rickard, Bambrick et al. also suggest positive results found by those such as Costa-Giomi (2004) may be attributed to the fact that private instruction was delivered as an extra-curricula activity. In an effort to explore this, the second of their studies investigated a novel program designed and delivered by external staff, but still in a mainstream classroom. Again, however, they failed to find benefits. As such, they conclude that "instruction might literally need to be offered outside of school" (2012, p. 13) to achieve psychosocial benefits. While this is supported by findings from Hietolahti-Ansten and Kalliopuska (1990), it does not explain the lack of benefits reported in Schellenberg’s (2004) investigation of a private, externally provided music instruction program for students. Nor does it explain the results of Costa-Giomi’s (2004) study, which investigated participation in schools (albeit in private settings). The idea that it is the classroom setting that limits the ability for studies to report psychosocial outcomes is reinforced by numerous studies that did report some positive benefits for programs delivered in school, yet outside of a regular classroom environment. This included programs that were delivered privately (Jones et al., 2004), and those that were delivered in groups removed from traditional, or mainstream, curriculum-based classrooms settings. For example, while perhaps delivered in a physical classroom, participants were not selected based on class-groups, and programs were not curriculum-based (Baker & Jones, 2006; Cheong-Clinch, 2009; Choi, 2010; Devroop, 2012; Gooding, 2011; Kim et al., 2006; McFerran & Teggelove, 2011; McFerran et al., 2010; Rickson & Watkins, 2003). Again, however, the adjudication that regular classroom settings are incapable of procuring psychosocial benefits is challenged by the numerous programs have found benefits in classrooms settings (Kirschner & Tomasello, 2010; McKegg et al., 2012; Rusinek, 2008; Vaughan et al., 2011). This suggests, while it may play a factor, the setting of a program alone cannot account for inconsistent evidence in this field of study. Other factors. Further factors reported in the research literature to play a potential role in the success or effectiveness of a music program or intervention, include the group size and choice of facilitator. Rickson and Watkins (2003) report in their article that smaller 63

group sizes may be more suitable when working with students with behavioural issues, and that future studies using smaller groups may avoid the disruption experienced in their sessions. In respect to facilitation, one of Grimmett et al.’s (2010) key recommendations is that programs should aim towards employing and maintaining staff skilled in program delivery to maximise outcomes. Rusinek (2008) also makes a particular note that the attributes of the program facilitator in their study were instrumental in achieving positive outcomes. This included significant musical skill and industry knowledge, openness to collaboration, and their professional industry-based (rather than educational) approach to the program and participants. Conclusion As this chapter has shown, researchers identify many potential challenges to investigation in this area, which serve as potential explanations for the inconsistent results reported both within and between studies. However, just as reported benefits in this area are inconsistent, so too are the limitations offered to account for this. Authors often attribute their inconsistent, insignificant or absent findings to several factors within the one study, making it difficult for readers or future researchers to determine which factors had the most impact, and are most important to address in future research. Further complicating this issue is the fact that in some cases, a limitation reported by one author is directly contradicted by evidence or a limitation reported by others. This would suggest the challenges in this area of research are not well understood, and there is no clear consensus regarding the best way to approach the design and facilitation of studies in this area. Despite this lack of consensus, several patterns do seem to emerge as reasonably strong predictors of whether a study will report positive benefits or not. In respect to study design and methodology, it would appear that qualitative research approaches, in general, are better suited to research in this area. In regards to the type of musical participation, it seems programs delivered outside of mainstream classroom settings, employing sufficiently engaging activities, and tailored to schools and young people are more likely to be beneficial for psychosocial outcomes. Interestingly, despite the fact some of the issues identified here have been evident within literature for 15 years or more, researchers have continued to investigate in-class music education programs using quantitative methodologies (even when they themselves have reported the challenges of these approaches). 64

When combined with the continuing inconsistency of findings in this area, the fact researchers continue to use research approaches already identified as challenging suggests there are a number of persisting presuppositions in this area. These include an apparent assumption that quantitative methods are appropriate for capturing the psychosocial wellbeing benefits of school-based music participation, and the presumption that class-based music instruction is appropriate for facilitating them. The continued use of such approaches is, however, perhaps understandable. Each and every challenge reported in this area contains some level of ambiguity. Further, while limitations are reported, they are very rarely discussed at length in the literature, often articulated in one sentence, or at most a paragraph, in limitation or discussion sections. Indeed, in the process of undertaking this PhD project, at no stage has any synthesis or review of the research challenges related to this particular area become apparent. Again this suggests that while limitations are identified, they are not sufficiently understood or articulated to guide future research. However, while limitations reported in this chapter do not provide a definitive explanation for why research has so far been unsuccessful in determining whether musical participation in mainstream schools promotes psychosocial wellbeing, it does provide areas for further study. Firstly, it suggests that in order to reach consensus in this area through research, a more informed understanding of the potential challenges in undertaking investigation in this field must be reached first. Secondly, it appears that an understanding of challenges must be sought in two distinct areas: research design and methodology, and the nature of musical participation. Perceivably, only when challenges in these two areas are sufficiently identified, understood and acknowledged, can researchers hope to undertake the type of research needed to fully comprehend the relationship between musical participation in mainstream schools and psychosocial wellbeing. Further it is only when researchers comprehend and report consistent evidence regarding this relationship that policymakers can be expected to write and enact appropriate policy.

65

Chapter 5: Project Methodology and Design Project Rational Despite claims in policy literature that musical participation in mainstream schools leads to increased student psychosocial wellbeing (Parliament of Victoria, 2013), research has so far provided limited evidence for this causal relationship (Rickard, Bambrick, et al., 2012). This lack of consistent evidence has been linked to the research methods used (Kim et al., 2006) and the attributes of music programs investigated (Shields, 2001). Project Aim The aim of this project was to identify the challenges of investigating the psychosocial wellbeing benefits of musical participation in mainstream schools. The outcome of the project was a series of recommendations to guide the type of research needed to inform policy makers. These recommendations address inconsistencies between policy claims and research evidence in this area. Research Questions The research question used to achieve this aims was: What are the challenges of investigating the psychosocial wellbeing benefits of musical participation in mainstream schools? This question was addressed by asking two sub-questions: 1) What are the challenges related to research method and design? 2) What are the challenges related to the attributes of a music program? Methodology The philosophical and methodological stance taken in this project aligns with aspects of both participatory and critical theory research paradigms (see Table 5.1). While some scholars (Lincoln, Lynham, & Guba, 2011) argue that these paradigms present different approaches to research, they are often commensurable and share many attributes (as indicated by the green shaded cells in Table 5.1).Taken together, the tenets of these paradigms support the theoretical approach taken in this project, locating the project within current theories and practices in contemporary social science. The theoretical framework of the current project is presented here using the language of

66

participatory and critical theory paradigms, along with more specific theoretical concepts also congruent with the project. Table 5.1 Methodological Elements of Project Aligning with Participatory and Critical Theory

Values (Research is valuable when it…)

Nature of knowledge (Reality/knowledge is …)

Participatory

Critical Theory

“Transform[s] the world in the service of human flourishing”

Promotes social justice by changing “social policy and practice”

Critically engages with research and knowledge generation to serve best interests of society

Undertakes critique/revelation/historical revisionism to erode social oppression

Engages participants as coconstructors of knowledge for political reasons

Engages participants as co-construtors of knowledge for political & practical reasons

Inherently subjective, yet can be understood as subjective-objective reality

Inherently subjective, yet can be known through qualitative and quantitative methods

Individually & socially constructed

Socially constructed

Non-rational

Rational

Multifaceted (different versions may co-exist) Known through experiential connection with a phenomenon Representative of and connected to power structures Knowledge accumulation (New knowledge is known through…)

Inquiry posture (The researcher approaches inquiry as…)

Practical and participative experience of a phenomenon in the context of interest Deconstruction of dominant notions of what is right, good or normal

Critique/historical revisionism of established processes and beliefs

The primary voice of research, gained through critical self-reflection An activist and “transformative intellectual” who knows how to use knowledge/research to produce fairness and social justice

Note: Shaded green cells show points of commensurability between paradigms. Adapted from Paradigmatic controversies, contradictions, and emerging confluences, revisited (pp.101-15), by Y. S Lincoln, S. A Lynham, and E. G. Guba, in The SAGE handbook of qualitative research by N. K. Denzin & Y. S. Lincoln (Eds.), (4th ed., pp. 97-128), 2011, Thousand Oaks, CA: SAGE Publications Inc.

Values. The values underpinning a project can be understood as the personal beliefs of an individual that serve as the motivators or rationale for them engaging in research as a

67

human endeavour. The values underpinning this project include a number of enquiry goals and value judgements that were not necessarily addressed directly by this project, but rather serve to communicate beliefs about the role of scientific enquiry more generally. These are presented here to illustrate the underlying beliefs and commitments to research which informed and shaped this project, otherwise known as axiology: the values of an individual that are seen to influence all stages of the research or inquiry process, including “choice of problem, choice of paradigm to guide the problem, choice of theoretical framework, choice of major data-gathering and data-analytic methods, choice of context, [and] choice of format(s) for presenting findings” (Lincoln et al., 2011, p. 116). The goal of supporting human flourishing was the core value driving this research, where flourishing is defined as optimal social and psychological functioning, health or wellbeing (Huppert, 2013; Keyes, 2002). This goal has driven the decision to gain a more informed understanding of issues involved in investigating the relationship between musical participation and student flourishing, and the intention to communicate subsequent recommendations to researchers and policymakers in this area. This value is consistent with the participatory paradigm position that resesarch should aim towards “action to transform the world in the service of human flourishing” (Lincoln et al., 2011, p. 108). Linked to the goal of human flourishing was the desire to promote social change and equity, embodied here as an intention to work towards equitable access for all students, via policy support, to the wellbeing benefits of musical participation. This value more closely aligns with the critical theory paradigm presented by Lincoln et al. (2011), where social justice is both the reason for, and the intended outcome of scientific inquiry. Critique itself represents another value that underpinned this project. A commitment to critique was considered here as the willingness to challenge presuppositions or power structures that exist in theory, practice or policy, and to explore how these may obstruct goals of flourishing or equality. This position is consistent with the critical theory position that research should be oriented towards revelation which will “improve social justice and remove barriers and other negative influences associated with social oppression” (Lincoln et al., 2011, p. 112). This project was also based on the belief that critical engagement is the responsibility of those who act as the authoritative constructors of knowledge in a society. This is so knowledge, 68

both in its construction and practical application, serves the best interests of a society. Such a position aligns with the participatory framework, where knowledge is created to “change the world [which] implies engagement [and] responsibility” on the part of the researcher (Lincoln et al., 2011, p. 112). Another key value was that research should be collaborative. This was defined here as the need to seek the voice of participants or players in a field (seen here as students, teachers and researchers), and valuing their individual representations of reality. This is consistent with participatory and critical theory paradigms, both of which value an approach to research in which participants are offered the opportunity to become co-creators of knowledge (Lincoln et al., 2011). Nature of knowledge. The approach to knowledge that shaped the way this project was undertaken is commensurable with the participatory paradigm. An important element of this is recognition of the utility of both subjective and objective conceptions to knowledge. The dual use of subjective and objective notions of knowledge can be understood in terms of Heron and Reason’s (1997) concept of “subjective-objective reality”, which Lincoln et al. (2011) locate within the participatory paradigm, and is interpreted as follows: while what we know is essentially known subjectively, in order to communicate what we know, and collaborate in a shared pursuit of common knowledge, we can or must formulate concepts and tools for knowledge creation that take an objective form (in order for them to be known or understood by others). While these tools or concepts are socially constructed, and understood subjectively, their construction aims towards a shared knowledge or understanding of what can be known. This can be seen as a way to navigate or negotiate between extreme subjectivity (where nothing is real because everything can be reduced to a social or mental construction), and the pragmatic needs of communicating (and investigating) our everyday lived experiences, where some degree of objectivity is necessary. This last point aligns with several authors from fields such as education (Bleazby, 2011), research theory (Andrews, 2012) and philosophy (Bernstein, 2011) who suggest that, in a pragmatic sense, to focus solely on either subjective or objective notions of reality is unpractical, problematic, and potentially unproductive. For example, Bernstein (2011) argues that only when we move past dichotomies related to objective and (the potentially nihilistic depths of) relative philosophy, and take a hermeneutic 69

approach to knowledge, will we achieve praxis; or “communal action among equals in which public freedom becomes tangible” (Bernstein, 2011, p. 225). This, in part, explains the use of both quantitative and qualitative data in this project. Importantly, however, while this project valued the role of objectivity as a pragmatic tool in the generation and dissemination of knowledge, it operated on the underlying premise that reality is inherently subjective. This was based on the understanding that knowledge is socially constructed; something that is known in the eyes of the knower, not derived from an external or objective reality. Lincoln et al. (2011) note this perception common to both critical theory and participatory paradigms. However, following Kilgor’s (2001) distinction, this project assumed knowledge is multifaceted and not necessarily rational, which suggests a closer alignment to Lincoln et al.’s description of participatory theory, than critical theory (see Table 5.1). Following the supposition that knowledge is socially constructed and multifaceted, this project was grounded in the belief that there is no one absolute truth; that multiple constructions of reality may coexist between groups or individuals. While this position also fits with both constructivist and social constructivist positions (Andrews, 2012), it can be distinguished from these forms of constructivism. This is because constructivism is said to focus on an individual’s mental or cognitive construction of knowledge, while social constructivism focuses on the role of social or environmental factors (Young & Collin, 2004). A participatory approach, on the other hand, is seen to accommodate individual and collective constructions of reality, formulated through both the experiences of an individual and their relationship to their environment (Lincoln et al., 2011). This view accommodated the use and discussion of different versions of reality reported by different players (or reporters) in this project. This included versions that were formulated predominantly via mental or cognitive processes, and those that were informed or constructed in relation to social or contextual factors. During this project importance was also placed on the role of gaining knowledge through practical experiences, or the need for an experiential connection with a given phenomenon in order to properly understand it. For this reason, the project explored the practical and lived challenges of investigating wellbeing in a school context. An investigation of these challenges necessitated real-world experience of the processes and contexts involved. This is consistent with the participatory paradigm, which advocates the “Primacy of practical [or] living knowledge” (Lincoln et al., 2011, p. 106), 70

where practical knowledge or “knowing” “presupposes a conceptual grasp of principles and standards of practice […] and experiential grounding in the situation within which the action occurs” (Heron & Reason, 1997, p. 281). This project was also undertaken with the acknowledgement that social or systemic power structures may impede social justice or human flourishing through a preference for certain forms of knowledge; i.e. the preference for objective knowledge by policymakers (Torrance, 2011) and academics (Finlay, 2002a). The recognition that systemic-based preferences for knowledge, and knowledge itself, may reflect and or represent power structures aligns with critical theory (Lincoln et al., 2011). Inquiry posture. The voice of the primary inquirer in this project (i.e. me) is the primary voice through which knowledge was known and communicated. This choice was particularly pertinent given the study’s use of critical reflection as the main analysis approach, or method of knowing. Knowledge in this study was not generated through the rigorous use of established or empirical research tools and the presentation of subsequent data. Rather, it was through reflection on and critique of data gathered – as well as the processes and contexts of data gathering – that knowledge was produced. In this way, it was not empirical data that was considered the authoritative voice, but reflections on research data, processes and contexts in which this took place. This is consistent with the “inquirer posture” in the participatory paradigm, where the “Primary voice [becomes] manifest through aware self-reflective action” (Lincoln et al., 2011, p. 101). Also congruent with the participatory approach, secondary (or participant) voices were seen as “illuminating theory” (Lincoln et al., 2011). Participants were important contributors to knowledge generation, as they provided practical insight into a given phenomenon not necessarily accessible to a researcher in isolation. Further, empirical data was consciously and purposively appropriated in a way that critiqued existing research approaches, and was used to form an argument for better quality research that will ultimately lead to more informed policy and more equitable access to music in schools. This is consistent with the critical theory inquirer posture, which sees the researcher as an activist who understands how to use research and intellectual processes to promote social justice (Lincoln et al., 2011).

71

Quality criteria. To address “legitimacy” in research findings, reflexivity has been employed throughout the dissertation to provide a transparent account of the research process. Transparency is often used for addressing rigour, or “trustworthiness” in non-positivist reporting methods (Shenton, 2004; Sinkovics & Ghauri, 2008). It involves engaging in an ongoing dialogue throughout the research process that reflects on how data or results may be affected by things like context, the researcher’s interaction with data, participants, or the research question. Through this dialogue, an explicit account is provided of intentions, biases, dilemmas and other factors that may impact results (Finlay, 2002a, 2002b; Stige, Malterud, & Midtgarden, 2009). The purpose of incorporating reflexivity into this study was to provide a transparent account of the research project, and offer markers by which not only qualitative and quantitative methods could be evaluated, but also the use of critical reflection as the central methodological tool. This has partly been addressed by stating the motivations and background of the researcher in the introduction. Further efforts to achieve transparency have been made by providing explicit reflections on how data was collected and analysed, and openly acknowledging where decisions, observations, results, or conclusions were based on personal reflections or interpretations. Use of the first-person. To be explicit about where a “reflexive dialogue” (Finlay, 2002b) with researcher intentions, experiences of data collection, data, and results occurred, this dissertation sometimes employs the use of the first-person narrative. The first-person voice is used only where such a narrative is consistent with the epistemological nature of that research activity – for example where reflection itself is used as the primary source of narrative, data collection, or analysis. As such, the “first-person” voice is used in the introduction, the reporting of interview and observation results, and in the write up of the critical reflection analyses. The use of the first-person was not considered necessary or appropriate when communicating the research approach or methods used, or when reporting quantitative results. Importantly, the restricted use of the first-person does not suggest that researcher interpretation or bias was not present throughout the whole project. In keeping with Finlay (2012), the position taken here is that the influence of the researcher is present throughout all research, regardless of whether it is reported. Therefore, the selective use

72

of the first-person in this dissertation does not aim to marginalise reflexivity in certain parts of the project, but rather to celebrate where it is considered a fundamental element. Project Design As depicted in Figure 5.1, the overall design of this project was inherently emergent in nature, and comprised of two distinct stages. The first of these involved the analysis of data taken from two evidence-based studies, each of which aimed to provide empirical evidence for the psychosocial wellbeing benefits of different school-based music programs. Both studies used quantitative and qualitative methods, and data analysis showed inconclusive evidence for a relationship between musical participation and student wellbeing.

Figure 5.1 Flow chart of overall project design, showing the two stages of analysis and the stage of reflexive evaluation. The inconclusive nature of results from theses studies, combined with several other factors outlined in the introduction chapter, led to a state of doubt (or dissonance) regarding the suitability of established research approaches in these settings (research “approach” refers not only to the methods used, but also the type of musical participation investigated). To explore this notion, the decision was made to critically reflect on the methods, data, results, contexts, and music programs of the two studies in order to identify the challenges of investigating the psychosocial wellbeing benefits of school-based musical programs. This process of critical reflection constitutes the second stage of the project, and consisted of two independent systematic analyses. The first aimed to explore the challenges of investigating the psychosocial wellbeing benefits of musical participation in schools that relate to research methods and designs. The second focused on the 73

challenges related to the types of musical participation being investigated. Each critical reflection analysis used all data from both studies in the first stage, as well as observational and contextual data gathered in relation to the research processes used in the first stage. By critically reflecting on the research approaches of the first stage, each analysis in the second stage identified several challenges of investigating the subjective wellbeing benefits of music in schools. These challenges were subsequently used to form a set of recommendations for the design and facilitation of future research. While it is the results and recommendations from the critical reflection analyses (the second stage of the project) that represent the knowledge generated in this PhD, a comprehensive account of the data collection, analysis, results and context of the first stage are presented first. This has been included for two reasons. First, to give a linear account of the activities and process that occurred throughout the project (as shown in Figure 5.1). Second, to present a detailed description of the material that informed the critical reflection analysis. Without intimate knowledge of the methods, data, results, music programs, and contexts of the studies in the first stage, the subsequent critical reflection analyses on these studies would not have been justified, nor possible (for an account of the theoretical or philosophical approach to knowledge generation in both stages, see Appendix A). Following the aforementioned linear approach, the remainder of this chapter gives a brief account of the different elements included in the project design (shown in Figure 5.1), including a description of the methodological approach used for the critical reflection analyses. To establish a consistent set of labels to explain the different elements of this design, and for ongoing reference throughout the dissertation, it was considered necessary to utilise three main terms: “empirical study”, “critical reflection analysis”, and “project”.

74

Table 5.2 Definition of Terms Used to Describe Overall PhD Design Term

Definition

Use of term in dissertation

Abbreviation

Empirical study

Form of investigation using established research tools to evaluate effectiveness of an intervention

Used to refer to two evidence-based studies in stage one of project:

Study/ studies

Reflective evaluation and critique of process, belief or assumption. Used to identify problematic elements of said process, belief or assumption

Used to refer to two systematic analyses in stage two of project:

Critical reflection analysis

Project

All stages, processes & activities that make up a research project

- Keys to Success - Bereavement Group CRA

- CRA of research method & design - CRA of music programs investigated Used to refer to overall PhD project, including both stages, reflexive evaluation, and all research activities

As shown in Table 5.2, “empirical studies” (or “studies”) is used to refer to the two existing studies that provided evidence-based data via standard research approaches. While it is acknowledged the term “empirical” has a complex etymology (Williams, 1985), it is used here to denote the action of employing scientific materials to conduct experiments in order to produce knowledge. This is closer to the common usage of the term between the 16th and 19th Centuries than the contemporary use of the term; where is often used to refer to positivist models of scientific enquiry (Williams, 1985). The particular appropriation used here aimed to distinguish between investigations that use research tools (i.e. surveys or interview schedules) to produce new evidence that is verifiable (through either quantitative or qualitative methods), and investigations that use critique and reflection to produce knowledge through the deconstruction and revision of existing knowledge or practices. “Critical reflection” (or “critical reflection analysis”) is used to signify the process of analysis in which these “studies” are reflected upon and critiqued. The word “project” is used to indicate the PhD project as a whole. Stage 1: empirical studies. Data for the empirical studies was made available by an existing research team for analysis in this project. A full account of the design and method of the empirical studies are given, however, this is to provide background information relevant to the critical reflection analysis only. 75

Empirical study 1: Keys to Success. The first empirical study collected both quantitative and qualitative data in order to investigate the benefits of a class-based music program. To use mixed methods research (MMR) terminology, this study used a sequential mixed design (Teddlie & Tashakkori, 2006). Survey data was collected and analysed first. Based on the perceived need to explore the results further, follow-up interviews were conducted with selected students and the class teacher. Observational reports were also collected throughout the study. Qualitative and quantitative data sets for this study were analysed independently. A detailed explanation of the collection and analysis of survey and interview data is presented in Chapter 6, and results are presented in Chapter 7. Empirical study 2: Bereavement Group. The second empirical study also collected quantitative and qualitative data. Here, the aim was to evaluate the design, facilitation, and benefits of a school-based music therapy program delivered for bereaved students. Again using MMR terminology, this study followed a concurrent design (Creswell & Plano Clark, 2011), meaning qualitative and quantitative data sets were collected independently, and in parallel. Analysis of these data sets was also undertaken independently, and observational data was again recorded during the study. Details regarding the collection and analysis of survey and interview data are presented in Chapter 6. All results are presented in Chapter 8. Reflexive evaluation. Analysis of data from the two empirical studies (see Chapter 6) returned inconclusive results (see Chapters 7 and 8), and thus provided little insight into the relationship between music in schools and psychosocial wellbeing. The manifestation of these ambiguous results coincided with a realisation that the majority of published research in this area had reported similarly inconclusive findings. A review of this literature (see Chapter 4) suggested these inconclusive findings were due to challenges embedded within the research approaches used, including methodological and design issues, and the types of music programs examined. These challenges also appeared to be present in the two empirical studies included in this project. Further, it appeared research challenges were underpinned by assumptions regarding the suitability of common research approaches in this area. Similar 76

assumptions – particularly regarding the type of music program deemed appropriate for facilitating psychosocial wellbeing – also appeared in policy literature regarding music in Australian schools. These factors led to the decision to use the data from the two empirical studies to critique the research approaches they employed, rather than simply reporting the insignificant outcomes. A particular method of critical analysis, akin to the “critical reflection” process described by Mezirow (1990), was identified as the most suitable way to reflect on the available data from these research approaches, and identify the problems and challenges they contained. Stage 2: critical reflection analyses. The methodology used in the critical reflection analyses is presented here as it represents the methodology used to generate the primary outcomes and recommendations of the project as a whole. It therefore represents the primary methodological approach of the PhD. The method of analysis used in this approach can be likened to John Dewey’s process of “reflective thinking”, which he describes as the logical organisation of thoughts aimed at the generation of knowledge, to inform the “acceptance or rejection of something that is reasonably probable or improbable” (Dewey, 1933, p. 4). Dewey claims “reflective operations” involve two steps, or sub-processes; “(a) a state of perplexity, hesitation, or doubt; and (b) an act of search or investigation directed toward bringing to light further facts which serve to corroborate or to nullify the suggested belief” (1933, p. 9). In this project, the first step represents the perplexity that arose when I was faced with the lack of evidence supporting a link between psychosocial wellbeing and music in mainstream schools. How can this be? I thought, there is a wealth of evidence that shows music has a profound impact on people’s wellbeing in other areas (such as chronic illness, mental health, disabilities, and in community settings). Why hasn’t there been consistent evidence reported regarding the wellbeing benefits of music in schools? It just doesn't make sense! This extended to the results of my own work: I’m sure the music programs had a significantly positive effect on students. The staff seemed to think it did. So why didn't this effect show up strongly in all interviews, or in any survey results at all? The other source of bafflement was the fact that, despite this lack of evidence, writers in the field of policy have made strong claims about these benefits: 77

how can these documents be claiming that music will increase the emotional and social wellbeing of all students if there is little to no rigorous evidence supporting this idea? The doubt and hesitation Dewey also includes in this first step corresponds to the growing doubt I had about the suitability of the research designs and methods predominantly used in this area – including both those written up in the literature, and those used in the two empirical studies included in this dissertation. Informed by my experience of undertaking research in schools, I questioned, did students really understand or take the research seriously? If not, can we expect that these results reflect their real experiences? I don't think so. I believe we may need to take a different approach if we want to really find out what is happening here. Further, I began to doubt whether participation in the form of class-based music education, recommended in policy and preferred by researchers in this area, was really capable of promoting wellbeing to a level that would be reported by students. The focus on this type of program seems to be based on the assumption that class-based music education will achieve wellbeing because it achieves the other musical benefits valued in schools, I observed, but that doesn't fit with my experience at school, or with theoretical models from fields such as music therapy. If such doubts were justified, they appeared to offer an explanation for the lack of evidence in the area. No wonder there are few reports of increased wellbeing, they are using inappropriate tools to investigate inappropriate programs. The subsequent critical reflection analyses can be seen as representing the second step in Dewey’s process; the “act of search or investigation directed toward bringing to light further facts which serve to corroborate or to nullify the suggested belief” (1933, p. 9). In this project, the beliefs of interest were: that existing research methods and designs were appropriate for this area and, that all types of musical participation are capable of promoting wellbeing. To assess these beliefs, two critical reflection analyses were undertaken using the information gathered relating to the two empirical studies. This information was used to identify any factors that presented challenges to investigating the psychosocial wellbeing benefits of the two music programs (i.e. was there anything about these studies that prevented them from reporting positive results). Once identified, these challenges were used to critically examine whether the two beliefs would be supported or contradicted by what occurred in these studies – if the research methods favoured in this area were suitable, and if class and curriculum-based music programs were capable of promoting wellbeing for all 78

students. By critically assessing the validity of these assumptions or beliefs in these two studies, this process of analysis aimed to provide insight into their validity in this field more generally, and thus provide a basis from which to make recommendations for further research (and policy). Terminology wise, similar analytic approaches have been used in other fields. “Critical reflective analysis” is sometimes used in nursing research (Chambers, 1999; Ray, 1990), and “reflective analysis” has been reported in social work literature (Osmond & Darlington, 2005). However, these are predominantly used to describe methods of practice evaluation where a certain “process” is considered the primary source of data. While the “processes” of research and musical participation were used as sources of data in this project, a more comprehensive range of sources was reflected upon. These included sources described in Mezirow’s account of “critical reflection”, and relate to “epistemological, logical, ethical, psychological, ideological, social, economic, political, ecological, scientific, spiritual, or [those which] pertain to other aspects of experience” (1998, p. 186). In further keeping with Mezirow’s conceptualisation, reflection was used here to both enable “us to correct distortions in our beliefs and errors in problem solving” (1990, p. 1), and to go deeper by undertaking a “critique of the presuppositions on which our beliefs have been built” (1990, p. 1). Therefore, the term “critical reflection analysis” was used here to capture the two-step process described by Dewey’s idea of “reflective thinking”, and the focus on problem solving and comprehensive data sources used in Mezirow’s concept of “critical reflection”. It also aimed to appropriate the affordance provided by both theorists to consider the impact of underlying beliefs or presuppositions. Critical reflection analysis 1. For the first critical reflection analysis, the “act” of investigation aimed to “corroborate or to nullify” the belief regarding the suitability of using established (and often positivist) research methods to investigate the psychosocial wellbeing benefits of music in mainstream schools. To do this, I reflected on how attributes of the data collection methods utilised in both studies affected the subsequent reporting of psychosocial wellbeing benefits. This included reflection on the methods and contextual factors surrounding data collection; student, teacher and researcher perceptions of this process; the data itself; and any other factors related to data analysis.

79

For example, the raw survey data was checked to see if students filled out surveys correctly; survey results were examined for irregularities or inconsistencies; teacher and staff comments regarding the challenges of the research activities were documented; the contextual characteristics of the class were noted; and researcher reflections on survey sessions were recorded. Together, this information led to reflections on the research methods such as: I don't believe it is reasonable to assume that these students understood the survey questions. The raw data doesn't support this, nor do the results. Further, the teachers told me students wouldn't understand the questions, and the students told me this themselves during interviews. What’s more, when I was collecting this data, it was obvious that students were really struggling with the language. As such, this led to the conclusion that students’ ability to understand the language in surveys presents a serious challenge to investigating the psychosocial wellbeing benefits of musical participation. If they don’t understand the question, it is unlikely that their answers will be valid. In turn, this conclusion served to challenge the assumption that using surveys is a valid approach to investigation in this area. All findings of this analysis are presented in Chapter 9. This includes an explanation of how findings were informed by data from the two empirical studies, and a discussion of relevant literature. Critical reflection analysis 2. The second critical reflection analysis assessed the viability of the second belief that any type of music program (particularly class and curriculum-based programs) can and will enhance the wellbeing of all students. To do so, I critically reflected on the attributes of music programs investigated in each empirical study, and the effect they had on reported psychosocial wellbeing benefits. This involved reflecting on practical aspects of the programs such as content, length and size; contextual factors such as setting; what the participants said about the programs; and how different program aspects aligned with the benefits (or lack of benefits) reported. For example, it was observed that more benefits were reported in the Bereavement Group study (Empirical Study 2), which investigated a program run in a semi-private setting, than the Keys to Success study (Empirical study 1), which investigated a class-based program. Further, students in the Bereavement Group study reported in interviews that the private setting was important for the benefits they experienced. Conversely, interview data from students in the Keys to Success study 80

suggested the class-based setting sometimes meant students didn't engage with the program, and therefore experienced less wellbeing-related benefits. This led to the conclusion that investigating a class-based music program poses a challenge to research in this area, as the class setting is less suited to promoting psychosocial wellbeing benefits than more private settings. This challenges the belief that class-based programs are suitable for promoting the wellbeing of all students. Findings from this critical reflection analysis are presented in Chapter 10, along with an explanation of how each finding was informed by relevant data. Also included in this chapter is a discussion of literature considered relevant to the findings. Recommendations. Through the two critical reflection analyses mentioned above, several important challenges for research in this field were identified. These challenges were subsequently used to inform a number of recommendations for future research. These recommendations are arranged in relation to two main topics. Those in the first topic are informed by the first critical reflection analysis, and include recommendations for the design and methods to be used in future studies. For example, in order to address the challenge of student comprehension, it is suggested that researchers carefully consider the language used in research tools, and develop age appropriate and culturally relevant tools where necessary. The second set or topic of recommendations was informed by the second critical reflection analysis, and contains recommendations for the design and delivery of school music programs to be investigated for psychosocial wellbeing benefits. For example, to overcome the challenge that class settings can have on facilitating increased student wellbeing, it is recommended that programs be delivered outside of class, and in private settings. Where possible, all recommendations are discussed in relation to or supported by existing literature. Practical and pragmatic considerations for implementing recommendations are also presented, including a discussion of recommendations which appear critical in any future study, as well as those which should be taken into consideration, and attempted where possible. Such recommendations are considered the primary results of this project, and are predominantly aimed at informing researchers. They are also seen to have potential in guiding future policy development, as well as program delivery undertaken by schools 81

or music practitioners. These recommendations have been published in an academic journal, and presented in this form as Chapter 11 of the dissertation. Discussion. While the aforementioned recommendations chapter includes a discussion of findings, a more general discussion of the project is undertaken in Chapter 12. This extended discussion gives an overview of the challenges identified in the critical reflection analyses, and argues that these challenges are underpinned by assumptions about research in this area, and the social sciences more generally. It then argues the ongoing need to actively critique such assumptions, and the need to implement the recommendations of this PhD to overcome their negative impact in this field. This final discussion chapter also presents a number of recommendations for the field of social and educational policy. This includes the argument that current policy advocacy is based on a misguided assumption that class-based music education will achieve wellbeing in mainstream schools. Further, it argues that this advocacy should be tempered or revised in future policy documentation. Conclusion. The PhD finishes with a reflexive conclusion. This chapter reflects on the candidature journey as whole, and how the initial motivations, experiences and beliefs that drove the decision to undertake the PhD have been affected by candidature. For example, the need to critically engage with one’s own beliefs, motivations and values, as well as those that exist within a given field of enquiry, is discussed as a central element of the PhD. Indeed, this in itself presents one of the most significant points of learning enabled through this project.

82

Chapter 6: Design and Method of Empirical Studies The studies presented in this chapter were designed and carried out as part of a larger multi-school project called Music Matters, run by the Music Therapy Department at the University of Melbourne (UoM). Music Matters ran independently to this PhD project, meaning the two studies presented here were originally conceived for different purposes. However, many of the data collection activities were carried out as part of this PhD with the agreement that the data would be made available for analysis in the dissertation. The method and design described for the two empirical studies presented here do not represent the method used for the PhD project as a whole (see Table 5.2, Chapter 5, for definition of “empirical studies” and “project”). As described in Chapter 5, this project uses critical reflection as its main methodological tool for knowledge creation, rather than provide empirical evidence of any kind. However, given this critical reflection was undertaken on the method, design and data from these two empirical studies, it is considered necessary to provide a thorough account of these elements. As such, this chapter presents a detailed explanation of the design and methods used for both empirical studies. While these studies were originally designed for different purposes, there were some cases in which qualitative methods were shaped by the emergent nature of this project. An explanation of how and when this was done is provided in this chapter, along with a description of the process and rationale for collecting observational reports, which are also used to inform the critical reflection analyses presented later in this dissertation. How these activities aligned with critical theory and participatory research paradigms (Lincoln et al., 2011) is also briefly discussed. Empirical Study 1: Keyes to Success Design overview. The first empirical study was called Keys to Success (KTS). It aimed to investigate the subjective psychosocial wellbeing outcomes of a curriculum-based music program delivered within a mainstream classroom context. Using a mixed method research (MMR) lexicon, the study employed a sequential MMR design (Creswell & Plano Clark, 2011; Teddlie & Tashakkori, 2006). This means it began with

83

an initial phase of quantitative data collection, which was followed by a qualitative phase of data collection. The quantitative phase used a quasi-experimental design (Shadish & Luellen, 2005), in which two Year 7 classes were non-randomly assigned to intervention and control conditions. Students in both classes filled out baseline and follow-up surveys. Upon analysis of quantitative data, the decision was made to undertake interviews to further explore program outcomes and the survey process. Nine participants from the intervention class and their teacher were recruited for interviews. These used semi-structured schedules to interrogate survey results, assess the suitability of using the survey tool with this group, and explore other themes related to the music program not covered in the survey. Interview data analysis occurred independently to the analysis of survey data, yet given interviews were designed to explore the suitability of the survey, some discussion of the survey is included in the interview results. Observational reports of data collection activities were also recorded throughout the study. This included note taking during and directly after data collection activities, documentation of de-briefing sessions between researcher(s) and other Music Matters project staff, documentation of discussions with school staff, and personal reflections. The music program. The music program investigated in the Keys to Success (KTS) study was part of a pre-existing curriculum-based program at the school. This existing program (which contained no musical elements) was also called Keys to Success, and aimed to teach students about life-skills, such as perseverance. All Year 7’s in the school participated in this program for one school term per year (approximately 10 weeks). A music therapist working for Music Matters was invited to facilitate a four-week musical version of the program to provide an alternate way of delivering it to students. While the goal of the existing KTS program was to educate students about life skills that may lead to increased wellbeing, the musical version investigated was not specifically directed at wellbeing outcomes. Instead, it focused on using music to help deliver mainstream curriculum material in a way that was fun and engaging for the students. In this way, the program did not appropriate music as a way to increase psychosocial wellbeing, but used it as a vehicle for curriculum-based learning. The music therapist was predominately responsible for delivering program content and directing session activities, yet the regular class teacher was present and 84

acted as co-facilitator at some times. Importantly, while delivered by a music therapist, this was not a music therapy program. Therefore, it did not follow the collaborative or participatory program models indicative of contemporary music therapy practice (Stige & Aarø, 2012; Stige et al., 2010). This was partly due to the structure of the pre-existing curriculum initiative, and the music program’s focus on achieving curriculum (rather than wellbeing) outcomes. The program lasted for four weeks, with weekly 50-minute sessions. During sessions, students were encouraged to participate in group music-making activities, and work towards the performance of a collective class song, which was videorecorded at the end of the program. Those students who did not feel comfortable performing were encouraged to take either directing or audio-visual roles (see program protocol in Appendix B for more detail). During the four-week study period the control group class participated in the existing non-musical version of the life-skills program, also in weekly 50-minute sessions. Following a waitlist control design, the control group class began the same four-week musical version of the KTS program with the music therapist upon completion of the study. Quantitative method. Participants. Forty-two students and one teacher were recruited from a mainstream government school in Melbourne’s western suburbs. Each student was a member of one of two Year 7 classes whose teachers had agreed to participate in the musical version of the Keys to Success program. Participants were assigned to study groups as classes, with 20 students in the intervention class, and 22 in the control class. This followed a clustered sampling approach (Teddlie & Yu, 2007) in order to access a study population representative of mainstream secondary school students in the state of Victorian. Both classes were considered mainstream in that they contained no elements of specialised learning, and student groups were not identified as at-risk of low psychosocial wellbeing. This study was not designed to conduct analyses related to age, sex, or identification with ethnic or cultural background and therefore data for these variables was not collected. However, the Victorian Government states the upper and lower age limits for Year 7 students are 12 and 15 years of age (DEECD, 2015), and consistent 85

with the overall school demographic, both classes contained students from both genders, and a diverse range of cultural and linguistic backgrounds. All 42 students were invited to take part in the surveys. Information statements were provided for students (see Appendix C) and families (see Appendix D), and active consent forms (see Appendix E) were sent home to parents before the program began. Only students who returned these were included in the study. It was made clear that participation in the survey was not a requirement for inclusion in the program. The UoM Human Research Ethics Committee (HREC; see Appendix F), and the Victorian Department of Education and Early Childhood Development (DEECD; see Appendix G) approved this study. Materials. The survey instrument included scales from several sources. Primarily, scales were taken from the Victorian Adolescent Health and Wellbeing Survey (termed HowRU for administration; DEECD, 2010), which itself uses measures from several sources. The HowRU has been used to assess the wellbeing of young people in Victorian school settings (Jonkman, Steketee, Tombourou, Cini, & Williams, 2014), including large population-based studies conducted by the Victorian Government (Victorian Government, 2010). Data from these government studies has been used to assess national social inclusion indicators for young people (Victorian Government, 2011). The HowRU scales included in the present survey address several areas of student psychosocial wellbeing. The present tool also contained measures taken from the Communities That Care® Ltd. Youth Survey (CTC; Arthur, Hawkins, Pollard, Catalano, & Baglioni, 2002; Bond, Thomas, Toumbourou, Patton, & Catalano, 2000). While the HowRU already contained some CTC measures relating to school, extra measures were included which related to community and neighbourhoods, peer relationships, and other individual factors. The CTC measures used in this study are also consistent with established indicators of psychosocial wellbeing and social inclusion in young Australians, particularly in the areas of community belonging, school engagement, and social behaviour (AIHW, 2011; ASIB, 2009, 2010; Ryan & Sartbayeva, 2011; Saunders, 2011; Scutella & Wilkins, 2010). Furthermore, CTC scales were developed specifically for administration in school settings, and while used internationally (Hemphill et al., 2011), have been adapted for the Australian context (Bond et al., 2000). All CTC scales 86

measure what are termed “risk” and “protective factors” of student psychosocial wellbeing. Evidence shows higher “protective factor” scores and lower “risk factor” scores are linked to better wellbeing outcomes for students (Arthur et al., 2002). For a full list of CTC scales and response options, and the use of these scales in the Australian context, refer to available online resources and fact sheets (www.rch.org.au/ctc/). For an explanation of interpreting risk and protective scores see Bond et al. (2000). Together, the measures used in this study cover six domains of wellbeing: School, Emotional control, Social support, Psychological distress, Psychological wellbeing, and Psychological needs. While there is some overlap between these domains, they have been grouped in this way for reporting purposes. Similarly, while only the CTC scales are explicitly referred to as risk and protective factors in the literature, all measures were classed as either a risk or a protective factor for consistency. This categorisation is based on whether they measure negative or positive constructs. The survey instrument (see Appendix H) also contained a scale called the Healthy Uses of Music (HUMS), which was being piloted by the Music Matters team (publications forthcoming). The HUMS was not used for this project, so is excluded from all discussion of methods and results. Summary details of each individual measure included in this project are presented in Table 6.1. More detailed information for measures, including scoring protocol, is provided in Appendix I.

87

Table 6.1 Scales Details for Keys to Success Study Domain

Scale

Factor

Item wording example

Response option example

Scale score range

Community

Low neighbourhood attachment

Risk

I like my neighbourhood

‘YES!' (1) to 'NO!' (4)

1-4

Community

Opportunities for prosocial involvement

Protective

Which of the following activities for people your age are available in your community? 'Sports teams'

‘Yes' (1) to 'No' (2)

1-4

Community

Rewards for prosocial involvement

Protective

There are people in my neighbourhood who encourage me to ‘YES!' (1) to 'NO!' (4) do my best

1-4

School

Academic failure

Risk

Are your school marks better than the ‘YES!' (1) to 'NO!' (4) marks of most students in your class?

1-4

School

Low commitment to school

Risk

How interesting are most of your school subjects to you?

‘Very interesting' (1) to 'Very boring' (5)

1-5

School

Opportunities for prosocial involvement

Protective

Teachers ask me to work on special classroom projects

‘YES!' (1) to 'NO!' (4)

1-4

School

Rewards for prosocial involvement

Protective

The school lets my parents know when I have done something well

‘YES!' (1) to 'NO!' (4)

1-4

‘Ignore her' (1) to 'Act like it’s a joke and ask her to put it back' (4)

1-4

‘YES!' (1) to 'NO!' (4)

1-4

I know how to relax when I feel tense ‘YES!' (1) to 'NO!' (4)

1-4

Peer and individual

Social skills

Protective

You’re looking at CD’s in a music store with a friend. You (…) see her slip a CD under her coat (…) What do you do now?

Peer and individual

Belief in the moral order

Protective

I think it is okay to take something without asking if you can get away with it

Emotional control

Emotional control

Protective

Social support

MSPSS-SO

Protective

There is a special person who is around when I am in need

‘Very strongly disagree' (1) to 'Very strongly agree' (7)

1-7

Social support

Trusted adult in life

Protective

I have an adult(s) that I trust and would turn to for advice if I was having problems

‘Very strongly disagree' (1) to 'Very strongly agree' (7)

1-7

Psychological distress

K10

Risk

Did you feel worthless?

‘None of the time' (1) to 'All of the time' (5)

10-50

Psychological wellbeing

MHI-PWB

Protective

Were you a happy person

‘All of the time' (1) to 'None of the time' (6)

14-84

Psychological needs

BPN-SF (whole score)

Protective

I feel like I am free to decide for myself how to live my life

‘Not at all true' (1) to 'Very true' (7)

9-63

Psychological needs

BPN-SF (autonomy subscale)

Protective

I feel like I am free to decide how I live my life

‘Not at all true' (1) to 'Very true' (7)

3-21

Psychological needs

BPN-SF (competence subscale)

Protective

Most days I feel a sense of accomplishment from what I do

‘Not at all true' (1) to 'Very true' (7)

3-21

Psychological needs

BPN-SF (relatedness subscale)

Protective

People in my life care about me

‘Not at all true' (1) to 'Very true' (7)

3-21

88

Procedure. Surveys were administered to intervention and control groups on the same day, before and after the music program. They were completed individually, using pen and paper, during class time, and under the supervision of class teachers and a researcher. The purpose of surveys, what they aimed to measure, and the confidential nature of answers were explained briefly to students before each survey session. Students were encouraged to seek clarification for unfamiliar terms or concepts used in the survey. Confidential support was available to students in the event that questions made them feel uncomfortable. Responses were checked, and data cleaned by hand using pen and paper. Responses were then entered manually into the online survey software program, SurveyMonkey (www.surveymonkey.com). Data files were then exported from SurveyMonkey in a format suitable for scoring and analysis using SPSS version 20.0 (IBM Corporation, 2011). Analysis. After calculating individual scale scores, paired sample t-tests were used to create an estimate of the mean change between baseline and follow-up scores. This was done for all scales in both intervention and control groups, and involved subtracting follow-up scores from baseline scores. These t-tests also provided confidence intervals for the mean change scores, and p values for evaluating their statistical significance. Only those participants who completed both baseline and follow-up surveys were included in this analysis. Independent sample t-tests were also used to estimate the mean difference between intervention and control group change scores for each scale, and to calculate confidence intervals and p values related to these mean differences. Results reported from this procedure did not use the equal variances assumption, as visual displays of the data suggested it was not appropriate for some scales. As such, equal variances were not assumed for all scales in order to maintain a consistent procedure. Mean difference scores were obtained by subtracting the control group’s mean change scores from the intervention group’s mean change scores. As such, for protective factor measures a positive mean difference score indicated the intervention group faired better than the control. For risk factors, negative mean difference scores indicated this.

89

The purpose of the independent sample t-tests was to determine if any mean change scores reported by the intervention group differed in magnitude from those reported by in the control group. However, due to elements of the study design such as the non-random sampling procedure and small number of participants, the ability to rule out potential confounders related to groups was limited. Nonetheless, visual displays of the change score data showed the distribution of change scores between the two groups was relatively similar. This suggested a normal distribution similar to samples from normal populations. Therefore the analysis of mean difference scores was deemed reasonable in this case. Along with t-test results, mean baseline scores for each scale were also observed to determine where each group was in relation to potential scale ranges at baseline. This was done to help interpret mean change and difference scores by identifying trends, such as possible ceiling or floor effects, which may have limited the amount of potential movement on a given scale. No data imputation procedures were used; meaning participants who missed necessary items on a given scale were not given a score for that scale. Mean change scores for each scale were calculated using only full data sets available for that scale. A final analysis was undertaken to identify which participants from the intervention group would take part in the subsequent interviews. This was achieved by identifying the 10 students with the most “intense” (largest) individual change scores on the nine CTC scales. All statistical data analyses were performed using SPSS version 20.0 (IBM Corporation, 2011). Graphical representation of results were created using Minitab for Windows 16th Edition (VSN International, 2008). Qualitative method. Consistent with sequential research designs described in MMR literature (Creswell & Plano Clark, 2011; Teddlie & Tashakkori, 2006), interviews were specifically designed to explore questions that arose from the analysis of survey data. Predominately these related to whether the survey provided a legitimate way to capture reports of psychosocial wellbeing. Therefore, interviews aimed to further investigate the perceived impact of the program on student psychosocial wellbeing, thus providing data to crosscheck benefits reported in the survey. Interviews also explored participant experiences of the survey to determine if they perceived it as an appropriate reporting 90

method. Lastly, interviews aimed to assess whether the musical elements of the program were seen to make a difference when compared to the non-musical version of the KTS program. This was to both evaluate the music program generally, and explore if any specific program elements were perceived as being connected to the reporting of psychosocial wellbeing benefits. While the decision to facilitate these interviews was driven by a desire to understand and interrogate the survey process and subsequent results, at the time they were undertaken they were still considered to be part of the empirical study. It was proposed that results would provide empirical (albeit it qualitative) evidence of program benefits, while offering an explanation of why these benefits were not evident in survey data. This was seen as a way to maintain the goal of providing evidence for policymakers, while also offering a critique of using quantitative methods in this area. Additionally it was proposed this could become a pilot study and the results would inform the choice of a better survey tool for future studies aiming to inform policy. Importantly, at this stage the decision had not yet been made to undertake the critical reflection analyses. However, it does reflect the shift towards critical thinking that was developing throughout the project, thus representing the emergent nature of the PhD. Participants. The 10 students from the intervention class that reported the most intense survey scores were invited to take part in interviews. Of these, one student was absent on the day of interviewing, thus nine students were interviewed. The intervention group’s class teacher was also interviewed in order to provide an alternate perspective. This teacher was provided with an information statement (see Appendix J), and signed a self-consent form (see Appendix K) before participating. Materials. Semi-structured interview schedules were chosen to conduct interviews with both the students and the teacher. While questions varied slightly between the teacher and student versions, a consistent structure was maintained throughout all interviews. The questions used for interview schedules, and the overall structure used for both, are presented in Table 6.2. The schedule for students began with open-ended questions such as “Can you tell me about the music activities that you took part in?” These questions aimed to invite

91

open responses from students about their general experiences of the program. These were followed by more directed questions such as “Have you changed since participating in the music program?” Such questions aimed to gather more specific feedback regarding the impact of the program. Further questions focused on particular constructs measured by the survey. For example, the question “Did the music program change how you felt about your connection to the school?” aimed to assess the impact of the program on students’ school engagement. Data from these direct questions was also crosschecked with participants’ survey responses to explore whether scale data accurately reflected the impact of the music program experienced by students. Direct questions regarding student experiences of completing surveys were also included to explore whether students were engaged in surveys, or understood their content. The teacher interview schedule asked questions intended to explore their perception of the program outcomes, the survey tool used, and the program in general. This aimed to provide an alternate source of data, or perspective, to afford wider insight into the impact of the music program, and the suitability of the survey. Teacher responses were also used to cross-examine the benefits of the program reported by students in both surveys and interviews (and vice versa). All interviews ended with an open invitation for participants to respond further to any topic covered in the survey, or comment on anything to do with the music program or research process. This intended to provide participants with the opportunity to articulate, in their own words, their perceptions of the program’s impact and the survey process. This approach was driven by the project’s value of collaborative enquiry, which aligns with both participatory and critical theory research approaches (Heron & Reason, 1997; Lincoln et al., 2011).

92

Table 6.2 Keys to Success Study Semi-structured Interview Schedules Focus

Student Interview Questions

Teacher Interview

Warm up/invitation for open responses regarding program in general

1.

Can you tell me about the music activities that took part in with [the MT] as part of the Keys to Success (KTS) program last year?

1.

Can you tell me about the music activities that your class took part in with [the MT] as part of the Keys to Success (KTS) program last year?

2.

How did you find it?

2.

How did you find it?

Perceptions of the musical element of the program

3.

What was the KTS program about?

3.

What was the KTS program about?

4.

How did the music fit in with that? Did it help? Was it useful?

4.

How did the music fit in with that? Did it help? Was it useful?

5.

Was there any difference between the music part of the KTS program and the other non-musical parts?

Psychosocial outcomes

5.

Do you think you changed as a result of the music program? How?

6.

Do you observe any changes in the students as a result of the music program? What were they?

6.

Did you learn anything? About life/friends? Different from non-musical KTS?

7.

Did they learn anything? About life/friends? Different from nonmusical KTS?

7.

Thinking just about the music part of the KTS program, how did it make you feel about being at school?

8.

Thinking just about the music part of the KTS program, were there any social benefits for the students? (engage with class more/foster friendships, networks/engage with school more/engage with subject matter more/able to express themselves)

8.

How did it make you feel about your place in community/society more generally?

9.

Was there anyone that stood out? How and why?

9.

How did it make you feel about being in class? With your classmates?

10. Did the music program change how you felt about your place/connection in/with the school? 11. Did it change how you felt about your teacher?

10. Did it change your relationship with any of the students? Or your view of them?

12. Did it change how you felt about your classmates? Perception of the use of music in class context

13. Would you like to do more music things in class/at school like this? Why?

11. Would you be open to doing more music things like this in class?

Perception of the survey tool used

14. Do you remember the survey that we did in class before and after the [music] sessions? What was that like?

12. Do you remember the survey that we did in class before and after the music sessions? What was that like?

15. Did the questions make sense?

13. Were the questions appropriate?

16. Did you feel like answering truthfully?

14. Did you feel like the kids answer these things truthfully?

Procedure. All interviews took place in the months following the program and were conducted one-on-one in meeting rooms at the school. Participants were informed their

93

responses would be confidential. With permission from participants, interviews were recorded using digital voice recorders, and transcribed using a word processor. Analysis. Interview data was analysed using a combination of deductive and inductive procedures. Fereday and Muir-Cochrane (2008) suggest this “hybrid approach” facilitates a method of qualitative analysis that addresses a priori research intentions while also allowing themes to emerge in their own right. This was considered an appropriate approach here as, while interview data was collected to satisfy predetermined research goals, the project’s commitment to collaborative research could also be accommodated. To elaborate, the use of deductive techniques allowed data to be organised and analysed in a way that addressed the pre-established project aim and research questions. 6 The complementary use of inductive techniques also enabled all issues or themes raised by participants to be identified and honoured, regardless of whether they had previously been identified as important to the study, or whether they existed in the literature. Consistent with participatory approaches (Heron & Reason, 1997; Lincoln et al., 2011), this approach aimed to give participants a voice by providing them opportunities to contribute to knowledge generation using their own language and ideas (however, it was acknowledged that it would be the interpretations of these ideas that occurred during analysis that would constitute the results – see “Inquiry posture”, Chapter 5). This hybrid analysis approach followed a three-step process. The first is presented in detail here, while the second and third steps are only discussed briefly in this chapter. Given their close links to the data, they are explained more thoroughly when presenting the results in Chapter 7. Step 1: creating a category template. The first step in analysing interviews was deductive. Using a process similar to that described by Crabtree and Miller (1999), this involved the creation of a “category template” to establish four predetermined data categories prior to analysis. These categories, which related directly to the established research goals included:

6

This was true for all iterations or stages of analysis (stages depicted in Figure

5.1, Chapter 5). 94

“Psychosocial wellbeing outcomes”, “Perceptions of the program”, “Perceptions of the survey” and “Emergent themes”. The “Psychosocial wellbeing outcomes” category provided a loose framework for grouping all data regarding the perceived impact of the program on student psychosocial wellbeing. This offered a point of comparison to the psychosocial wellbeing outcomes reported in the survey data, and thus offered an opportunity to interrogate the validity of quantitative results. It also aimed to identify the presence of any perceived benefits not covered in the survey. The second category, “Perceptions of the program”, was used to group participant responses regarding experiences and perceptions of the program. This included any responses that suggested or depicted a link between particular program elements and psychosocial wellbeing outcomes. The “Perceptions of the survey” category was designed to collect data regarding experiences and impressions of the survey itself. This included both perceptions of scale content and the survey process. Data gathered in the “Emergent themes” category included any responses that could not be categorised into previous categories, yet were considered relevant to the project. Relevance was determined by the perceived ability of codes to offer insight into: the impact of the program on psychosocial wellbeing; the strengths and limitations of delivering music programs in mainstream schools; how different program elements related to reported psychosocial wellbeing; and the strengths and limitations of research approaches used. It was through the use of this category that the study endeavoured to honour both the inductive coding process, and it’s participatory commitments. Step 2: coding interviews. After establishing the four categories, transcripts were read in order to identify “key statements” (McFerran & Grocke, 2007), or “important moments” (Fereday & Muir-Cochrane, 2008). When identified, these statements or moments were given “codes” which related to the nature of their content (i.e. a participant comment that the program “was fun”, was coded as “program was fun”). This was an inductive analysis procedure, as “codes” were allowed to emerge from the data on their own, regardless of their relationship to a category. This meant, for example, that any psychosocial wellbeing benefits reported by participants that were not covered by questions in the interview schedule, survey scales, or discussed in the literature, could be coded as 95

benefits, included in further analysis, and presented as results. It also meant any moments or interactions perceived as significant during the analysis process, but not identified as important before analysis started, could also be reported in the study. After all interviews had been coded, coded data was allocated to “Psychosocial wellbeing outcomes”, “Perceptions of the program”, or “Perceptions of the survey” categories based on their perceived relevance to these categories. Any codes that did not fit into these first three categories were allocated to the “Emergent themes” category. Step 3: creating themes. When all codes had been allocated to a category, they were grouped together to form “themes”. Themes were used to represent common or important ideas, perspectives, or concepts that arose from the codes in a given category (see Table 7.3, Chapter 7). Sometimes these were formulated from similar responses given by different participants, and other times they were based on a response from only one participant. This was also an inductive process, as themes were derived from the data rather than predetermined in relation to the literature, interview schedule, or survey. Themes formulated in this last step represented the results of the analysis. For example, any themes that emerged in the “Psychosocial wellbeing outcomes” category were considered to be program outcomes reported by participants during interviews. Observational reports. Observations were also recorded throughout the study, and are reported in the results section (see Chapter 7). While these observations were not included in the original KTS study design, nor used in the first stage of analysis (see Figure 5.1, Chapter 5), they were collected throughout this study and provide an important source of information for the critical reflection analyses (second analysis stage, see Figure 5.1). While observational reports in research are primarily associated with participant or direct observation methods (see Guest, Namey, & Mitchell, 2012), the observational approach used in this study more closely aligns with the concept of reflexivity. Finlay (2002b) outlines how reflexivity can be used in three main stages of research: the “Preresearch Stage”, “Data Collection Stage” and the “Data Analysis Stage”. While reflexivity is undertaken in all these stages throughout this project, it is the “Data Collection Stage” that incorporates the use of reflexivity as a data collection method (in regards to this study and the PhD project as a whole). Finlay states reflexivity during

96

this stage involves undertaking constant reflection on the research process throughout a study, and recording these reflections by taking field notes after research encounters. Such notes are written up alongside research methods and findings, and can provide valuable insight into how data or results are affected by data collection activities. For Finlay (2002b), reflections refer to the impact of the researcher-participant relationship on data collection – a concept she terms “intersubjective reflexivity” (Finlay, 2002a) – and the relationship between the researcher and the data itself. While both are reported for this study, a wider approach is taken to include observations and reflections regarding the influence of context, setting, research methods (including the use of different tools), and the nature of participants. This is similar to Stige, Malterud, and Midtgarden’s notion of “engagement”, which “refers to the researcher’s continuous interaction with and relationship to the phenomenon or situation studied” (2009, p. 1508). They too advocate using reflexivity as a form of data collection by recording and presenting reflections on how “engagement” impacts knowledge generation. They go on to claim such reflections are useful for addressing critical theory and postmodern (or participatory) research goals, such as critiquing the validity of a given research approach within certain socio-political contexts. Observational reports for this study were collected during the “Data Collection Stage”. They related to perceptions of research processes, and covered both Finlay’s notion of “intersubjective reflexivity” and Stige et al.’s concept of “engagement”. Reports were collected from: notes taken during and after data collection sessions; notes and voice recordings made during supervision and research debrief sessions; documentation of discussions with school staff members; and the documentation of personal reflections on research activities. Reflections and observations regarding the context of data collection activities, including the nature of participant groups, participant behaviour, and staff comments are compiled and presented in narrative form in a separate section within the results chapter for the KTS study (see Observational Reports, Chapter 7). Observations and reflections regarding intersubjective researcher-participant relationships and interactions are inextricably linked to the data itself, and are therefore interwoven within the presentation of qualitative results to avoid repetition. All observational reports are also used as a tool of critique and deconstruction in the critical reflection analyses (see Chapters 9 and 10).

97

Empirical Study 2: Bereavement Group Design overview. The Bereavement Group (BG) study was originally designed to evaluate the impact of a music therapy program on the psychosocial wellbeing of six bereaved students in a mainstream school context. It also aimed to evaluate the delivery of the program in this setting, including barriers to student participation, and the suitability of program structure and content. To use MMR terminology, data collection activities followed a “concurrent mixed design” (Teddlie & Tashakkori, 2006). This means the decision to conduct both surveys and interviews was made at the design stage, and were carried out in parallel. The survey tool used was similar to that used in the KTS study, and was administered before and after the delivery of the music program to detect change on constructs of student psychosocial wellbeing. Interviews were undertaken within the same week as the follow-up survey, meaning the interview schedule was not informed by survey results. Due to the nature of the music therapy program, a small sample of five participants was recruited for the study, each of which participated in both questionaries and interviews. The music program. The music program investigated in this study was a bereavement program specifically designed and delivered to help students deal with issues of grief and loss in a school setting. The school’s wellbeing coordinator had previously offered non-musical bereavement programs to students suffering from grief or loss. In this case, the same music therapist from the KTS study was invited to run a music therapy-based version of the existing bereavement program, again as part of the Music Matters project. Following existing school protocol, the wellbeing coordinator recruited students who were currently dealing with grief or loss for the program. Weekly Sessions were run over five weeks in a semi-private group setting and during class time. Each session lasted approximately 50 minutes. Times alternated weekly to ensure students missed different classes for each session. This BG music program differed in some important ways from the music program in KTS study. It had a specific wellbeing focus and was not related to curriculum content. Also, sessions were run privately, out of class, and with a small, targeted group. Further, the music therapist, who had significant experience in 98

designing and facilitating music therapy-based bereavement programs, designed the content and structure of the program. Also, the music therapist was able to facilitate sessions in keeping with her music therapy training and experience. Therefore the program was specifically focused towards promoting psychosocial wellbeing outcomes and followed a music therapy model. This meant the program was tailored to the needs of students, and facilitated in a way that fostered democratic student participation and input into the program. While there was a pre-determined schedule for sessions, the program also followed a reflexive model; through ongoing participant consultation, students’ needs were re-assessed throughout the duration of the program to ensure activities met the ongoing or emerging requirements of all involved. After an initial introduction session the program was divided into four main sessions corresponding to the seasons of the year, each of which addressed and facilitated different stages of the grieving process. During sessions, students were invited to share their stories of grief and loss, choose a song that they felt represented their feelings of bereavement, and engage in free improvisation and emotional expression through drumming. During final sessions, the group worked on composing lyrics to their own song, which they sang together at the end of the program. This was recorded by the music therapist and distributed to each group member (see program protocol in Appendix L for more detail) Quantitative method. Participants. The school wellbeing coordinator recruited participants. Each were known to be experiencing some form of grief or loss in their personal life, and therefore invited to take part in the program. While participants were recruited from the same school as the KTS study, and there was a level of continuity between samples, the BG group differed in important ways. All were identified as at-risk given they were dealing with experiences of loss and grief. Also, the sample was smaller than that in the KTS study: six students participated in the program in total, and only five of these were present on the days of data collection, and participated in the study. Three of these five participants were female (60%), while three (60%) were from Anglo-Celtic backgrounds, and two (40%) were from Arabic backgrounds.

99

Another lesser distinction between the BG and KTS samples was age. Of the five students, one (20%) was in Year 8 (upper age limit of 16 years; see www.study.vic.gov.au), and four (80%) were in Year 9 (upper age limit of 17 years; see www.study.vic.gov.au). Information statements were provided to students (see Appendix C) and their parents (see Appendix D), and participants were asked to return active consent forms (see Appendix E) signed by their parents before participation. Information statements informed parents and students that choosing not to participate in the study would not affect students’ ability to partake in the music program. Ethics approval for this study was received from the UoM HREC (see Appendix F) and the DEECD (see Appendix G). Materials. An adapted version of the Victorian Adolescent Health and Wellbeing Survey (HowRU; DEECD, 2010) was again used to collect survey data. While the version of the HowRU used in this study was largely similar to that used in the KTS study, there were some differences in the scales and domains included. An extra measure was included in the School domain, termed here as Educational expectation, which was taken from the HowRU survey, and categorised as a protective factor. Also, the Low commitment to school scale contained two less items. Further, the CTC scales used in the first study for the Community and Peer and individual domains, were not included in this study. This meant the domains investigated by the survey were reduced to six: School, Emotional control, Social support, Psychological distress, Psychological wellbeing, and Psychological needs. Summary details of all measures used are presented in Table 6.3, and scoring details for the Educational expectation and Low commitment to school are provided in Appendix M. All other measures were scored using procedures described in the first study (see Appendix I). The use of risk and protective factors described in the KTS study were also used to analyse and report BG data. For a full copy of the survey, see Appendix N.

100

Table 6.3 Scales Details for Bereavement Group Study Domain

Measure

Factor

Item wording example

Response option example

Range

School

Academic failure

Risk

Are your school marks better than the marks of most students in your class?

‘YES!' (1) to 'NO!' (4)

1-4

School

Low commitment to school

Risk

How interesting are most of your school subjects to you?

‘Very interesting' (1) to 'Very boring' (5)

1-5

School

Opportunities for prosocial involvement

Protective

Teachers ask me to work on special classroom projects

‘YES!' (1) to 'NO!' (4)

1-4

School

Rewards for prosocial involvement

Protective

The school lets my parents know when I have done something well

‘YES!' (1) to 'NO!' (4)

1-4

School

Educational expectation

Protective

Emotional control

Emotional control

Protective

I know how to relax when I feel tense

‘YES!' (1) to 'NO!' (4)

1-4

Social support

MSPSS-SO

Protective

There is a special person who is around when I am in need

‘Very strongly disagree' (1) to 'Very strongly agree' (7)

1-7

Social support

Trusted adult in life

Protective

I have an adult(s) that I trust and would turn to for advice if I was having problems

‘Very strongly disagree' (1) to 'Very strongly agree' (7)

1-7

Psychological distress

K10

Risk

Did you feel worthless?

‘None of the time' (1) to 'All of 10-50 the time' (5)

Psychological wellbeing

MHI-PWB

Protective

Were you a happy person

‘All of the time' (1) to 'None of 14-84 the time' (6)

Psychological needs

BPN-SF (whole score)

Protective

I feel like I am free to decide for myself how to live my life

‘Not at all true' (1) to 'Very true' (7)

9-63

Psychological needs

BPN-SF (autonomy subscale)

Protective

I feel like I am free to decide how I live my life

‘Not at all true' (1) to 'Very true' (7)

3-21

Psychological needs

BPN-SF (competence subscale)

Protective

Most days I feel a sense of accomplishment from what I do

‘Not at all true' (1) to 'Very true' (7)

3-21

Psychological needs

BPN-SF (relatedness subscale)

Protective

People in my life care about me

‘Not at all true' (1) to 'Very true' (7)

3-21

What is the highest level of Year 10' (1) to 'University' (6) education you would like to get

1-6

Procedure. Participants completed baseline surveys in the first music program session, before activities started. Follow-up surveys were completed in the last session upon conclusion of program activities. The purpose of surveys was briefly explained to students, who were encouraged to seek clarification of terms and language. Students were also informed that their responses would remain confidential. Surveys were filled out individually using pen and paper under the supervision of the music therapist.

101

Data was checked and cleaned manually using pen and paper before being entered into SurveyMonkey. Data was exported from SurveyMonkey in CSV format and analysed using SPSS version 20.0 (IBM Corporation, 2011). Analysis. An estimate of mean change scores for each scale was calculated using paired samples t-tests. These t-tests also provided confidence intervals and p values for assessing the statistical significance of the mean change scores. Scores indicate the mean difference between baseline and follow-up responses, obtained by subtracting the follow-up scores from the baseline scores. No data imputation procedure was used and therefore, if one participant failed to respond to sufficient items on a given scale, the change score for that scale was based on the remaining full sets of data only. The group’s mean baseline scores were also calculated, and have been reported along with potential scale ranges to aid in the interpretation of mean change scores. This was done to highlight potential ceiling or floor effects that may have affected the ability for movement from baseline to follow-up on a given scale. SPSS version 20.0 (IBM Corporation, 2011) was used for all statistical data analyses. Minitab for Windows 16th Edition (VSN International, 2008) was used for creating a graphical representation of results. Qualitative method. Qualitative data for the BG study was originally collected for a different purpose than which it is used in this project. Unlike the sequential design of KTS study, interviews occurred concurrently with surveys; meaning both were designed and carried out in parallel. Therefore, interviews were not informed by survey results and were not designed to interrogate quantitative data nor evaluate the survey. Rather, they aimed to evaluate the program in a more general sense. This included exploring motivations for participation, commitment to the program, barriers to participation, and suggestions regarding the content and structure of future programs. Interviews also provided an alternate way for students to report program benefits, and identify outcomes not targeted by the survey. To reiterate, however, the reporting of benefits in these interviews intended to supplement or support reports in the survey data, not critique them. Even though initially collected for a different purpose, this data does suit the needs of this project. It provides qualitative accounts of program benefits that can be

102

compared to those reported in survey data. Also, interview responses provided insight into the attributes of the music program perceived as productive and counterproductive for facilitating benefits. Further, interviews offered students opportunities to actively participate in knowledge creation, thus maintaining collaborative research goals. Participants. All five participants who filled out surveys were invited to participate in an interview. All five agreed. Materials. Interviews were conducted using a semi-structured interview schedule. This began with open-ended questions such as “What music activities have you been involved in at school this year?” and “Can you describe why you chose to participate in those projects before you began?” The purpose of such questions was to enable participants to give general responses regarding their participation. These were followed by more directed questions regarding barriers to participation such as commitment to the program and issues of participating during class time. Such questions represent the duel purpose of the interviews; they also served as an evaluation of the barriers to running such programs in mainstream school contexts. The interview then focused on experiences of participation which related to the musical elements of the program, including questions such as “Has your relationship with music changed since you became involved in the program?” Questions also focused on how different stages and elements of the program affected participants’ ability to deal with issues of grief and loss, and how they felt they had changed since participating in the program. This aimed to assess participant perceptions regarding the efficacy of the program, while also allowing them the opportunity to articulate its perceived impact on their psychosocial wellbeing using their own language. Further questions asked participants to identify elements of the program that worked for them, and those they felt could be improved. Again, these questions aimed to both evaluate elements of the program beneficial for psychosocial wellbeing, and explore any further participation benefits experienced by participants. The interview finished with an invitation for participants to respond to any element of the program they felt was not covered, or that they wanted to expand on. In this question, participants were prompted to relate both positive and negative

103

perceptions of the program; an approach taken throughout the interviews in which students were encouraged to respond freely. This was done by reminding participants that their opinions were important to us; that negative feedback was just as important as positive feedback, and that responses were confidential. This approach represents the study’s commitment to a participatory research approach, in which participants were motivated to engage in the interview process as valued co-creators of knowledge (Heron & Reason, 1997; Lincoln et al., 2011). A full list of the questions used in the schedule is presented in Table 6.4. This also includes an overview of how questions were grouped to address research goals. Table 6.4 Bereavement Group Study Semi-structured Interview Schedule Focus

Interview questions

Warm up/invitation of open responses to the program in general

1.

Can you tell me a little about the music activities you have been involved in at the school this year?

2.

Can you describe why you chose to participate in those projects before you began?

Potential barriers to participation

3.

Do you/did you feel committed to participating in the program?

4.

Was it hard to find the time to participate? Did you have to make any sacrifices to join in?

Psychosocial outcomes (and perception of program structure)

5.

Has your relationship with music changed since you became involved in the program? How? (nature and quality)

6.

Your program was broken up into four parts, each which was related to a different season of the year, and had a different focus, is that right? Can I ask you a little about how you found each session? 6.1. WEEK 1: Autumn – sharing stories of grief and loss with the group 6.2. WEEK 2: Winter – sharing a song with the group that expresses your grief and loss 6.3. WEEK 3: Spring – Instrumental improvisation-playing emotions on drums 6.4. WEEK 4 & 5: Summer – Creating a song together to celebrate passed love ones

Feedback regarding program attributes

7.

Have you changed since you participated in the program?

8.

Were other people you knew involved in the program?

9.

Having participated, what would you say has been the best bit?

10. And what would you say has been the most challenging? 11. Do you have any other ideas about what you would like to do in a future program like this?

Procedure. Interviews took place over the two weeks following the end of the program, and were conducted one-on-one in meeting rooms at the school. All participants gave verbal

104

consent for interviews to be recorded using a digital device. Recordings were transcribed using a word processor. Analysis. The same three-step “hybrid approach” to analysis used for the KTS interview data was also used here. In this case, however, the “Perceptions of the survey” category was not used. This was because interviews in this study did not aim to interrogate the survey process and content in the same way. The remaining categories are consistent between the two studies. This continuity was maintained to enable comparability of interview responses and themes between studies. Observational reports. In keeping with notions of reflexive reporting described by Stige et al. (2009) and Finlay (2002a, 2002b), observational reports were also collected for the BG study, and are reported on in this project. These included the documentation of supervision sessions, research meetings, and personal observations of the research process. These are presented in their own section in the BG study results chapter (see Observational Reports, Chapter 8). Reflections connected to the analysis of interview data are integrated into the reporting of these results.

105

Chapter 7: Keys to Success Results This chapter reports the primary outcomes of a school-based study called Keys to Success (KTS). The KTS study explored the psychosocial wellbeing benefits of a class-based music program delivered in a mainstream school. These benefits were investigated using both quantitative and qualitative methods. Details of the music program and research methods used are described in detail in Chapter 6. The primary outcomes of this study are not reported here to provide empirical evidence for the potential of school-based music programs to increase student psychosocial wellbeing. Rather, they are reported to present data, results and contextual information used to inform two critical reflection analyses. These critical reflection analyses interrogate the legitimacy and suitability of the research approach used in the KTS study; including the methods of investigation, and the program investigated. Along with the primary psychosocial outcomes, further evidence used to inform the critical reflection analyses is also presented here. This includes participant perceptions of research activities and the music program reported during interviews; observations and reflections on data collection recorded by researchers throughout the study; comments from school staff; and further contextual information relevant to the study. The results from the quantitative analysis are reported first, followed by the results of the qualitative analyses. The chapter concludes with the presentation of observational data. All results are presented independently; discussion of the similarities and discrepancies between quantitative and qualitative data sets, and how these are related to observational data, is reserved for subsequent chapters. Each set of data is presented in a style consistent with the epistemological nature of the research method involved. For example, quantitative results are reported objectively, with little in-depth interpretation or reflection. Qualitative findings are presented using the first-person to indicate the presence of the researcher’s own reflexive interpretations in the analysis process. Observational data is reported in the first-person narrative to reflect the process of observation, reflection, and interpretation involved. This is done to give contextual information, and communicate what occurred during research activities from the point of view of the researchers and school staff.

106

Methodological Overview Quantitative data was collected from an intervention group and a (non-musical) control group using baseline and follow-up surveys. Mean change scores were calculated, along with confidence intervals and p values, to determine whether the program had a statistically significant impact on the intervention group’s psychosocial wellbeing. Mean difference scores, confidence intervals and p values were also calculated to test wether there were any statistically significant differences between mean change scores from the intervention group and the control group. See Chapter 6 for full description of quantitative method used. Interviews were conducted after an initial analysis of the quantitative data with selected students and the teacher from the intervention group. Interviews initially aimed to further explore psychosocial outcomes reported in the qualitative data, while also exploring issues related to the program and the survey process. A combined deductiveinductive approach was used for analysis. Codes were inductively identified and allocated to four predetermined categories of interest. Once allocated, themes were created from codes within each category. See Chapter 6 for full description of qualitative method used. Quantitative Results Participants were asked to complete 18 scales which covered risk and protective factors in eight different domains of student psychosocial wellbeing (see Table 7.1 and Table 7.2). Of the 20 participants in the intervention group, 18 responded at baseline, 15 at follow-up, and 14 completed both baseline and follow-up questionnaires. Of the 22 participants in the control group, 21 responded at baseline, 17 at follow-up, and 17 completed both questionnaires.

107

Table 7.1 Protective Factors: Mean Change and Difference of Mean Scores Mean change scores for protective factors Scale

Domain

Opportunities for Community prosocial involvement Rewards for prosocial Community involvement Opportunities for School prosocial involvement Rewards for prosocial School involvement Social skills

Peer and individual

Belief in the moral order

Peer and individual

Emotional control

Emotional control

MSPSS-SO

Social support

Trusted adult in life

Social support

MHI-PWB

Psychological wellbeing

BPN-SF (whole score)

Psychological needs

BPN-SF (autonomy)

Psychological needs

BPN-SF (competence)

Psychological needs

BPN-SF (relatedness)

Psychological needs

Group Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control Intervention Control

n

p

M

SD

95% CI

12 17 13 17 14 16 13 17 13 17 14 17 14 17 11 17 14 17 13 16 11 14 12 15 14 17 13 16

0.01 0.65 0.17 0.43 0.84 0.20 0.05 0.55 0.55 0.89 0.71 0.17 0.17 0.53 0.93 0.81 0.27 0.25 0.10 0.03 0.37 0.62 0.17 0.78 0.32 0.31 0.42 0.64

-0.50 0.13 -0.26 -0.18 0.01 0.17 -0.29 -0.07 0.07 -0.02 0.04 -0.22 -0.13 -0.10 -0.02 -0.09 -0.36 -0.53 -4.00 -6.56 -2.27 -1.93 -1.67 -0.40 -0.79 -1.06 -0.62 -0.69

0.58 1.19 0.63 0.89 0.27 0.50 0.48 0.50 0.42 0.58 0.35 0.64 0.32 0.67 0.85 1.48 1.15 1.84 8.07 10.60 8.01 14.19 3.89 5.53 2.86 4.19 2.66 5.71

-0.87, -0.13 -0.48, 0.75 -0.63, 0.12 -0.63, 0.28 -0.14, 0.17 -0.10, 0.44 -0.58, 0.00 -0.33, 0.18 -0.18, 0.32 -0.32, 0.28 -0.17, 0.24 -0.55, 0.11 -0.31, 0.06 -0.45, 0.24 -0.59, 0.55 -0.85, 0.67 -1.02, 0.31 -1.48, 0.42 -8.88, 0.88 -12.21, -0.91 -7.66, 3.11 -10.12, 6.26 -4.14, 0.81 -3.46, 2.66 -2.44, 0.87 -3.21, 1.10 -2.22, 0.99 -3.73, 2.36

Mean Difference 95% p M CI 0.07 -0.63 -1.32, 0.06 0.78

-0.08

-0.65, 0.49

0.30

-0.15

-0.45, 0.14

0.24

-0.21

-0.58, 0.15

0.63

0.09

-0.28, 0.47

0.17

0.26

-0.12, 0.63

0.91

-0.02

-0.40, 0.36

0.88

0.07

-0.84, 0.97

0.75

0.17

-0.94, 1.28

0.47

2.56

-4.56, 9.68

0.94

-0.34

-9.69, 9.00

0.49

-1.27

-5.01, 2.48

0.83

0.27

-2.33, 2.88

0.97

0.07

-3.26, 3.41

Note. CI = confidence interval. Mean change scores calculated using paired sample t-tests with baseline scores subtracted from follow-up scores. Mean difference scores calculated using independent sample ttests with control group scores subtracted from intervention group scores and equal variances not assumed. Blue p values indicate statistically significant findings. Green mean change scores indicate positive changes and red mean change scores indicate negative changes.

Table 7.2 Risk Factors: Mean Change and Difference of Mean Scores Mean change scores for risk factors Scale

Domain

Low Neighbourhood Attachment

Community

Academic failure

School

Low commitment to school

School

K10

Psychological distress

Group Intervention Control Intervention Control Intervention Control Intervention Control

n

p

M

SD

95% CI

14 16 13 14 14 17 13 17

0.14 0.86 0.94 0.77 0.22 0.65 0.06 0.59

-0.19 -0.04 0.01 0.04 -0.13 0.06 2.15 1.00

0.45 0.92 0.42 0.56 0.37 0.55 3.69 7.57

-0.45, 0.07 -0.53, 0.45 -0.24, 0.26 -0.28, 0.37 -0.34, 0.09 -0.22, 0.34 -0.08, 4.39 -2.89, 4.89

Mean Difference 95% p M CI 0.57 -0.15 -0.68, 0.39 0.86

-0.04

-0.43, 0.36

0.27

-0.19

-0.53, 0.15

0.59

1.15

-3.18, 5.49

Note. CI = confidence interval. Mean change scores calculated using paired sample t-tests with baseline scores subtracted from follow-up scores. Mean difference scores calculated using independent sample ttests with control group scores subtracted from intervention group scores and equal variances not assumed. Blue p values indicate statistically significant findings. Green mean change scores indicate positive changes and red mean change scores indicate negative changes.

108

Community domain.

Low neighbourhood attachment

Music No music

Opportunities for prosocial involvement

Music No music

Music

Rewards for prosocial involvment

No music Risk factor: red line Protective factor: blue line

1

2 3 Mean baseline scores (with 95% CI)

4

Figure 7.1 Mean scores at baseline and confidence intervals by group for Community domain On average, reported baseline scores placed groups in the positive half of scale ranges for both risk and protective factors in the community domain (see Figure 7.1). While marginal for most scales, reported mean scores for the intervention group show potential ceiling effects on the Opportunities for prosocial involvement scale, and potential floor effects for the Low neighbourhood attachment scale. This is especially the case for the former, where reported scores for the intervention group showed low variation across the scale range. The intervention group reported an improvement on the Low neighbourhood attachment scale, with a mean change score of -0.19, p = .14, 95% CI [-0.45, 0.07]. As a risk factor a decrease in this score indicates improvement, however, this was not statistically significant. The only statistically significant (and largest) mean change for the intervention group in this domain was for the Opportunities for prosocial involvement scale, with a mean change of -0.50, p = .01, 95% CI [-0.87, -0.13]. As a protective factor, this decrease suggests the group was worse at follow-up. This group also reported a negative outcome on the Rewards for prosocial involvement scale, yet this was not statistically significant (see Table 7.1).

109

Low neighbourhood attachment

Music No music

Opportunities for prosocial involvement

Music No music

Music

Rewards for prosocial involvment

No music Risk factor: red line Protective factor: blue line

-4

-3

-2 -1 0 1 2 Mean change scores (with 95% CI)

3

4

Figure 7.2 Mean change scores and confidence intervals by group for Community domain The control group reported improvement on the Low neighbourhood attachment and Opportunities for prosocial involvement scales, with mean change scores of -0.04, p = .86, 95% CI [-0.53, 0.45], and 0.13, p = .65, 95% CI [-0.48, 0.75] respectively. The control group’s mean change score for the Rewards for prosocial involvement scale in this domain showed a negative outcome. No mean change scores for the control group were statistically significant in this domain (see Table 7.1 and Table 7.2). As shown in Figure 7.2, reported mean change scores for both groups on all community domain scales were only marginally different from zero. Given the possible range of one to four, this indicates minimal change. The exception was for the intervention’s mean change score on the Opportunities for prosocial involvement scale, which moved, on average, 12.5% of the scale range. The mean difference in reported change scores for this domain indicate the intervention group showed greater improvement for the Low neighbourhood attachment scale, with a mean difference of -0.15, p = .57, 95% CI [-0.68, 0.39]. For both the Opportunities for prosocial involvement and Rewards for prosocial involvement scales, the control group improved more than the intervention with mean differences of -0.63, p = .07, 95% CI [-1.32, 0.06] and -0.08, p = .78, 95% CI [-0.65, 0.49] respectively. Mean differences were not statistically significant.

110

The only statistically significant reported change in this domain was the intervention group’s decrease on the Opportunities for prosocial involvement scale. While this suggests the scale was able to capture change in this area, the fact it showed a decrease from an already high mean baseline score potentially indicates this decrease was due to variation that could only go in one direction; this can arise from regression to the mean. That remaining reported mean change scores show no statistically significant variation from zero could suggest neither the control nor the intervention conditions had an impact on the remaining measures. Alternatively, results may reflect low power. School domain. As shown in Figure 7.3, reported baseline scores suggest both groups were fairing well in this domain before the program. All reported means were towards the high end of protective factors scales, at the lower end for risk factors, and show little variation.

A cademic failure

Music No music

O pportunities for prosocial inv olv ement

Rew ards for prosocial inv olv ement

Music No music Music No music 1

2

3

4

Music Low commitment to school

No music

Risk factor: red line Protectiv e factor: blue line

1

2

3 Mean baseline scores (w ith 95% C I)

4

Figure 7.3 Mean scores at baseline and confidence intervals by group for School domain 111

5

The intervention group reported no improvement on either the Academic failure or Rewards for prosocial involvement scales, with mean changes scores of 0.01, p = .94, 95% CI [-0.24, 0.26] and -0.29, p = .05, 95% CI [-0.55, 0.00] respectively. The intervention group did report improvement on the Low commitment to school and Opportunities for prosocial involvement scales, with mean change scores of -0.13, p = .22, 95% CI [-0.34, 0.09] and 0.01, p = .84, 95% CI [-0.14, 0.17]. None of these mean changes achieved statistical significance. The control group showed improvement only on the Opportunities for prosocial involvement scale, with a mean change score of 0.17, p = .20, 95% CI [-0.10, 0.44]. The control group reported no other improvements in this domain, and no scores were statistically significant (see Table 7.1 and Table 7.2).

A cademic failure

Music No music

O pportunities for prosocial inv olv ement

Rew ards for prosocial inv olv ement

Music No music Music No music -4

-3

-2

-1

0

1

2

3

4

Music Low commitment to school No music

Risk factor: red line Protectiv e factor: blue line

-5

-4

-3

-2

-1 0 1 2 Mean change scores (w ith 95% C I)

3

4

5

Figure 7.4 Mean change scores and confidence intervals by group for School domain Reported mean difference between groups in this domain show the intervention improved more on average than the control for both risk factors, with mean difference scores of -0.035, p = .86, 95% CI [-0.43, 0.36] and -0.19, p = .27, 95% CI [-0.53, 0.15]

112

for Academic failure and Low commitment to school respectively. For protective factors, the control improved more than the intervention for both the Opportunities for prosocial involvement and Rewards for prosocial involvement scales, with mean differences of -0.15, p = .30, 95% CI [-0.45, 0.14] and -0.21, p = .24, 95% CI [-0.58, 0.15] respectively. No mean difference scores were statistically significant. Reported mean change scores were minimal in relation to scale ranges (see Figure 7.4) and did not achieve statistical significance; suggesting scales did not capture any movement in this domain. This could be explained by low power, random response variation, or the idea the program did not impact constructs measured by these scales. Peer and individual domain.

Social skills

Music

No music

Beleif in moral order

Music

No music

1

3 2 Mean baseline scores (with 95% CI)

4

Figure 7.5 Mean scores at baseline and confidence intervals by group for Peer and individual domain Figure 7.5 shows mean reported baseline scores for both groups in this domain are towards the high end of scale ranges. Combined with small confidence intervals, this suggests both groups reported high subjective peer and individual wellbeing at baseline on these measures. This indicates ceiling effects for these scales.

113

Social skills

Music

No music

Beleif in moral order

Music

No music

-4

-3

-2 -1 0 1 2 Mean change scores (with 95% CI)

3

4

Figure 7.6 Mean change scores and confidence intervals by group for Peer and individual domain The intervention group reported a positive increase for both Social skills and Belief in the moral order scales, with mean changes of 0.07, p = .55, 95% CI [-0.18, 0.32] and 0.04, p = .71, 95% CI [-0.17, 0.24] respectively. The control group reported a decrease for both scales, with a mean change score of -0.02, p = .89, 95% CI [-0.32, 0.28] for Social skills, and -0.22, p = .17, 95% CI [-0.55, 0.11] for Belief in the moral order. Mean difference scores suggest the intervention group, on average, faired better than the control for both scales (see Table 7.1). Given the intervention group’s reported mean scores improved while those for the control group did not, results could be interpreted to indicate the music program positively impacted students in this domain. However, the fact changes were minimal (see Figure 7.6) and not statistically significant (see Table 7.1) suggests scales did not capture any real change. This may be due to the high baseline scores, or because the music program had no effect.

114

Emotional control domain.

Music

Emotional control

No music

1

2 3 Mean baseline scores (with 95% CI)

4

Figure 7.7 Mean scores at baseline and confidence intervals by group for Emotional control domain Figure 7.7 shows both groups reported high emotional control at baseline. The relatively small confidence intervals suggest a potential ceiling effect, especially for the intervention group.

Music

Emotional control

No music

-4

-3

-2 -1 0 1 2 Mean change scores (with 95% CI)

3

4

Figure 7.8 Mean change scores and confidence intervals by group for Emotional control domain

115

Both the intervention and control groups reported an average decrease at followup for Emotional control, with mean change scores of -0.13, p = .17, 95% CI [-0.31, 0.06] and -0.10, p = .53, 95% CI [-0.45, 0.24] respectively. As a protective factor, this indicates both groups reported lower emotional control at follow-up. However, as Figure 7.8 shows, changes were minimal. When comparing reported change scores, the control group’s score decreased less than the intervention, with a mean difference of -0.02, p = .91, 95% CI [-0.44, 0.36]. No scores for this domain were statistically significant (see Table 7.1). Combined with potential ceiling effects and minimal reported movement, this suggests no change was captured in this domain for either group. Social support domain.

MSPSS-SO

Music

No music

Trusted adult in life

Music

No music

1

2

3 4 5 Mean baseline scores (with 95% CI)

6

7

Figure 7.9 Mean scores at baseline and confidence intervals by group for Social support domain Average reported baseline scores for both groups show high levels of perceived social support (see Figure 7.9). Combined with confidence intervals of less than two points for each mean score, this suggests potential ceiling effects. Reported mean change scores for the intervention group were -0.02, p = .93, 95% CI [-0.59, 0.55] and -0.36, p = .27, 95% CI [-1.02, 0.31] for the MPSS-SO and Trusted adult in life scales respectively. Table 7.1 shows the control group reported similar changes for both scales. A comparison of change scores showed the intervention decreased less than the control on both MPSS-SO and Trusted adult in life scales, with 116

mean differences of 0.07, p = .88, 95% CI [-0.84, 0.97] and 0.17, p = .75, 95% CI [-0.94, 1.28] respectively. While both groups reported lower levels of social support at follow-up, changes were minimal (see Figure 7.10) and not statistically significant. This suggests these scales did not capture any significant change in social support.

MSPSS-SO

Music

No music

Trusted adult in life

Music

No music

-7

-5

-3 -1 0 1 3 Mean change scores (with 95% CI)

5

7

Figure 7.10 Mean change scores and confidence intervals by group for Social support domain Psychological distress domain. Mean reported scores place both groups at the lower end of the psychological distress spectrum at baseline (see Figure 7.11). This was particularly evident for the intervention group. The intervention group reported a mean change score of 2.15, p = .06, 95% CI [0.08, 4.39], while the control group reported a mean change of 1.00, p = .59, 95% CI [2.89, 4.89]. As a risk factor, this indicates reported psychological distress rose marginally for both groups during the study. However, given the high baseline scores, change scores may indicate a regression to the mean. Mean difference scores showed the intervention rose more (1.15, p = .59, 95% CI [-3.18, 5.49]), yet changes were minimal (see Figure 7.12). Minimal change and the lack of statistical significance suggest no significant change in psychological distress was captured using the K10 scale for these groups. 117

Music

K10

No music

10

15

20 25 30 35 40 Mean baseline scores (with 95% CI)

45

50

Figure 7.11 Mean scores at baseline and confidence intervals by group for Psychological distress domain

Music

K10

No music

-40

-30

-20 -10 0 10 20 Mean change scores (with 95% CI)

30

40

Figure 7.12 Mean change scores and confidence intervals by group for Psychological distress domain

118

Psychological wellbeing domain.

Music

MHI-PWB

No music

14

20

30 40 50 60 70 Mean baseline scores (with 95% CI)

80 84

Figure 7.13 Mean scores at baseline and confidence intervals by group for Psychological wellbeing domain Figure 7.13 shows reported mean scores at baseline were towards the high end of the scale, indicating high levels of reported psychological wellbeing before the program began. Paired with small confidence intervals, mean scores suggest potential ceiling effects for this measure – particularly for the intervention group. Both intervention and control groups reported decreased mean scores at followup, with mean changes on the MHI-PWB scale of -4.00, p = .10, 95% CI [-8.88, 0.88], and -6.56, p = .03, 95% CI [-12.21, -0.91] respectively. This suggests the psychological wellbeing of both groups decreased during the program. Yet, given the range of 14 to 84 for this scale (shown in Figure 7.14), decreases were minimal. The mean change score for the control group was statistically significant, while the intervention group’s was not. This difference in statistical significance suggests the control group’s psychological wellbeing decreased less than the intervention group from baseline to follow-up. This is supported by a mean difference score of -3.28, p = .57, 95% CI [14.88, 8.31], which shows the drop in reported psychological wellbeing was more pronounced for the control. This difference, however, was not statistically significant.

119

Music

MHI-PWB

No music

-70 -60

-40

-20 0 20 40 Mean change scores (with 95% CI)

60

70

Figure 7.14 Mean change scores and confidence intervals by group for Psychological wellbeing domain Results suggest this measure was able to capture reported change in psychological wellbeing for the control group, although this may again reflect a regression to the mean. As results were not statistically significant for the intervention group, a conclusion that the music program affected this domain cannot be made. Psychological needs domain. As shown in Figure 7.15, mean reported baseline scores for the intervention group were higher than the control for the BPN-SF scale and its subscales. Intervention group means were also high on the scale range, suggesting possible ceiling effects. Control group means are lower, yet are also in the upper half of the scale ranges. Both groups reported decreased scores for all scales. Mean change scores for the overall BPN-SF scale were reported as -2.27, p = .37, 95% CI [-7.66, 3.11] and 1.93, p = .62, 95% CI [-10.12, 6.26] for intervention and control groups respectively. Decreased mean changes of similar a magnitude were reported for all BPN-SF subscales (see Figure 7.16). No reported mean changes in this domain were statistically significant (see Table 7.1).

120

M usic

BP N-SF (w hole score) No music

10

BP N-SF (autonomy )

15

20

25

30

35

40

45

50

55

60

63

20

21

M usic No music

BP N-SF (competence)

M usic No music

BP N-SF (relatedness)

M usic No music 3

4

6

8

10 12 14 16 M ean baseline scores (w ith 95% C I)

18

Figure 7.15 Mean scores at baseline and confidence intervals by group for Psychological needs domain Reported mean difference scores of -0.34, p = .94, 95% CI -9.69, 9.00] and -1.27, p = .49, 95% CI [-5.01, 2.48] showed the intervention group decreased more than the control for the BPN-SF (whole score) and the autonomy subscales respectively. Reported mean difference scores show the intervention also decreased less than the control on competency and relatedness subscales (see Table 7.1). No reported mean difference scores were statistically significant. Minimal change scores (see Figure 7.16) and lack of statistical significance (see Table 7.1), suggests measures in this domain did not capture any significant change in students’ perceived satisfaction of psychological needs.

121

M usic

BP N -S F (w hole score)

N o music

-54 -50

BP N -S F (autonomy )

-45

-40

-35

-30

-25 -20

-15

-10

-5

0

5

10

15

20

25

30

35

40

45

50 54

M usic N o music

BP N -S F (competence)

M usic N o music

BP N -S F (relatedness)

M usic N o music -17

-15

-10

-5 0 5 M ean change scores (w ith 95% C I)

10

15

17

Figure 7.16 Mean change scores and confidence intervals by group for Psychological needs domain Quantitative results summary. Data collected for this study was unable to provide a definitive statistical outcome regarding the effect of the music program on any psychosocial wellbeing constructs measured by the survey. Likewise, there was no statistically significant difference between the mean change scores reported by control and intervention groups. Undoubtedly this is affected by several study limitations, including sample size and potential ceiling effects. A more detailed discussion of such limitations is presented in the first critical reflection analysis (see Chapter 9). Student Qualitative Results Nine students and the class teacher from the intervention group were recruited to participate in semi-structured interviews. All student interviews were conducted one-onone and lasted between approximately five and 20 minutes. The teacher interview lasted approximately 25 minutes. Interview data was analysed using a combined inductivedeductive qualitative approach that included three main steps (explained in Chapter 6).

122

Step 1: creating category templates. In the first step, a “category template” was created to help organise data into the main areas of interest to the project. These categories were: “Psychosocial wellbeing outcomes”; “Perceptions of the program”; “Perceptions of the survey”; and “Emergent themes” (see Chapter 6 for category description). This step took place before further “analysis” procedures were undertaken. This was to ensure data would be (loosely) arranged in a way that enabled each of the a priori research goals to be addressed after the following two inductive analysis steps (described below) were completed. The same category template was created and used for both student and teacher interview data. The remaining analysis steps, while following the same process, were applied to student and teacher data independently. Examples and explanations of the student analysis are presented first, followed by the analysis of teacher data. In both cases, the first-person voice is used to indicate the process of interpretation. Step 2: coding student interviews. After creating the category template in the first step, I identified “key statements” (McFerran & Grocke, 2007) and “important moments” (Fereday & Muir-Cochrane, 2008) within interviews and gave them a code. Where possible, I created codes using the participants’ own words in an attempt to stay connected with the data. In naming and identifying codes, I also attempted to follow the criteria offered by Boyatzis (1998) and Fereday and Muir-Cochrane (2008), in that codes should “capture […] the qualitative richness of the phenomenon” under investigation (Fereday & Muir-Cochrane, 2008, p. 83). This proved an important definition for this study given that many transcripts contained only sparse and or ambiguous data. Often it was the character of a response, the perceived meaning behind its ambiguity, or when there was no response at all that made “moments” within an interview important for me. For this reason, I believe the fact I had conducted interviews myself greatly aided my analysis; I could remember the participants and the interviews. This included body language, hesitant pauses and silences that, while meaningless in a written transcript, I remember as representing meaningful moments in regards to the subject matter. I believe creating codes based on a simple reading of the printed transcripts could easily have led to a misinterpretation of the meaning embedded within these interview “moments”. As such, I often relied on my memory of the interactions and referred back to the digital recordings of interviews when identifying codes.

123

This analysis approach is consistent with Poland and Pederson’s claim that “texts generated through the transcription of interview recordings, even so-called verbatim transcripts, are necessarily only partial accounts of the original interactions, which are themselves imperfect windows into naturally occurring experiences” (1998, p. 302). Further, they suggest “transcriptions and qualitative research based on transcripts cannot adequately capture […] a broad range of methods of expression that are primarily nonverbal or that do not rely on written text” (1998, p. 303). I resonated with these comments, and feel they validate my interpretation of non-verbal cues and the unspoken meaning behind certain responses. My interpretation of silences was also validated by Poland and Pederson’s conclusion that interviewee silences in qualitative interviews are often laden with meaning, and can be as important as verbalised responses. While it is not possible to represent all of these nuances using text, I have in some cases attempted to include short descriptions of non-verbal or contextual elements of communication in interview excerpts presented here, or in their interpretation. I have often used the term “interaction” to refer to pieces of data I felt told me something meaningful. This is to signify the fact such pieces of data represented interactions between the participants and myself, and often included non-verbal elements. Each interview was coded one-by-one in order to immerse myself in the context and memory of each individual perspective. As codes were identified, they were allocated to one or more of the categories in the category template based on my interpretation of their meaning and relevance to those categories. Sometimes the analysis of one interaction with a student resulted in several codes. For example, an interaction with participant 7a19 contributed to the formation of three separate codes allocated to the “Psychosocial wellbeing outcomes” category: Interviewer: Do you think you changed as a result of the music program? 7a19: Yep, made me more confident. Interviewer: How could you see that, or how could you feel that? 7a19: I talked to people more, and made friends. Here this boy suggests that participating in the program increased his confidence, which was evidenced by the fact he talked more, and made friends. While there appears to be three outcomes here, I treated this as a single interaction based on my perception that these outcomes were interdependent. I also decided to break them up into three 124

separate codes. First, the most obvious outcome was their suggestion of increased confidence, which I coded as “confidence”. Second, I interpreted the statement “I talked to people more” as meaning the program facilitated increased communication for this boy, and thus coded it as “communication”. Lastly, the second statement in the last sentence suggested that this boy had made friends as a result of the program, and was simply coded “made friends”. The fact that this boy had offered these responses to the question “Do you think you changed as a result of the music program?” indicated to me they were considered important outcomes for him. An example of coding in the “Perceptions of the program” category is taken here from an interaction with 7a10: Interviewer: Did you learn anything through the music [program]? 7a10: (Silence) Interviewer: Was it mostly just fun or was there…? 7a10: (Jumps in) Yeah, fun. Interviewer: Did you learn anything say, about life or friends through it as well? 7a10: Just for fun. This represents the typical type of interaction I experienced with this boy during our interview; he was regularly unresponsive and the session was beset by frequent pauses, silences, and short, reserved answers that often seemed to be led by my questions rather than his own personal experience. What is evident from reading this passage is that he did not identify any learning outcomes related to his participation in the program. He did, however state that it was fun. Worried that I had led him to the “fun” answer by my initial prompt, I followed with further prompts related to a number of learning areas the program specifically aimed to address. Despite these prompts, he maintained his original position, and strengthened it by adding it was “just” for fun. I interpreted this interaction to mean this boy did not consider the music program to have any educational focus at all, either in an academic or wellbeing sense. Rather, it appears that for him the program was solely about having fun. Consequently, this was coded “just for fun” and allocated to the “Perceptions of the program” category. An example of how data was coded into the “Perceptions of the survey” category comes from participant 7a03’s response to my directed question regarding his experience of the survey and survey process: 125

Interviewer: […] how did you find it? 7a03: Pretty useless. Interviewer: Pretty useless? Ok, cool. How come? 7a03: Probably because the questions all seemed not to make sense. Interviewer: Yeah, ok. What was it that didn’t make sense? Was it the way they were worded, or the actual words in there, or…? 7a03: They way they were worded. Here, this boy reports the way the survey was worded meant it didn’t make sense to him, and was “useless”. My interpretation of this was that, not only were the questions and answers used in the survey meaningless to him, but also the purpose of the survey was not clear. This resulted in two codes. The first related to his reported lack of comprehension and was labelled “didn’t make sense”, while the second aimed to capture my interpretation that it made the survey meaningless for him, and was named “pretty useless”. As mentioned, the “Emergent themes” category was created to allow for and capture codes that did not fit neatly into any of the three categories above. Interestingly a significant amount of data was coded into this category that, while not specifically related to either “Perception of survey” or “Perceptions of program” categories, I perceived to be central in addressing the research questions. One example of a code in this category was taken from the following interaction with 7a01: Interviewer: What was the Keys To Success program about? 7a01: It was like, how to be good, and how to be like not, ah, how to like respect and…(trails off) Interviewer: Ok. How did the music fit in with that? Did the music help that program do you think? 7a01: Yeah it was music too. Interviewer: Ok, but how was the music part of it different to the [non-musical version of the Keys to Success] class? 7a01: Ummm…(long silence)…it was good. Interviewer: Yeah? Was there anything that was really different from say, the music [classes] that you were doing and the… 7a01: (Jumps in) Yeah, we did the music, like, we have to do the music, like, no one writ it, [our class] did it. 126

Again, this was typical of the sort of interaction experienced with this girl in the interview. Some responses were only partially related to the questions I was asking, while others felt like they were either an attempt to give me the “right” answer, or any kind of response to get through the interview. The latter seemed due to a struggle to understand the concepts or questions, something also evidenced to me through long pauses, trailing off, or offering incomplete or nonsensical answers. I got the impression she felt the interview was akin to a test or academic assessment; to which she was offering the minimum responses she felt were necessary to get through it. In the above passage she is able to identify two tenets of the program (“to be good”, and “to respect”), though her articulation of them was strained. Her response to the next question, however, is coherent only in that she recognises that music was involved in the program. She seems to miss the fact that I was asking about “how” the music element related to the program and its content, simply confirming that music was involved. This appeared to indicate that she did not understand the concept or the question, something further strengthened by her response to my following prompts – which aimed to clarify that I was asking about the impact of the musical element. The “ummmm…”, followed by a silence, and then “it was good” implied to me she still did not grasp the concept of what I was asking. She could have said it made no difference or that she did not know (as other participants did), instead it appeared she tried to deal with this by either silence or offering a generic “coverall” statement. The fact she then cut me off and replied confidently to my last prompt for this question indicated to me that she felt like she finally understood what I was asking. That her last answer was again unrelated to the question further strengthened my perception that she did not understand this particular interview topic. As such, I coded this interaction in two ways: first, I felt this interaction showed she didn’t understand my questions, and thus coded it “didn’t understand interview questions”; and more fundamentally, given the repeated prompts, I also felt this girl had trouble grasping the concepts contained within the question, thus also coded this interaction as “didn’t understand the concept”. Step 3: creating themes. Once I was satisfied with the coding, I sat with the data in each category and looked for patterns. This involved identifying when interactions, or codes, from different participants suggested a similar pattern. It also involved identifying a pattern in several interactions or codes from only one participant. From these patterns, I created 127

themes that represented the essence of a given pattern. Other themes were based on only one code or interaction that seemed to me a significant representation of participant experience relevant to the study. This differs from traditional content analysis methods, in that the frequency in which a theme or code arose did not influence the importance I attributed to it (Fereday & Muir-Cochrane, 2008). The purpose of creating themes was to organise the data in a way that presents the results of the qualitative data, which could be used for further analysis, and discussion in relation to the main research questions. A visual display of how codes were grouped to form each theme is presented in Table 7.3. This is followed by an example of how each theme was created from codes.

128

Table 7.3 Keys to Success Student Interviews: Codes, Themes and Categories Category Themes Codes

Category Themes Codes

129 Category Themes Codes

Category Themes Codes

Psychosocial wellbeing outcomes Outcomes extended Social connection School engagement beyond KTS class Getting along In other classes too Wanted to come to school Made friends Enjoyed being at school Communication Being in class was better We did more Felt like part of the school together Different view of teacher

Engagement with learning

Prosocial behaviour Music calms me down Bullying More relaxing I don't fight anymore More happy Affect

Self-efficacy Absence of outcomes

I learnt some stuff Confidence No outcome Music is my way of learning Coping The same as before Its more fun than just writing Don't know Music makes studies more fun Probably not and easy Perceptions of the program We chose We did it I had a part The role of It was Made class It was just about It was about music We had to Same as Not just We did it the music together to play music fun better having fun education do it normal KTS the music We voted No one else did Everyone did it There was Music helps us It was Class was It was about We learnt to use I had to Outcomes It was the it, we did something understand fun more fun having fun music sing/ direct came from acting for everyone whole KTS We got to We wrote it We played more I got to do We learnt stuff We had Being in class Just fun We learnt about We had to The music It was choose together my part through music a laugh was better musical stuff choose didn't make making the video We had to We sang it We all joined in Music helped a Want to Class was We left the It was about being We had to a big difference decide bit do more more relaxing classroom musical do it We made it I interacted more The music was because Better than It was about fun It was to learning to about other stuff it’s fun other classes and learning music be more musical Everyone did it with others Perceptions of the survey Couldn't see the point Didn't really like it Lack of comprehension It was pretty useful Questions not fully comprehended We pretty much took it seriously It was useless Didn’t like it Dubious answer It made me think about my life Some yes, some no I did it seriously Didn't know what it Followed my question It was good for the school Didn't made sense Some people are dishonest was for Didn't understand It was pretty good Suspicious answer I'm pretty honest Most people are honest Don't know Emergent themes Didn't grasp the concept Responding to what I said Contradicting or changing stories Off the track Non-committal Its hard to explain Just suspicious Didn't understand the program Used my wording Backtracking Off topic Path of least Hard to explain Coverall term resistance Didn't understand the interview Just repeating Changed story Nonsensical Simple answer Suspicious question answer Didn't understand the survey Just agreeing Contradiction Dubious answer Coverall response Didn't understand the concept Same response Dubious answer Dubious answer

Psychosocial wellbeing outcomes. Social connection. Codes within the “Psychosocial wellbeing outcomes” category suggested there were some distinct psychosocial wellbeing outcomes reported by some students. While these outcomes were rarely reported by more than half of the participants, and often not directly related to the musical element of the Keys to Success program, they suggest the program did affect participants in several ways. The first of these related to the “communication” and “made friends” codes mentioned above, and a code named “we did more together”. Collectively, responses in these codes indicated the music program helped bring people together, whether this meant making “more friends”, increasing “communication” between students, or facilitating activities in which they “did more together”. Participant 7a13 explained it like this: Interviewer: Do you think the music was useful, when you’re talking about getting along and… 7a13: Yeah, people get along and, like, join in with each other…yeah. Interviewer: When you say getting along and joining with each other, your saying the music helps with that? 7a13: Yeah. Interviewer: How? 7a13: For communication with others, like we get along and have fun with each other too, in class, with other classes too. Interviewer: When you’re doing the music, you mean? 7a13: Yeah. In response to a previous question, this participant identified that, for him, the KTS program was about “getting along”. Here we can see that when asked whether the music helped with that, he replies “yeah”, and then suggests that it does so by increasing “communication”, as well as “getting along” and having “fun” with others. I interpreted the “getting along” and having “fun” to mean that students were engaging with each other in a rewarding and amicable manner. Four of the nine participants made comments that were grouped into this theme, which suggested to me that, through musical activities, the program facilitated a “friendly” connection between some students. As such I labeled this theme “social connection”, using “connection” to

130

indicate that these students were interacting in a manner that went beyond coparticipation in class activities. Further, “social” was used to signify that this connection was at the social level, as is articulated in 7a13’s above statement that it extended to other classrooms. Outcomes extended beyond KTS class. This idea of extending past classroom boundaries was also captured in a similar theme named “outcomes extended beyond KTS class”. This theme was based on data from two participants, including both the interaction above, and another with participant 7a19, who, when asked how the program made him feel about being in class, responded: 7a19: More happy and confident. Interviewer: More so than before? 7a19: Yeah. Interviewer: And was that for all classes, or just when you were in music class? 7a19: All classes. This suggested that, for this boy, not only did the program lead to him being “more happy and confident”, but these outcomes were also extended to “all” of his classes. To me this indicated that, for some students, psychosocial outcomes experienced were not necessarily confined to the music program, but extended to other areas of their school life. This theme was named “outcomes extended beyond KTS class”. School Engagement. The notion of being happier in class also helped form another distinct theme in this category. Several students suggested the music program made them “want to be in class more”, or made class “more fun”. Some students also noted that because of the program, they “just wanted to come to school”, or were “enjoying” being at school more. Participant 7a17 described it in this statement: 7a17: You actually want to be in the class, like if it’s just a normal class you know it’s going to be the same, but it was kinda different.

131

Here this girl explains how the music program classes were “kinda” different to her regular classes, which meant she wanted to attend them. The first statement, “You actually want to be in the class”, suggests she did not always want to attend other classes. Further, it appears the music element, which offered something “different” from her normal classes, was what increased her willingness to attend this class. This, and similar comments from other students (eight in total) indicated to me that the program helped increase some students’ motivation to attend school and class, and revealed a pattern of increased school engagement – even if sometimes limited to a single class. These codes and interactions were grouped here under the theme “school engagement”. Engagement with learning. For several students, this “school engagement” went further. Four of nine participants made comments suggesting the music program helped with their learning. This was expressed by 7a05 as follows: Interviewer: […] did you feel like you were still learning when you were doing the [musical] stuff? 7a05: Yeah, yeah, I did. Yeah, I learnt, I learnt, pretty much, some stuff. Interviewer: Ok, cool. Where you learning anything when you were doing the [non-musical] work? 7a05: Errr…not a lot. Initially this boy’s response that he learnt “some stuff” in the musical program suggested to me that his learning was minimal. However, considering his admission that he did not learn “a lot” during non-musical classes – and, seeing that this statement was prefaced by an utterance and pause, the (somewhat guilty) admission implied that “not a lot” was probably more like “nothing” – it seemed this was a significant outcome for him. For participants like 7a03, however, the effect of the music program on learning seemed more pronounced: Interviewer: Would you like to do more music things like this in class or at school? 7a03: Yes, definitely. Interviewer: Why’s that? 7a03: Because I like music, I find music is my way of learning things, it helped me a lot.

132

Interviewer: Just with class work or social stuff as well? 7a03: Yeah, both. This suggests the program helped this boy engage with learning in both social and academic spheres. I perceived this an important outcome for him given both the question he was responding to, and the confidence of his response. These and similar interactions formed the theme “engagement with learning”. As a side note, part of me felt some responses that informed certain themes above were somewhat suspicious. For example, 7a03’s claim that music was his “way of learning” may have been an effort to provide feedback that would lead to more music and less work in classes (as I was sure he perceived there was a evaluative element to the interview), rather than a profound reflection on the impact of music on learning. Nevertheless, in many cases I felt my commitment to representing the voice of participants meant I needed to give students the benefit of the doubt, and take their responses as they were articulated. With some responses, however, I did not find this viable, as they appeared too suspicious both in content and or context to be taken at face value. These are discussed later in the “Emergent themes” category. Affect. Also populated with potentially suspicious statements was the “affect” theme. This theme represented participant reports that the program influenced their mood, emotions or “feelings”. While there were few of these, one vocal example was from 7a05: Interviewer: And do you think you changed as a result of the music program? 7a05: Ahh, yeah, probably a bit. Interviewer: And how do you think you changed? 7a05: I think like, before that, I never used to listen to music, and […] I used to always get in fights when I was angry, and like, since she told me that [if you listen to music it will calm you down], every time I want to do something bad, I pretty much sing a song. I have labeled this interaction suspicious for two reasons. Upon sharing the response with my supervisor (also the music therapist who facilitated the program), she expressed doubt that this is was what the boy learned from the program, since students were not “told” listening to music would automatically calm them down. Further, the

133

notion that this boy would stop and “sing a song” when facing violent impulses or physical confrontation seemed questionable to me. However, I recognise my doubt was based on a perception I had formed through minimal contact. While he might not have been “told” that listening to music will calm him down, it is entirely possible he made that interpretation during the program. These reasons, coupled with my commitment to represent participant voices, meant this exchange and brief comments from two other participants that the program made them “happy” or “relaxed”, formed the “affect” theme. Prosocial behaviour. Two students claimed the program helped them address social behaviour issues. For example, 7a17 offered this explanation of the musical video they made as part of the program: 7a17: It was about like how your friends sometimes don’t talk to you and bully you and stuff, so how to like cope with it and get confidence. Interviewer: And you learnt about this through doing the music? 7a17: Yeah. Interviewer: Did that work well doing it with music? 7a17: Yeah. Interviewer: Why? 7a17: Because the video and that was all about bullying, like how your friends bully you and you don’t have someone, and you don’t have any confidence, but then another person comes in and helps you and gains you confidence. Here this participant describes that the video included themes related to bullying and helping others. My interpretation of what this girl was describing was that the program helped her to learn about the effects of bullying, and what you can do to help others who are being bullied. Several other students made similar comments which, combined, formed the theme, “prosocial behaviour”. Importantly, in suggesting that this represents a psychosocial outcome of the program, I concede that a student merely mentioning prosocial behaviours does not translate to them engaging in them. However, I feel the fact students mention them here in relation to the program means they did

134

engage in thinking about them and their importance, which I believe is a crucial step to their adoption, and an outcome in itself (albeit it a somewhat suspicious one). Self-efficacy. The confidence mentioned in 7a17’s above statement also informed the “selfefficacy” theme. Included within this theme were comments from five participants. Some made references to “coping”, and knowing “that there are always things worth fighting for”. However, predominately students described this in terms of “confidence”. This included confidence in dealing with “bullying”, “talking to others”, and “just confidence”. For example, 7a03 explained that the music part of the program was about: 7a17: Like, giving us confidence, and building our abilities to do things in front of others. This suggests the program helped increase this boy’s self-efficacy by developing his confidence to “do things in front of others”. Despite Bandura’s (1997) caution against equating confidence with self-efficacy, for me, the use of confidence here did relate to self-efficacy given it met Bandura’s own criteria of affecting “beliefs in capabilities to organise and execute the courses of action required to manage prospective situations” (1997, p. 2). Absence of outcomes. The data and themes above indicate students reported a number of program outcomes during interviews. Despite this, most interview data indicated outcomes were largely absent. This was predominantly evidenced by answers to questions in the “psychosocial outcomes” section of the interview schedule (see Table 6.2, Chapter 6). These answers were often short, and indicated little or no change from participating in the program. For example, when asked whether the program made a difference to how they felt about classmates, of the nine participants three simply said “nuh”, one responded “no”, two said “not really”, and one shook their head. Another two responded “nah”, explaining that because they “all knew each other pretty well” before the program, “it didn't make a difference”. Similarly, when asked if the program changed how they felt about their connection with the school, three said “nuh”, one “no”, another “not really”, and 7a03 claimed: 7a03: Nah, it just kept me in the same position. 135

Five students denied the program had a significant impact on their learning, six claimed it made no difference to their relationship with their teacher, and four reported it had no impact on how they felt about being at school. Four students also stated their feelings about being in class were the same as before the program, with 7a03 offering this explanation: Interviewer: How did [the music program] make you feel about being in class? 7a03: Umm…Alright. Interviewer: Still not awesome? 7a17: Nah, not yet. Interviewer: Why not awesome? 7a17: It’s class. This suggested that, while the music program might have been fun for this participant (an acknowledgement he makes elsewhere in the interview), the program did not change his negative perception of being in a classroom environment. Further, it appears the mere fact that it was conducted within in a classroom potentially mediated some program benefits. When students were asked how the program made them feel about their place in community or society more generally, one said it was the same, while another said they didn't know. Two more sought further explanation of these concepts, and upon clarification also said there was no difference. In total, all nine participants reported an absence of at least three different outcomes addressed in the interview schedule. These responses (and a number of suspicious responses discussed later) led me to the conclusion that, while some students did report outcomes, these outcomes were not experienced by all students, or in all areas. The above responses were placed under the theme “absence of outcomes” to indicate that in many cases, no benefits or outcomes were reported.

136

Table 7.4 Raw Data and Codes Used to Form the “Engagement With Learning” Theme Category Theme Code

Data

Psychosocial wellbeing outcomes Engagement with learning Its more fun Music is my way of I learnt some stuff than just learning writing 7a05 7a03 7a13 7a05: I learnt pretty much some Int: Would you like 7a13: It was stuff. to do more music more fun than Int: Were you learning anything things like this in doing just when you were doing the [nonclass? writing, it’s musical class]? 7a03: Yes, definitely better like that. 7a05: Ahhh, not a lot. Int: Why? 7a03: Because I like 7a13 music, I find music is Int: Did you learn anything [during my way of learning the music program]? things. It helped me a 7a13: (pause) Sort of, yeah. lot. Int: What sort of things did you Int: Just with class learn? work or social stuff 7a13: Like, I don’t know, it’s hard as well? to explain 7a03: Yeah, both Int: Was it the sort of stuff that KTS was about? 7a17 7a13: It was mainly about the keys Int: Would you like to success to do more music things like this in 7a13 class? Int: Can you still learn stuff in class 7a17: Yeah […] when it’s fun? because it makes the 7a13: Yeah, sort of studies easier and Int: Only sort of? more fun 7a13: Yeah, it’s fun, it’s half-half Int: Half fun half learning? 7a13: Yeah 7a17 Int: Did you think [learning through music] works with the social stuff […]? 7a17: Yeah. Int: Do you think it would help learning about other things as well? 7a17: Mostly Keys To Success.

Music makes studies more fun and easy 7a17 Int: Was learning with the music any different from when you were doing the non-musical KTS stuff? 7a17: It was actually more fun […] Int: Did that make it easier to learn, or just the same? 7a17: Easier to learn Int: Why? 7a17: Because when you have the fun bit in there it’s easier, but when it’s all boring you cant learn anything. 7a17 Int: Would you like to do more music things like this in class? 7a17: Yeah […] because it makes the studies easier and more fun

In total, eight themes were identified in the “Psychosocial wellbeing outcomes” category, all of which have been summarised here. An example of how the raw data and codes were combined to create themes in this category is presented in Table 7.4. Perceptions of the program. Themes in this category were divided into two sub-categories. The first related to particular elements of the program perceived as important to participants (i.e. activities), and which potentially aided in the facilitation of any psychosocial wellbeing outcomes. The second related to participants’ general perceptions of the program, including its nature and purpose. 137

We chose the music. In the first subcategory, elements of the program reported as beneficial or significant for students often related to musical and other creative activities. For example three students identified the fact that they “chose” or “picked” the songs they worked with as memorable. For example, 7a05 explained: 7a05: Like we were doing some music, we get to make up our own song, we did a remix, like everyone gets to choose a song, and votes the most won, gets to choose the song, and we changed the remix, and it was the Eminem one, the one we picked. Like similar comments made by other students, this statement was offered while giving a general description of the program at the start of the survey, indicating they identified strongly with this element of the program and that it was important. Further, it appeared to me this choice, or opportunity to influence the content of their own classes, was an uncommon occurrence for these students. Based on this perception, and the fact many students used the phrase, this theme was named “we choose the music”. We did it. Similarly, students often identified that they “made up”, “wrote”, “sang”, “played”, or “remixed” the music. This idea of authorship was articulated by eight of the nine participants, suggesting it was a significant aspect for the group. This authorship was not limited to music, but extended to making the video: 7a16: We made a song, produced it, made a video of it. While simple, this boy’s response was important for me because it was the only description of the program he offered, and in an interview that lasted less than five minutes, was one of few responses containing more than two or three words. Therefore, I interpreted this to be one of, if not the most, important aspect of the program for this boy. This and similar responses formed the theme “we did it”. We did it together. Five participants reported doing things together as a group during the program was important. Again, this related mostly to artistic activities, as explained by 7a01: 7a01: It was really good, coz […] we went out and everyone sing-ed it, and everyone watched it, it was good.

138

This girl is responding to an open question at the start of the interview regarding her perception of the music program. First she claimed it was “really good”, and then she suggested this was because “everyone” participated. What gives this statement strength for me is that she mentions “everyone” twice, and that it was a response to an open question. This and similar responses formed the theme “we did it together”. I had a part to play. Three participants expressed the fact that everyone had a role to play was important. Participant 7a13 made particular reference to the fact there were different roles for different people in the class: 7a13: There’s producers, some people record people when they sing and some people write like, there are keyboards, there are people for keyboards, and other people like recording and singing different like…yeah. For me, this statement was powerful because it was in response to an initial open question. This suggested it was his strongest perception or memory of the program. I also inferred that everyone having their own special role to play in reaching a common goal was something participants did not always experience in class, which for some is was what made the program special. Participant 7a10’s response when asked what they thought of the program supports this idea: 7a10: Great. Coz I was part of the singing the video, I had to be director. My interpretation of this was that the boy felt recognised as a valued team member, which was again an uncommon experience. Such data was placed under the theme of “I had a part to play”. The role of music. Data in this first subcategory also contained references to the role of music in the program. Some made direct reference to the fact musical activities were about addressing issues such as “confidence” and “respect”, while others made suggested music helped them learn, or understand program content. For example, when asked if music helped address issues such as behaviour, participant 7a03 replied: 7a03: Yeah, it made it easier since teens […] understand through music and stuff.

139

For me, this meant 7a03 found music an appropriate vehicle for learning about program content. This and comments from three others intimated a recognition that music had unique role in the program, and formed the theme “the role of music”. It was fun. All students clearly identified that the program was “fun”, and that they “had a bit of a laugh” with each other. While several suggested having “fun” was an outcome on its own, for others it appeared that it was the “fun” element that led to potential outcomes. For example, 7a10 suggested it was “because it was fun and we laughed” that he had “more friends”. Yet most commonly it was this element that made participants want to do more music things in class. This data formed the theme “it was fun”. Made class better. The music program was also reported to improve students’ experience of being in class. For some it made class more “relaxing” or “fun”, while for others it made it just “different” enough to want to be there. As participant 7a17 described: 7a03: You actually want to be in the class, like if it’s just a normal classes you know it’s going to be the same, but it was kinda different. This indicated to me that the musical element of the program was enough to keep some students engaged in the classes. This is typified here by this girl’s comment, “You actually want to be in class”, which suggests she was not typically engaged in, or wanting to attend, other classes. While the rest of the statement implied that it was only the novelty that engaged her, it nonetheless indicated to me that this girl was more engaged in class as a result of the music program. In total four students made comments which contributed to this theme, which was named “made class better”. It was just about having fun. That the music program was “fun”, and sometimes only valuable in its novelty, leads to the next subcategory of themes in this category. While the above themes represent program elements perceived as important and potentially facilitating psychosocial outcomes, several interactions and codes in this category tell a different story. For example, there appeared to be some confusion about the KTS program, and the role of music. When asked what the best bit of the music program was, several

140

students claimed it was being allowed out of class (to make the video in the playground), or as 7a05 put it, “we stopped doing work and […] just had started having fun”. While such comments could be interpreted to mean the musical element meant learning became fun, and didn't feel like work, for me a distinct pattern emerged that suggested otherwise. Several students appeared to have either missed the fact that the music program was also about learning life skills, or, it didn't help in learning them. One interaction that illustrates this pattern was with 7a16: Interviewer: What was the [larger] KTS program about? What was the point of that? 7a16: Confidence…organisation…and I think that's it. I’m not sure. Interviewer: Do you reckon the music helped, you know, learning about confidence and organisation and all that kind of stuff? 7a16: A bit. Interviewer: A bit? Not a lot though? 7a16: Nuh. Interviewer: Nuh? No worries. So, was it useful do you think in that class or more fun? 7a16: Just fun. Here this boy acknowledges the role and purpose of KTS program, but appears hesitant when asked whether the music played a role. This perceived hesitance is more pronounced when considering the tone of voice and avoidant body language during this interaction. Both suggested the need give an answer that didn’t get him in trouble. Further, I perceived a level of relief that he didn’t have to maintain this act in his grunted “Nuh” to my further probe of “Not a lot though?”. The rest of the interaction went on to confirm my feeling that this boy considered the program to be more about having fun than learning. This and interactions with six other students suggested that, for them, fun and avoiding work were what made the program valuable. As such these interactions formed the theme “it was just about having fun”. It was about music education. Another perception that emerged was that the program was about becoming more musical, either in terms of learning about music or gaining musical skills. Interactions with four participants formed this perception, one of which was with 7a05:

141

Interviewer: So, do you think you learnt anything through that music program? 7a05: Ahhh…probably a bit, yeah. Interviewer: Ok, yeah. What sort of things did you learn? 7a05: Probably how to like, rhyme words, and pretty much that’s it. Despite this participant’s earlier (and later) claims that they did learn other nonmusical things from the program, this interaction suggests that for them music education was one of the main purposes of the program (it also represents a level of contradiction among responses which will be discussed later). This pattern was represented by the theme “it was about music education”. I considered this an interesting pattern within the data regarding program perceptions, as music education was not a focus for the music therapist or class teacher. We had to do it. Another theme that emerged in this category was based on the word “had”, which appeared in several descriptions of the program. Specifically, three participants used this word when describing program activities. One example was from 7a10: 7a10: I had to do this music thingy to umm, make up a song as a class, and became successful, and we had to make a video. While at first I thought this a subtle and insignificant grammatical detail, it came up so regularly that I decided to pursue it further. Initially, I thought students were describing that they were made to do musical activities whether they liked it or not, but then dismissed the idea thinking that it was probably a reflection of the way these students talked. In discussion with the music therapist, I discovered that, given the class setting, students were not given the choice to participate in musical activities. Therefore, I interpreted the use of this word to signify students’ recognition of this lack of choice. While this potentially contradicts the agency suggested in the “we got to choose what we did” theme reported above, my perception was, while they could choose which song and activity they participated in, participation was not an option. This theme was named “we had to do it”. Same as normal KTS. Two students suggested reported psychosocial wellbeing outcomes were not a result of the music program. This is exemplified in the following interaction with 7a03: 142

Interviewer: So, with the music stuff in Keys to Success, did you learn anything? 7a03: Yeah, not to bully. Interviewer: Was this different from the non-musical parts of KTS? 7a03: (Pause) Probably not. Interviewer: Probably not? It was pretty much the same? The music didn’t make a difference when you were learning about… 7a03: (Jumps in) It made it a lot more fun, like, other than that, nuh. While acknowledging that he had learnt “not to bully” through the music program, when questioned further he suggests this was no different from what he would have (or had already) learnt in the non-musical KTS sessions. I found this response offered an interesting comparison between musical and non-musical versions of the KTS program, and used it and another similar interaction to construct a theme named “same as normal KTS”. It must be noted this theme (like many reported here) does not represent the experience of all participants. This is based on interactions with only two students, and other students made statements directly contradicting this idea. Not just the music. Two students also suggested it was not necessarily the musical element of the program that facilitated outcomes. This included 7a16: Interviewer: Have you changed since you participated in the program? 7a16: A bit, yeah. Interviewer: How? 7a16: I have more confidence. Interviewer: What was it about the music that helped with that? 7a16: Acting. Here 7a16 explains suggests it was his acting experience rather than the “music” that facilitated his confidence. This was important for me not because it meant music had no impact, or was not an important part of the process, but rather because it suggested “music” is not always sufficient in itself to achieve outcomes: sometimes it is the opportunities afforded through music programs that are important, not the musical content. Here it appears that it was not the songs or instruments that helped this boy, but

143

rather making a music video meant he was given the chance to act. This theme was labeled “not just the music”. Twelve themes were identified in the “Perceptions of the program” category, and have been summarised above. An example of how the raw data and codes were combined to create themes in this category is presented in Table 7.5. Table 7.5 Raw Data and Codes Used to Form the “It Was Just About Having Fun” Theme Category Theme Code Data

Perceptions of the program It was just about having fun It was about having fun 7a01 Int: Would you like to do more music things like this in class? 7a01: Yeah […] me and my friend do because […] it was fun. 7a05 Int: Was it different [from the non-musical KTS class] when you did the music? 7a05: Yeah, we used to do like different stuff, we never used to sing songs and things like that, and we used to do like, lots of work with [the teacher], and then she [the music therapist] came in and she pretty much saved our life. Int: What do you mean by saved our life? 7a05: Like stopped doing work, and we just started having fun. 7a05 7a05: We all had a fun day. And we started laughing at the stuff, and whenever we used to sing, some of us used to laugh and joke around and all that. 7a19 7a19: It makes the class more fun. 7a17 7a17: It was actually more fun.

We left the classroom 7a03 7a01 Int: Was [learning] different 7a01: […] it was from the non-musical parts of good, it was really KTS? good. Coz […] we 7a03: (Pause) Probably not. went out [of the Int: […] The music didn’t classroom] and make a difference when you everyone singed it, were learning about…? and everyone 7a03: (Jumps in) It made it a watched it, it was lot more fun, like, other than good. that, nuh. 7a03 7a10 Int: How did you Int: Was it mostly just fun or find it [the was there….? program]? 7a10: (Jumps in) Yeah, fun. 7a03: It was actually pretty good Int: […] did you learn Int: What was good anything say, about life or about it? friends through it as well? 7a03: For me, when 7a10: Just for fun. got to sing my part. 7a19 Int: Anything else? Int: How did you find it [the 7a03: Probably the program]? fact that we were 7a19: It was fun. allowed out of class Int: Anything else? and stuff? 7a19: Just fun. Just fun

7a16 Int: Do you reckon the music helped […] learning about confidence and organisation and all that kind of stuff? 7a16: A bit. Int: A bit? Not a lot though? 7a16: Nuh. Interviewer: […] was it useful do you think in that class or more fun? 7a16: Just fun.

144

It was about fun and learning music 7a05 Int: Would you like to do more music things like this in class? 7a05:Yeah, I’d love to. It’s just that like, I would like to learn more about music and just to calm me down, and have more fun.

Perceptions of the survey. Couldn't see the point. The first themes that emerged in this category relate to general perceptions of the survey tool used. These perceptions varied among participants, for example 7a03’s “useless” comment, described earlier in the chapter, indicated that for him the survey was pointless because it didn't make sense. Responses from two other participants, such as 7a13, suggested they too failed to see its point: 7a13: I don't know. I didn't know what it was for. These responses formed the theme “couldn't see the point”. This aimed to capture the idea that the confusion surrounding the survey made it hard for some students to determine its purpose. Didn't really like it. Participant 7a11 responded that the survey was “all right” but when further prompted as to whether he liked it, responded “not really”. This combined with above responses led me to the conclusion that there was a general level of dissatisfaction regarding the survey among some students. This theme was based on responses from four participants and named “didn't really like it”. Lack of comprehension. While other students suggested the survey was “good”, often these suggestions seemed unconvincing. For example, one boy sounded so uncertain in his answer of “good” that I asked if he was sure – he laughed nervously, which I translated as an admission I had caught him out, and went on to suggest the survey did not entirely make sense. Another example is from an interaction with participant 7a10: Interviewer: What did you think about [the survey]? 7a10: (Pause)…Um…(long silence) Interviewer: Was it good, was it bad, was it annoying? 7a10: (Pause) It was good. It seemed this boy was not going to respond to my first question regardless of how long I waited, so I followed up with a prompt. Again this was met with an initial pause, and finally he said, “it was good”. It was not clear what the initial pause and long silence meant, however, it felt to me this boy either didn't remember the survey, didn't

145

pay attention while doing it, or it was confusing for him. For me, any or all of these could have meant he was unwilling to respond for fear of sounding silly, or revealing that he didn't understand or remember it. Alternatively, perhaps it meant he was unable to respond. Based on the initial perception that his comprehension was not sufficient to offer a detailed answer, I offered some possible responses to demonstrate that a simple answer was acceptable. The following pause indicated to me he was still struggling with this subject, and that the final “it was good” response was an attempt to offer a “safe” answer that also meant we could move on. Taking this lead, I did move on, yet then and later this interaction led me to believe that this boy did not understand the survey. I have grouped this and a similar response from one more participant under the theme “lack of comprehension”. While the “couldn't see the point” theme was also based on a lack of understanding, for me responses from these two participants reflected a lack of comprehension regarding the survey at a more fundamental level. It was pretty useful. Responses from two participants, however, suggested some students did partially understand the survey. While one participant thought it made sense and was “pretty good”, others articulated why they thought it was useful: 7a17: It actually made me think more about my life and my friends and teachers, and my school life. This suggests this girl grasped some of the survey content, in that it was addressing issues about school relationships and engagement, as well as peer and individual factors. Similarly, participant 7a01 reported: 7a01: Yeah, it was good for the school, every question was about the school and if you like it and the music and that stuff, yeah. Again this girl seemed to grasp the idea that it was asking questions regarding satisfaction with school life, and was related to music in some way. Further, to me the comment that “it was good for the school” suggested this girl also recognised the reflective or evaluative purpose to the survey. These responses were placed under the theme “it was pretty useful”.

146

Questions not fully comprehended. Predominately, students claimed questions made sense. In some cases this seemed a reasonable response, especially for the two participants whose responses formed the “it was useful” theme above. However, again several responses seemed questionable, included one student who said that they didn't understand what the survey was for and then went on to claim the survey made sense (it is possible, however, that this student understood the questions, yet could not see the point of why they had to answer them). Another source of scepticism came from students who gave suspicious responses regarding their overall perception of the survey. While they claimed that the questions made sense, I was under the impression they were either trying to give the “right” answer, or just get through the interview by giving answers that did not invite further discussion. Again, I perceived these answers did not reflect the reality of how these students experienced the survey. That 7a03 clearly articulated “all” questions did not make sense, and then specified that it was question wording that caused this, further fuelled my doubt that all students understood all questions. The pattern emerging for me here was that comprehension of survey questions varied from student to student, and sometimes question to question – the latter idea being supported by 7a10’s response: Interviewer: Did everything make sense? 7a10: Some made sense, but, some [didn’t]. I placed the above data under the theme “questions not fully comprehended”, in order to reflect my perception that, while some students understood some questions, on the whole comprehension of survey content was an issue for students. We pretty much took it seriously. The last theme here referred to whether students answered survey questions honestly and or seriously. Care was taken to preface this interview question with the acknowledgement that answers were confidential. Responses again varied. Some suggested they didn't know or were “not sure” if people answered truthfully. One boy started off by suggesting that honesty was an issue, but then backtracked: Interviewer: Do people answer those things truthfully? 7a05: No, not really. I don’t think people do…some people do some people don’t. But I think in this school, there are some Year 9’s and they’re bad, and they probably don’t tell the truth.

147

I was unsure of what to make of this response, it felt the boy realised he had revealed too much, and shifted the focus to other students. It was also possible he took the question generally. Subsequently I suggested that sometimes it’s not about being “bad”, but being bored and not taking it seriously. He answered: 7a05: I don’t think anyone done that in my class. I think they probably put their head into it. This last comment was consistent with responses from others in the group, who suggested students “mostly” answer truthfully. Some stated they themselves were “pretty honest”, or “weren’t joking” when completing the survey. I felt I needed to take these students’ word, and concluded that honesty was not an issue. In terms of taking the survey seriously, I was less convinced, but decided perhaps the level of seriousness was consistent with normal class activities. I named this theme “we pretty much took it seriously” to indicate an overall intention to take it seriously, but that perhaps this waivered, or was not at the level of “seriousness” desired in empirical research. In total, six themes were identified in the “Psychosocial wellbeing outcomes” category, all of which have been summarised here. An example of how the raw data and codes were combined to create themes in this category is presented in Table 7.6. Table 7.6 Raw Data and Codes Used to Form the “Couldn't See the Point” Theme Category

Perceptions of the survey

Theme Code Data

Couldn't see the point It was useless 7a03 Int: What did you think of that [the survey]? 7a03: Pretty useless. Int: Pretty useless? Cool. How come? 7a03: Probably because the questions all seemed not to make sense. Int: Yeah, ok. What didn’t make sense, the way they were worded or the actual words in there, or…? 7a03: They way they were worded.

Didn't know what it was for 7a16 Int: What did you think of that [the survey]? 7a16: I didn't really know what that was for, it was probably just about KTS stuff. 7a13 Int: What did you think of that [the survey]? 7a13: I don’t know, I didn’t know what it was for…

Emergent themes. Throughout data collection and analysis several themes emerged that did not directly relate to the above three categories, but which I perceived as important to the study. Primarily, these themes revolve around a pervasive sense of uncertainty I experienced during interviews and data analysis. This uncertainty was based on several

148

factors that emerged for me as persistent patterns in the data, and related primarily to understanding, comprehension, and suspicious interactions. Didn't grasp the concept. Understanding and comprehension have been discussed several times in relation to the above data, and appeared as a pattern throughout interviews that was not restricted to any one category. For example, the following interaction with 7a10 is one example of understanding surrounding program perception: Interviewer: This was part of KTS, which was part of a bigger program, right? 7a10: Yeah. Interviewer: So what was the actual KTS program? What was the point of that? 7a10: (Pause)..um..(pause) music matters, with um, education. Interviewer: So, that was what [the music therapist] was doing […] but the stuff that [the music therapist] was doing was only part of a larger KTS program, right? 7a10: Yeah. Interviewer: Yeah? So what was that larger KTS stuff about? 7a10: (Pause)…um…(pause) I don't know. Interviewer: Don't know? Ok. So was it about […] how to be successful […] something like that […]? 7a10: Um…(pause) to educate…like, music. Such interactions were not uncommon, and signified several things for me. Firstly, this boy seemed to have difficulty either distinguishing between the larger KTS program and the music program, remembering them, or being able to articulate their purpose. This was despite several leading prompts. Given the boy’s English was sufficient to articulate answers to other questions, and that he remembered some parts of the program, I felt in this case he struggled to grasp the concept of the larger KTS program, and distinguish between it and the musical element. Further, his responses contained many pauses suggesting he was thinking carefully about his answers. Despite this, his answers were simple, and almost childlike. It seemed he was grasping at anything he could remember to formulate an answer, yet had trouble presenting these with cohesive language. Further, even though I indicated that his first answer was off

149

track, he basically repeated it, only rearranging the words. This suggested to me this boy’s level of comprehension was quite limited: not only did he struggle to understand what we were talking about, but he also struggled to express himself in a coherent manner. Several similar interactions were experienced with other students, which caused me to think that it many cases these students were not grasping the concept of either the larger KTS project, or the music program. Another example was from 7a01: Interviewer: Did you learn anything through those [music] classes? 7a01: Yeah we learnt. Interviewer: What sort of things did you learn about? 7a01: We learned like, umm…(silence)…we learned like, umm, (pause), we learned like, music and how to sing it and that stuff…(trails off) Interviewer: Cool. Did you learn anything about life or friends, or anything like that? 7a01: Yeah learned how to be like, good friends. Interviewer: Yeah? Ok, cool. And what was the musical part of that […] does that make sense? 7a01: Oh yeah, because, um mister (?) was telling us what to do and he was doing the same, and like, we got some information. Again there were many pauses that signified to me this girl was having trouble either remembering, articulating, or had not understood, the program. Also, like the previous interaction, there was recognition of a musical education element, which was interesting given that this was not the focus of the program. This made me wonder how much of the program these students grasped. I experienced interactions like this with each participant at least once in relation to different aspects of the study and interview questions, and for some students like 7a01, this non-comprehension continued throughout the entire interview. This suggested to me that, overall, these students struggled to understand everything I was talking about. The theme “didn't grasp the concept” was created from over 20 interactions, and represents this pattern across these different categories and areas.

150

Responding to what I said. The above interactions also demonstrate another pattern that emerged throughout interviews (particularly with two participants). This was the perception that often participants were just responding to what I said. In the above example with 7a10, despite my perception that this boy did not fully understand my questions, he was still answering many of my prompts with “yeah”, suggesting that he did. This caused me to wonder how many short answers from participants were based more on my questions than their own experience. For me, this was also evident in the above interaction with 7a01. Her “Yeah, we learnt” answer to my first question seemed to be based on the fact that I was talking about learning, and a desire not to be found out that she did not learn much. This was reinforced by her laboured answer to my follow-up question, in which she appeared to have trouble articulating what she learnt. Further, my prompt about “anything about life or friends” was met with a quick response suggesting she learnt to be “like, good friends”. Again this appeared to be a response to my use of the word “friends”. Overall, responses from at least four participants led to my belief that, in many cases, students were just responding to words used in my questions. I named this theme “responding to what I said”. Contradicting or changing stories. There were also several instances in which students seemed to be contradicting themselves, or changing their stories. For example, in the interaction with 7a05 described in the “we pretty much took it seriously” theme above, the participant started off saying that people are not honest when answering surveys, and then changed this view. Likewise, in 7a03’s interaction (described in “it was just about having fun”), he began by saying he learnt things through the program, but when pressed admitted that he didn't. Interactions with these two participants, and similarly contradictory interactions with at least three others, led me to formulate the theme “contradicting or changing stories”. Off the track. There were also responses that seemed unrelated to my questions or prompts, which caused me to question their validity. A good example is again from an interaction with 7a01, also described in the “Identifying and coding key statements” section:

151

Interviewer: Was there anything that was really different from say, the music [classes] that you were doing and the… 7a01: (Jumps in) Yeah, we did the music, like, we have to do the music like, no one writ it, [our class] did it. In this instance, I was exploring whether this girl perceived any difference in how the music program differed from the non-musical KTS program in teaching them “how to, like, respect” (a learning outcome she articulated previously). To me her response, which explained that they were the ones who “did” the music, was quite off track to my question regarding how the music program differed from the non-musical KTS program. It was unclear to me whether this was due to a misunderstanding of my question, trouble to grasp the concepts of what we were talking about, or an attempt to avoid the question. In order to capture all of these possibilities, this theme was named “off the track”, as that is what these responses seemed to be. There were at least four participants that had interactions in this theme. Non-committal. Another theme that emerged was that all nine students sometimes seemed to offer non-committal answers. These included short answers of “yes”, “good”, or “it was fun”. An obvious example was from 7a01: Interviewer: How did [the music program] make you feel about being in class? 7a01: Good. It was fun and…good…(trails off) It felt such responses were based on a desire to give answers they thought I wanted to hear, and which would not invite further questions. While I do not dispute the program was “fun” and “good” for many students, these felt to me like safe answers that precluded commitment to the subject matter. This theme was named “non-committal”, and while some participants seemed to contribute more than others, every participant had at least one interaction that led to its creation. It's hard to explain. The issue of articulation was also a prominent theme throughout interviews. While often this was based on my interpretation that students were struggling to articulate themselves, sometimes it was the participants themselves who identified it

152

was an issue. For example, in the following response, 7a17 responds to my question about whether the program changed how she felt about her teacher: 7a17: Yes, a little bit, like…I don’t know how to explain it. This and comments from other participants confirmed my perception that they were struggling to articulate their perceptions in relation to the subject matter. These formed the theme “its hard to explain”, which is distinguishable from the “didn't grasp the concept” theme as here students were explicit about their struggle to articulate themselves. This theme was based on interactions with three participants. Just suspicious. Also suspicious to me was the perception that several participants were using certain words and phrases as a coverall responses to questions they didn't fully comprehend: Interviewer: Do you think you changed from taking part in the music program? 7a11: Ahh, yeah. Interviewer: Yeah? In which way? 7a11: Ahh, being nicer, and respect. Initially it appeared that becoming “nicer” and more respectful were ways that this boy had changed due to the project. However, the pauses preceding each response, and use of the word “respect” interested me. It seemed “respect” was added almost as an afterthought, and in a way that was out of context with the question. I wondered if this was a term he remembered from the program, and repeated without full knowledge of its meaning. I decided to give him the benefit of the doubt until in a later response he made a similar statement: Interviewer: How did taking part in the [the music program] make you feel about your place in community or society more generally? 7a11: Ahh, people are being more friendly... Interviewer: Yeah? 7a11: Yeah, and more respect. Interviewer: When you say people, do you mean people all throughout society, or people in the class, or…? 7a11: Ahh, people all throughout the society. 153

At this stage it seemed to me “more friendly” and “respect” were being used as go-to coverall terms for tricky questions. This was again signified by the pauses, and the improper use of the term “respect”, which further reinforced my suspicion that he did not quite grasp its use. Further, the use of these terms in this context seemed incredible to me: he appeared to be explaining how people all throughout society had become more respectful and friendly as a result of his participation. While he may have been referring to a change in his perception – he now recognised people were respectful and friendly – this still didn't seem a valid outcome of the program. Due to my uncertainty surrounding this and other interactions that seemed guarded, simplistic, or doubtful, and based on intuition as much the words that were spoken, I decided to name this theme “just suspicious”. All participants contributed to this theme, although again, some feature more prominently than others. In total seven themes were identified in the “Emergent themes” category. A visual representation of how the raw data and codes were combined to create a theme in this category is presented in Table 7.7. For me, the uncertainties embedded throughout the above themes called into question many of the responses given by students. While these themes were not explicit reflections on the outcomes of the programs, the perceptions of the program, or the survey, I do feel this is important data for this study. Considering the level of uncertainty surrounding student understanding of all aspects of the program, and their ability to respond to questions, I argue any discussion of results necessarily needs an added layer of interpretation based on these points.

154

Table 7.7 Raw Data and Codes Used to Form the “Responding to What I Said” Theme Category

Emergent themes

Theme Code Data

Used my wording 7a01 Int: Did the music help that program do you think? 7a01: Yeah it was music too. Int: Ok, but how was the music part of it different to the [nonmusical] class? 7a01: Ummm…(long silence)…it was good.

Responding to what I said Just repeating 7a10 Int: Did [the music program] make you more confident or want to be at school more or class more? 7a10: Oh, I want to be in school more.

Just agreeing 7a10 Int: Do you think you changed as a result of the music program? 7a10: Ahhh, Yes. Int: In what way? 7a10 7a10: Umm…(long Int: How did it make you feel about being pause)…just yes. at school? Int: Ok, in a good way? 7a10: Umm…(long silnce)…I don’t know. 7a10: yeah, in a good way. Int: Did it make you want to be at school more or less…? 7a10 7a10: More. Int: How did it make you feel 7a01 about your place in community Int: How did the music 7a10 or society more generally? stuff […] make you feel Int: How did it make you feel about being 7a10: (Long silence – no about being in school? in class? answer) 7a01: Yeah, I like being 7a10: Oh, umm…(pause)…I wanted to be Int: Do you know what I mean in school. in class more. when I’m talking about Int: Yeah, but how did Int: Did it make a difference to how you community or society? the music make you felt about being with the people in your 7a10: Oh….nah. feel about being in class? Int: That's alright. So, sort of school? Did it make 7a10: Yeah. like your neighbourhood or… any difference? Int: In what way, did you want to be with 7a10: (Jumps in) Oh, yeah. 7a01: Mm (nods) them more or? Int: Did that make any 7a10: More. difference to that do you think? 7a10: Yes. 7a10 Int: Yeah? Can you tell me Int: Did it change how you felt about your why? classmates? 7a10: (Silence)…just yes. 7a10: Oh yeah. Int: How? 7a10 7a10: I don’t know. Int: Do u remember it [the Int: Did you get to know anyone better? survey]? Make friends? 7a10: Oh yeah. 7a10: I had more friends […] Int: What did you think of it? 7a10: (Long silence – no 7a10 answer) Int: Do u remember [the survey]? Int: Was it good, annoying? 7a10: Oh yeah. 7a10: It was good. Int: What did you think of it? 7a10: (Long silence) Int: Was it good, annoying? 7a10: It was good.

Teacher Qualitative Results Teacher interviews were analysed using the same method as student interviews, therefore only the creating themes step is presented. There were, however, some differences in this data. No interactions were coded to the “Emergent themes” category, and due to minimal of data, many themes relied on only one code. An overview of how codes were grouped to form each theme in each category is provided in Table 7.8.

155

Table 7.8 Keys to Success Teacher Interview: Codes, Themes and Categories Category Themes Codes

Category Themes

156

Codes

Psychosocial wellbeing outcomes Everyone became involved in the class It involved everyone

They achieved the Engaged in learning program goals She actually took They achieved in a short part time

It involved those I couldn't Kids felt alive during class They were waiting for it every week They were involved

Everyone's involved They felt like they were achieving something Everyone's They achieved what we motivated wanted them to They felt like they They learned it were achieving something

Social connection

Self-efficacy

It brings students closer together

They can feel they succeeded in something

She became part of the group

They can feel they are achieving something We did that

Codes

They learnt how to treat each other

Improved parentstudent relationship I opened up

Perceptions of the program It provided the environment It was the environment

Embedding music in Music as a vehicle to their daily lives achieve outcomes Embedding music in Kids love music their daily lives Anticipation Music for engagement

They owned the music The created the music They sung the song

Facilitation The spirit of the music therapist They way she was doing it

Music for teaching Achieving through music Category Themes

Prosocial behaviour

Outcomes are conditional Benefits can take time

Even one success is worth it Even one success is worth it

Need to match the program to the outcomes

Perceptions of the survey They didn't take it seriously They didn't take it seriously

Couldn’t engage with the topic

Results don't reflect the reality

They need to know who, what and why

Concepts were too abstract

Results were positive, not negative

The didn't know what it was for

It was the topic

Results don't reflect reality

They need to know who you are If you did it again it would be better They need to know why you’re doing it You need to spend time with them

Perhaps another method would be better Using different methods

Psychosocial wellbeing outcomes. Everyone became involved in the class. The most overwhelming outcome reported by the class teacher was that, through the music program, students became “involved” in the class. This was explained in terms of student’s anticipation for the musical KTS class; that the music program helped them engage with class activities; and that it made students feel “alive” during class. Further, he suggested students he had previously struggled to integrate into classes became involved in both classroom activities and as part of the student group. This was extended in that the program was able to involve “everyone”: Teacher:

Because everyone you can see is involved in the same way.

To me this suggested that, during the music program, all students became engaged or involved in class at a significant level, and were more “present” in the class. Because this “presence” extended to all students, this theme was named, “everyone became involved in the class”. Importantly, “class” here represents both the group of people and the activities in the classroom. This theme contained five codes formed from nine individual data extracts. Engaged in learning. Linked to this perceived outcome was the idea that students became more engaged in learning. Sometimes the teacher reported this was because the program motivated students to participate in learning activities, other times it was because it enabled them to experience a sense of achievement. Again, this was extended to the whole class: Teacher:

Everyone’s involved, everyone’s engaged, everyone’s motivated to do some work, to have some input in this work.

This theme was based on four themes containing 10 data extracts, and was named “engaged in learning”. They achieved the program goals. The teacher also reported, as a result of this engagement, students were accomplishing what they (the school) wanted them to achieve through the KTS program:

157

Teacher:

[they were] engaged, yeah, and that was happening, and we’ve seen it, because the students actually achieved what we wanted from them through music […]

This seemed to indicate that the teacher was confident the expected learning outcomes of the KTS program were achieved through the music program. Subsequently, I named this theme “they achieved the program goals”. It was based on four codes and five data extracts. Social connection. Another reported outcome was that students became closer. This was articulated in two ways: Teacher:

I think music, in my opinion, in this classroom, it makes people get closer together, because if we are sitting together and try to create a song, we share ideas together, we work together, you know?

This suggested to me the perception that the program brought the whole class closer together as a group. The second way this increased “closeness” or connection was articulated related to previously isolated students: Teacher:

[…] because she came from overseas […] and she was really isolated, but I noticed when [the music therapist] came and we had [the program], she became alive again, and she became involved, and actually she took a part in the acting video. So […] that was a good success itself, because it really make some people involved into that. And I have a chat with [the student] even before this program, and I told her, “Please, you really need to communicate, you need to get involved with the whole classroom”. She kept saying, “Yes, yes”, but in actual fact, the music made her really get into that group more, so that was good.

This comment indicated that the teacher felt the program not only helped this girl become “alive again”, and take part in the class activities, but also become more connected with “that group” – which I interpreted to mean classmates. As such, I named this theme “social connection”. It contained two codes and two data extracts.

158

Self-efficacy. Another outcome articulated by the teacher was his perception that the program built students’, sense of “success”, “motivation”, and the belief that they could achieve something: Teacher:

At the end of the day, when they created the song and they did the [video]taping, and then they watch it, they thought, “We did something”, and that’s what you want.

For me, what the teacher was describing here was an increase in self-efficacy, or the students’ self-belief they could “achieve” or “succeed” in something. Subsequently, these responses were placed under the theme “self-efficacy”, which was made up of three codes and five data extracts. Prosocial behaviour. The last two perceived program outcomes or benefits reported by the teacher were based on only one comment each, yet appeared to be significant for them. The first was that students became more aware of how to treat each other in a socially responsible way: Teacher:

I think I can say that the students walked out of this learning that bullying shouldn’t be in our school, and when we treat each other, there is a nice way of dealing with each other instead of using different things that will lead into bullying.

Here the teacher appears to be reporting his perception that students learnt prosocial values through the program. This theme was based on one code, and the above quote only. It was labeled “prosocial behaviour”. Improved teacher-student relationship. The last outcome reported by the teacher was the positive impact of the program on his relationship with his students: Teacher:

I’m always known as a very strict person in the classroom, and I don’t smile too much, I don’t laugh too much. I shouldn’t say that now, but that’s my personality. But after that when I stood there, singing with them, laughing, smiling, it was a beautiful environment [and] I think, that was good. I felt like sometimes,

159

you have to really open a little bit with the kids, you don’t have to be very serious all the time, because it makes a difference with the students, it does, when they see you smiling, and laughing and happy, that reflects on them all the time. This statement suggests the teacher saw himself opening up through the music program, which enabled him to interact with his students on a different level. Thus, he could better connect with students, make himself more available or approachable, and model positive affect and behaviour. This theme was also based on one code that contained the passage above. It was named, “improved teacher-student relationship”. Perceptions of the program. It provided the environment. The teacher reported several specific elements of the music program that he felt were instrumental in achieving the benefits he reported above. For example, when asked what it was about the program that helped engage and teach students about the principals of the KTS program, he replied: Teacher:

The environment, I think it gave the students more involvement into the task and into the work, so integrating that music into our daily life situation. That was good.

Here the teacher identifies that it was the “environment” the program created within the classroom that helped engage them. To capture this perception, this and similar comments were used to create the theme; “it provided the environment”. Embedding music in their daily lives. Also expressed within the above comment was the perception that embedding music within the daily experience of students was also beneficial. One example of how this worked was the anticipation the music program created: “the kids were waiting every week [for the music therapist to come and run the class]”. This suggests that making music a part of their regular routine was able to increase student engagement in class. This resulted in the theme “embedding music in their daily lives”, which included two codes and two data extracts.

160

Music as a vehicle to achieve outcomes. The teacher also reported that music was an effective vehicle for facilitating outcomes. This included learning: Teacher:

Because most of the kids they love music, they like it, so it can teach them something through music.

I interpreted this to mean that, because music was something all students could relate to, it was a successful medium for engaging them in a topic, and therefore useful for delivering curriculum content. The teacher made similar comments regarding the ability for music to get students to work together, create something from which they could gain a sense of achievement, and engage in class. For me these comments were articulating something slightly different from the “it provided the environment” and “embedding music in their daily lives” themes. Here it seemed the teacher was articulating that it was the music itself that was facilitating outcomes, whether it was the creative process of group songwriting, creating a finished product they could be proud of, or having their own role to play in a group process. In order to reflect all of these processes, I created the theme “music as a vehicle to achieve outcomes” from four codes, and 10 data extracts. They owned the music. A more specific manifestation of this use of music I felt warranted its own theme was related to the idea that students created and performed the music themselves: Teacher:

Kids had to think of a song, create a song and put it into music, that was great actually, part of creating the song, and also sing it.

What emerged from this comment for me was the perception that students were able to gain a sense of ownership through the program, both of the music and the musical processes. This sense of ownership was reported by the teacher to lead to benefits such as self-efficacy and engagement. The theme “they owned the music” was created to capture this impression. It was based on two codes from the above quote. Facilitation. Another program element the teacher reported as important was the facilitator: Teacher:

And actually I have to say a few comments about [the music therapist] as well, her spirit and the way she was doing in the

161

classroom, it was good, engaging the students, their movement, their talking and all that. The teacher referred to the value of the music therapist’s “spirit” multiple times, exemplified here in its perceived effect on students’ engagement, movement and communication. While I do not dispute the exceptionality of the music therapist’s spirit, for me the use of the term here represents the skills and training of the music therapist. Furthermore, “the way she was doing in the classroom,” suggests the way she used these skills and training to facilitate the program were considered an important element that made it successful for the teacher. This theme was named “facilitation” to capture both the idea that it was the way the program was facilitated, and the perceived expertise of the facilitator that were important. This was based on two codes, and two data excerpts. Outcomes are conditional. While the above responses suggest that the program was a resounding success, there were a few comments that put this success into context. Firstly, the teacher seemed to suggest that the benefits were not immediate: Teacher:

I noticed that at the beginning […] when [the music therapist] came the first time, there was a few kids they didn’t actually react too much to what she was saying in the classroom, it didn’t excite them much. But I noticed later on, those people, they were the first people to stand up and create some of the songs and sing it in the classroom, and I was really surprised.

This comment suggests that it took a while for some students to react to the music program. While the teacher goes on to suggest that it was the change in these unresponsive students that “really surprised” him, this suggests to me the importance of contextualising reported benefits: the value of musical processes can sometimes take time to manifest in an observable way, and it may be unreasonable to expect immediately visible results. The teacher also stated that, while he wanted to do more music in the classroom: Teacher:

We have to think of the purpose of the music just to achieve some other goal […] So you really need to look or think […]

162

what goal we really need to [achieve], what message through this music we want to send to our student. I interpreted this comment as an acknowledgement of the need to first consider the intended benefits of having music in a classroom, and how music is expected to help achieve them, before expecting outcomes. For me, this and the above comment indicated that, while this teacher saw obvious benefits of having a music program in his class, he also recognised these benefits were potentially conditional, or they might take time, or not occur for all students, if the purpose of the program is not clearly identified and matched with its activities. As such I placed these responses under the overall theme of “outcomes are conditional”, which included two codes and four data excerpts. Even one success is worth it. Having suggested that outcomes are conditional, the teacher went on to suggest that even a small success can make a whole program worthwhile. Again this related to the previously mentioned “isolated” girl: Teacher:

She was really isolated, but [through the program] she became alive again, and she became involved, and actually she took a part in the acting video. So that was, in my opinion, […] a good success itself.

I interpreted this to mean that, even if a program can only reach one person, it is worth the time and effort involved. I used the theme “even one success is worth it” to represent this impression, which was based on one code and data excerpt. Perceptions of the survey. They didn't take it seriously. The teacher had strong opinions about the survey, the first of which was offered with little prompting on my part: Teacher:

I have to be honest with you, I don’t think the students took [the survey] seriously.

This was a view that he repeated several times throughout the interview, suggesting to me he was convinced of this point. The theme, “they didn't take it seriously” was used to capture this perception, and was based on one code and three data excerpts. 163

Couldn't engage with the topic. When elaborating, the teacher suggested that, for some students in particular, it was the topic of the survey that meant did not take it seriously. Further: Teacher:

The students really didn’t take that seriously, because first, it's on paper, it’s nothing in reality so they can see it and touch it they can feel it.

For me, what the teacher is referring to here is the inherently abstract nature of concepts contained within the survey (such as wellbeing and community). Further, this meant students struggled to engage with questions, and relate them to their own experience. Consequently student responses were not necessarily representative of their lived experiences. I used the theme “couldn't engage with the topic” to represent this idea. It contained two codes and two data excerpts. Results don't reflect the reality. The idea that student responses did not reflect their lived experience was supported by the teacher’s reaction to the survey results: Teacher:

Well to be honest with you, I mean, this to me, this data, maybe it means something to you, but to me as a person who lives with the students long time during the day, I would say that this results, and this figure doesn’t really reflect, I’m sorry to say that, the reality one-hundred percent.

While it seemed obvious to me the teacher was trying not to offend me by rejecting the legitimacy of the results, it was also clear from several comments that he considered them completely invalid and unrepresentative of his own reality. This theme, named “results don't reflect the reality”, was based on two codes and three data excerpts. They need to know who, what and why. The teacher also went on to offer some potential explanations for the discrepancy between survey results and “reality”. Firstly, he suggested that if more time was spent with the students, things might be different: Teacher:

I’m sure now, if you do the same activity after spending one year with them, knowing […] who you are, and what you are doing, I think the response will be totally different.

164

The teacher seemed to be suggesting that the students needed to be sufficiently aware of who we were, what we were doing, and why we were doing it, in order to get valid results. Further, he seems to suggest that this would have required a reasonable time investment. I used the theme “they need to know who, what and why” to encapsulate these sentiments. This theme included five codes and five data excerpts. Perhaps another method would be better. The teacher provided further feedback regarding the suitability of using surveys as a data collection method in classrooms, this time at a more fundamental level: Teacher:

It [the results] doesn’t [reflect the reality], that’s for sure. But I’m wondering, as I said, if there is a way with […] your team, if there is a way this figure can be different or change, or work through some other methods to make a better result.

Firstly this comments seems to suggest the teacher was enthusiastic to find a way in which to gather data that better represents what he experienced as the outcomes of the program. What was most interesting here to me, however, was his last suggestion of using “other methods” to “make a better result”, which I interpreted to mean using something other than the survey we used, to collect more data that was more representative of what “happened”. To represent this last impression, I used the theme “perhaps another method would be better”. It was based on one code and one quote. Qualitative results summary. Interview data shows both students and the teacher perceived the program to have several identifiable psychosocial wellbeing benefits, the most notable of which were learning and school engagement. However, most benefits were reported by a small amount of students only, and all students reported an absence of at least one outcome targeted by the interview schedule. Further, many of the benefits reported by the teacher, such as improved student-teacher relationships, were in contradiction with the majority of student reports. For students, reported perceptions of the program included the idea it was enjoyable and a welcome distraction from normal classroom activities, yet there were mixed reports regarding the program’s purpose, and significance of its musical elements. Some students reported that the musical element helped them learn or connect with others, while others suggested there was no significant difference between the musical 165

and non-musical versions of the KTS subject. Student reports regarding the purpose of the music program also varied. While there was some mention of its role in teaching the content of the KTS subject, more often it was equated with purposes unintended by facilitators such as music education, just having fun. The teacher had a clearer idea of the program’s purpose and its link to outcomes, however again some outcomes he reported directly contradicted student responses. Two students reported they understood the survey process, four were open about not understanding some elements, and the rest offered suspiciously non-committal responses in this area. This suggested a general level of misapprehension regarding survey content and its purpose. Teacher responses support this conclusion. Further, while difficult to determine from the students’ own responses whether they took the survey seriously, the teacher was sure they did not, and went on to claim results were not reflective of reality. A significant amount of misapprehension, avoidance and otherwise suspicious responses and interactions were evident within student interviews. This highlighted the low comprehension levels of some students, and called into question the validity of some responses. Observational Reports Survey sessions. I made several observations during the collection of the above survey data. Both intervention and control group class teachers commented individually during respective survey sessions that the literacy and comprehension skills of several students meant completing surveys would be challenging for them. This was apparent to me during the sessions. At least a quarter of the 42 students admitted that they struggled with words and concepts included in the survey questions and responses, most of which sought regular guidance in interpreting these. This seemed partly due to issues of cultural and linguistic diversity – an observation consistent with the school principal’s claim that over 40 different cultures were represented within the school community, and that as a whole, students were “significantly from non-English speaking backgrounds”. 7

7

Quote taken from interview with school principal collected as part of the larger

Music Matters project explained in Chapter 6 166

Both teachers also commented that conducting the surveys as a whole class activity meant some students would “play up”. These comments were consistent with my observations during these sessions. There was vocal discussion of questions and sharing of answers between about a quarter of students, and approximately five students in each class seemed to lose interest throughout sessions, either answering without reading the questions or disengaging completely. Interview sessions. The limited vocabulary and comprehension of students was also evident during interviews. Despite prompts, and efforts to explain difficult terms and ideas on my part, students often offered one-word answers, seemingly due to their struggle to understand concepts and questions addressed in the interview schedule. This was particularly evident for one student, whose interview lasted for only five minutes (about the time it took for me to ask all of my questions and prompts). Others appeared to be shy, or just uninterested. This was not the case with all participants – some were relatively engaged, and articulate in their responses – however, I feel it significantly limited what could be expected from the data.

167

Chapter 8: Bereavement Group Results This chapter reports on the results of a small school-based study called the Bereavement Group (BG). This study was delivered as part of the larger pre-existing Music Matters research project (described in previous chapters). The Bereavement Group (BG) study investigated both the benefits and implementation issues of a small music therapy program (also named Bereavement Group) that aimed to assist students experiencing loss and bereavement. It targeted psychosocial wellbeing outcomes and was conducted in a semi-private setting during school hours by a music therapist. Quantitative and qualitative data were collected concurrently in order to assess the program’s impact on psychosocial wellbeing, explore barriers to engagement, and evaluate the program. Observational reports were also documented in relation to research activities. Results not only provide insight into the benefits and features of the BG music program, but also enable comparison to KTS study results, thus offering important information for informing the critical reflection analyses. Methodological Overview A similar survey to that used in the KTS study was used. Baseline and follow-up measurements were taken, and mean change scores were calculated to quantitatively assess the effect of the program on areas of student psychosocial wellbeing. Confidence intervals and p values were calculated to estimate the statistical significance of change scores (see Chapter 6 for methods used). Interviews were conducted concurrently with follow-up surveys. These explored participant perceptions of the program, including: perceived benefits; how the program facilitated these benefits; aspects of the program they found valuable; and program elements they would change. Interviews were analysed using a combined deductive and inductive approach. Codes were inductively identified, allocated to three predetermined categories, and used to create subsequent themes (see Chapter 6). Quantitative and qualitative analyses are presented sequentially, followed by observational reports. Each set of results is presented in a style consistent with the epistemological nature of the research and analysis methods used. Quantitative results are reported objectively, while interview and observation results are reported using the first-person to reflect the process of interpretation involved in the collection and analysis of data. Discussion of similarities or inconsistencies between data sets is not 168

presented here, but rather reserved for the critical reflection analyses (Chapters 9 and 10), recommendations (Chapter 11) and discussion (Chapter 12) chapters. Quantitative Results Students were asked to complete a survey instrument containing 14 scales covering risk and protective factors across six domains of student psychosocial wellbeing. All five participants completed both baseline and follow-up surveys, yet there was missing data from one participant at follow-up. Subsequently, mean change scores for some scales are based on four participants only. See Appendix O for all scale results. School domain. The wide confidence intervals shown in Figure 8.1 indicate reported baseline scores for most measures in the School domain contained notable variation. This variation is further demonstrated in baseline scores for the Educational expectation scale in Figure 8.2, where variation across the scale rendered the use of confidence intervals uninformative, and raw data is presented instead. Figure 8.2 shows reported responses from three of the five students placed them at the top of the scale, with a fourth only one point below. While the small sample size precludes the conclusion that the fifth participant (who reported a score of one) is an outlier, the grouping of scores towards the high end of this scale suggests a potential ceiling effect. A potential ceiling effect is also evident for the Opportunities for prosocial involvement scale, especially given its reduced confidence intervals. Figure 8.3 reported mean change scores were mixed. For risk factors, a mean score of -0.20, p = .57, 95% CI [-1.10, 0.70] was reported for Academic failure, suggesting the sample improved. However, for the Low commitment to school scale, a mean score of 0.27, p = .07, 95% CI [-0.04, 0.57] indicated a worse score at follow-up.

169

A cademic failure

O pportunitities for prosocial inv ov lement

Rew ards for prosocial inv olv ement 1

2

3

4

Low commitment to school

Risk factor: red line Protectiv e factor: blue line

1

2

3 Mean baseline scores (w ith 95% C I)

4

Figure 8.1 Mean scores at baseline and confidence intervals for School domain

Educational expectation baseline scores

1

2

3

4

5

6

Figure 8.2 Raw baseline scores for Educational expectation scale For protective factor scales, a positive mean change score of 0.04, p = .86, 95% CI [-0.55, 0.63] was reported for Opportunities for prosocial involvement, again suggesting improvement. However, mean change scores dropped for both Rewards for prosocial involvement (-0.10, p = .72, 95% CI [-0.81, 0.61]; see Figure 8.3) and Educational expectation (-1.60, p = .23, 95% CI [-4.72, 1.52]; see Figure 8.4).

170

5

Academic failure

Opportunitities for prosocial invovlement

Rewards for prosocial invovlement Risk factor: red line Protective factor: blue line

-4

-3

2 0 1 -2 -1 Mean change scores (with 95% CI)

3

4

Figure 8.3 Mean change scores and confidence intervals for Academic failure, opportunities for prosocial involvement and rewards for prosocial behaviour While results indicate the group was worse overall at follow-up for this domain, mean change scores were minimal given scale ranges. The only scale to indicate a sizable mean change score was Educational expectation. However, given the potential ceiling effect noted in baseline data, this may represent regression to the mean. Overall, the absence of statistical significance in all of the above mean change scores prevents a conclusion as to the effect of the music program on this domain.

171

Low commitment to school

-5

-4

-3

-2

-1

0

1

2

3

4

5

Educational expectaion

-6

-5

-4

-3

-2 -1 0 1 2 M ean change scores (w ith 95% C I)

3

4

5

6

Figure 8.4 Mean change scores and confidence intervals for Low commitment to school and Educational expectation Emotional control domain. As Figure 8.5 shows, the mean reported baseline score suggests this group reported relatively high emotional control at baseline. However, the large confidence interval suggests this did not represent potential for ceiling effect. An average decrease was reported at follow-up, with a mean reported change of -0.15, p = .65, 95% CI [-0.99, 0.69]. As a protective factor, this indicates the group reported reduced emotional control after the program. However, this mean change was not statistically significant, and as shown in Figure 8.6, minimal in comparison to the scale range.

172

Emotional control

1.0

1.5

2.0 2.5 3.0 Mean baseline score (with 95% CI)

3.5

4.0

Figure 8.5 Mean baseline score with confidence intervals for Emotional control

Emotional control

-4

-3

-2

-1 0 1 2 Mean change score (with 95% CI)

3

4

Figure 8.6 Mean change score with confidence intervals for Emotional control

173

Social support domain.

MSPSS-SO baseline scores

1

2

3

4

5

6

7

6

7

Trusted adult in life baseline scores

1

2

3

4

5

Figure 8.7 Raw baseline scores for Social support domain As Figure 8.7 shows, reported baseline scores in this domain contained notable variation. This combined with the small sample size meant reporting mean scores and confidence intervals were uninformative, thus raw scores are presented. Despite the variation, data was skewed left suggesting that on average, the group’s reported social support was relatively high at baseline.

MSPSS-SO

Trusted adult in life

-7

-5

-3 -1 1 3 0 Mean change scores (with 95% CI)

5

7

Figure 8.8 Mean change scores and confidence intervals for Social support domain Both scales showed the same decline from baseline to follow-up, with mean 174

change scores of -1.60, p = .24, 95% CI [-4.85, 1.65] and -1.60, p = .33, 95% CI [-5.58, 2.38] for the MSPSS-SO and Trusted adult in life scales respectively. As protective factors, this reduced mean score suggests a decrease in social support at follow-up, especially given the potential score range for both scales was one to seven, (see Figure 8.8). However, results were not statistically significant, prohibiting a conclusion that the program had an effect on students’ reported level of social support. Psychological distress domain.

K10

10

15

20

25 30 35 Mean baseline score (w ith 95% C I)

40

45

50

-40

-30

-20

-10 0 10 Mean change score (w ith 95% C I)

20

30

40

K10

Figure 8.9 Mean baseline, and mean change score with confidence intervals for Psychological distress domain The group improved its mean score in the Psychological distress domain with an average change score of -1.50, p = .68, 95% CI [-12.01, 9.01] for the K10 scale. As a risk factor, this suggests lower reported psychological distress at follow-up. However, absence of statistical significance and small magnitude of mean change compared with the scale range (shown Figure 8.9) prevent this conclusion.

175

Psychological wellbeing domain. As Figure 8.10 shows, reported baseline scores for this domain contain notable variation. A mean change score of 9.50, p = .22, 95% CI [-10.23, 29.23] from baseline to follow-up indicates an average improvement on this scale of 13.57%, which is notable given the scale range of 14 to 84 (70 points). While this suggests a promising trend, the mean change score is based on data from only four participants and not statistically significant, prohibiting a conclusion the music program led to increased psychological wellbeing.

MHI-PWB

14

20

30

40 50 60 Mean baseline score (w ith 95% C I)

70

80

84

MHI-PWB

-70

-60

-40

-20 0 20 Mean change score (w ith 95% C I)

40

60

Figure 8.10 Mean baseline, and mean change score with confidence intervals for Psychological wellbeing domain

176

70

Psychological needs domain.

BPN-SF (w hole score)

10

15

20

25

30 35 40 45 Mean baseline score (w ith 95% C I)

50

55

60

63

20

21

BPN-SF (autonomy )

BPN-SF (competence)

BPN-SF (relatedness) 3

4

6

8

10 12 14 Mean baseline scores (w ith 95% C I)

16

18

Figure 8.11 Mean scores at baseline and confidence intervals for Psychological needs domain For this domain, notable variation was reported across all scales at baseline. This is particularly notable given scale ranges (see Figure 8.11). As Figure 8.12 shows, reported mean change scores for all scales showed a decrease from baseline to followup. This was most evident in the overall BPN-SF scale, with a reported mean change of -5.25, p = .60, 95% CI [-33.40, 22.90]: a drop of 10.10% on a scale range of nine to sixty-three. Decreases were also evident in subscales, with mean change scores of -2.00, p = .54, 95% CI [-10.24, 6.24], -2.25, p = .51, 95% CI [-11.74, 7.24] and -1.25, p = .62, 95% CI [-8.53, 6.03] for autonomy, competence, and relatedness scales respectively. While a decrease in mean change scores for all scales may suggest the sample reported lower satisfaction of psychological needs after the music program, there was high variability among reported scores, and a lack of statistical significance. Also, mean change scores (excluding the autonomy subscale) were based on only four participant

177

responses. Therefore, it is not possible to make any conclusions as to the impact of the music program in this domain.

BPN-SF (w hole score)

-54 -50 -45 -40 -35 -30 -25 -20 -15 -10

-5

0

5

10

15

Mean change score (w ith 95% C I)

20

25

30

35

40

45

50 54

BPN-SF (autonomy )

BPN-SF (competence)

BPN-SF (relatedness) -17

-15

-10

-5 0 5 Mean change scores (w ith 95% C I)

10

15

Figure 8.12 Mean change scores and confidence intervals for Psychological needs Quantitative results summary. Results suggest that the sample size was too small to make conclusions regarding the impact of the music program on domains of psychosocial wellbeing covered in the survey. This is evidenced by a lack of statistical significance in mean change scores, something likely compounded by missing data. Further, many reported mean change scores were of low magnitude when compared to scale ranges, with poor precision as indicated by wide confidence intervals. Further, while reported means for baseline data suggest ceiling and floor effects were not present, raw data plotted in Figure 8.2 and Figure 8.7 indicate potential for these effects in some cases. Qualitative Results Interview data was analysed using the same three-step inductive-deductive approach used in the KTS study (see Chapter 6). There was, however, one slight 178

17

alteration in that the category template did not include the “Perceptions of the program” category (for explanation, see Chapter 6). Step 1: creating category templates. The category template used in this study to organise interview data consisted of three categories. These included “Psychosocial wellbeing outcomes”, “Perceptions of the program”, and “Emergent themes”. Step 2: coding interviews. Interview transcripts were coded one-by-one using the same coding protocol described in the KTS project, including the use of “interactions” to encompass the context surrounding participant responses (see “Step 2: coding student interviews”, Chapter 7). Once identified, codes were assigned to categories listed above for further analysis. One example of a participant response (from BG2), its coding, and subsequent allocation to the “Psychosocial wellbeing outcomes” category is as follows: Interviewer: How did you find the songwriting process? BG2: It kinda cheered me up, because, on that day that we did it, I found out more information about my dad, and it made me really upset, but then when we had the group, it made me like feel it was ok, because it’s ok to cry, and all that stuff. Here, the student describes that attending a music program session helped her cope with difficult feelings. As she explained, the student became distressed upon receiving new information relating to the loss of her father. This distress seemed to be somewhat mediated by attending the group, which appeared to help her accept and express these feelings without feeling bad about it. This whole interaction was coded as “coping” as I felt the context of the response, including the description of events, provides a richer account of how the program was seen to help her to cope with experiences of loss. An example of data coded into the “Perceptions of the program” category is taken here from student BG4: BG4: […] being involved – everybody had a chance to get involved, it wasn’t always one person getting chosen, it was always, everybody had their option to have their say in whatever it was.

179

Here the student articulated several elements of the program she believed were important for her. She suggested it was a cohesive group, and then went on to explain that everyone was given the option to participate and have a say. This suggested to me that the participatory and inclusive nature of the program were what made it work for her. This interaction was coded as “everyone had a say”. Because the participant offered this response in relation to why she thought the program was “pretty good”, I allocated it to the “Perceptions of the program” category. One example of data allocated to the “Emergent themes” category is a comment made by participant BG4: Interviewer: And [having participated in the music group] what would you say has been the most challenging? BG4: It all went pretty well, but, the only time when I was really confused, was like answering all those questions in the, the question things that we got in the start and the finish, some of the questions were basically repeated, and some of them I didn’t really understand. Here, instead of identifying an aspect of the music program as challenging, BG4 suggests the questions and the purpose of the survey did not make sense to her. This piece of data was coded as “I didn’t understand the survey”, and allocated to the “Emergent themes” category. This allocation was based on my perception that, while these interviews did not seek information regarding the survey process, it seemed important to the participant, and was therefore considered relevant to study findings. Step 3: creating themes. The next step in the analysis process was to create themes within each category. Where codes within a category followed a similar patter, or related to a similar topic or idea, I grouped these together into themes. A visual presentation of which codes and themes were allocated to different categories is represented in Table 8.1. Psychosocial wellbeing outcomes. Social connection. Codes within this category suggested participants experienced a number of psychosocial wellbeing outcomes as a result of the program. All five participants indicated that through participation, they got to “know the other people even better” in 180

their group. For many this was through getting to “know each others stories” about grief and loss, knowing “how others were feeling” and that they “had stories that were sort of like mine”. For some this meant they became “better friends”, while for others it meant they could learn more about how their friends were “coping with it [grief and loss]”. These responses suggested to me that, through the program, participants were able to connect with each other on a deeper level. Therefore, codes articulating this connection were grouped into the theme “social connection”. Further, it appeared this connection was not restricted to within program boundaries, and often extended into their social world. The process in which these students became closer, and the outcomes of this new level of connection is described in the following response by BG2: BG2: One girl is in my class, and another girl in another class and we’re pretty close, and I didn’t know that she lost some of her friends, and then I finally found that out. Then another boy, he lost his mum, and I knew that, but I didn’t know how he was coping with it, and then I finally realised that he was coping with it kinda bad, and he wouldn’t talk to anyone, and when he found out that I lost my dad, me and him just talked a lot about happy stuff.

181

Table 8.1 Bereavement Group Student Interviews: Codes, Themes and Category Category

Psychosocial wellbeing outcomes Dealing with I can remember the Social Connection Social Support Letting it out Moving on Affect Conditional outcomes bereavement good things Codes We know each other Someone to talk Letting someone Dealing with bad news Connecting with old Moving on Makes you feel I'm still not coping with it, better now to know memories better/happy but it’s helped I know my friends are Not alone It's ok to cry Helps me get through the Remembering the good He's not in I understand my It’s made me at least a bit struggling too anymore day things pain anymore feelings better Sharing stories Helping others Letting the Helps me sleep Focusing on the good The next Mood Its not over feelings out parts stage regulation I know I'm not alone How they deal Building confidence Negotiating good and I'm less angry Some parts were hard, but with it bad memories now worth it Got me back on track Helps me deal with problems Category Perceptions of the program Themes Change in musical Gave us the place and How music helped Musicality not required Private group setting We also had fun Hard but worth it relationship space Codes Music can make you I don't like drums, but… Now I know how music can Chance to connect with Got to know each other Its been fun Talking about feelings feel better help me others better was difficult Music for expression I'd never played drums I appreciate music more Opportunity for Because they're like me Instrumental work Sharing songs can be and sharing before now expression was fun hard Themes

182

Music can help you deal with things

I listen to music more now

Because it’s confidential

Music provides a safe setting The music was fun

I understand music better now

Everybody could have their say

Category Themes The survey was confusing Codes Didn't understand the survey

Trying not to make Some parts were hard, it all sad but worth it

Emergent themes We missed class, but it was worth it It was just maths It was easy to catch up I wanted to go anyway

Maybe next time More people would be good Wasn't long enough

I’d recommend it to anyone I'd recommend it Good the way it was It met my expectations

This passage suggests sharing stories within group sessions allowed this girl get to know friends she already considered “close”, at a deeper level. It also described for me how she formed a new connection with another boy, whom she knew, but did not know was “coping (…) kinda badly” with the loss of his mother. Social support. Four out of the five participants said participation in the group meant they all had someone to “talk to” about their experiences of grief and loss. For some this was about seeking “advice” from someone who had been through something similar, while for others it was about “expressing” feelings to someone who understood, “knowing” someone was there if they needed, or “helping” and being there for others. I felt they were explaining a social support network that, while facilitated by the program, now existed outside session boundaries. I grouped these responses into a theme called “social support”. I felt “social” represented the fact the support was coming from the (school) community, instead of being institutional or professional. Further, “support” was used to represent my feeling that the interaction described went beyond mere friendship and extended to become a resource for their psychosocial wellbeing. For BG5, this was articulated as follows: BG5: Whenever you have like a bad problem, you always know that you have someone to go to to express them, and there is always someone to tell. Interviewer: […] has it helped you to deal with stuff? BG5: Yeah, knowing that there is always somebody there to help you out. This participant also noted: BG5: It was embarrassing to talk about the story, I didn’t have much to say, it was just mostly them, I didn’t stand up and say what happened. Interviewer: Ok, did you find that session helpful though? BG5: Ahh, yeah it was good, it just told us you always have someone to express your feelings to, and your problems. This comment suggested “social support” was a resource not only available in sessions, but something also accessible outside, even if not used right away. This 183

reinforced my perception that this support existed in their social world, and extended past group session boundaries. Letting it out. While BG5 suggested they did not “have much to say” in the group, for some participants, “letting it out” appeared to be a significant outcome. Two participants specifically referenced the fact that sessions gave them the opportunity to relate how they were feeling. This was articulated as both feeling “relieved that I could tell someone what I was going through”, and being able to “get my feelings out”. While there is overlap here with the “social support” theme, I assigned these comments to a new theme based on a distinction between “knowing” and “expressing”. While the “social support” theme related to descriptions of “knowing” they had someone to “talk to”, these comments appeared as if participants were focusing on the act of expressing, whether it be through talking, crying, sharing a song, writing lyrics, or playing an instrument. One example of this expression is given by BG4: Interviewer: Would you say you have changed since you participated in the program? BG4: Yep, I’m a lot happier after having so many people I can talk to, and letting all those feelings out. Interviewer: Do you think music played a big role in that? BG4: Yep, when we were writing the song, I got to say a lot about how I was feeling, and I got to hear how others were feeling, and how they found a way to move forward, and sort of get happier in life. I interpreted this to mean she had been keeping her feelings of grief and loss to herself, and the act of “letting [them] out”, or giving them voice through group discussion and song-writing, facilitated a process of venting which possessed its own cathartic properties. Further, it seemed to me that the expression of these internalised thoughts and feelings brought them out into a space where they could be acknowledged and dealt with, and left behind. Finally, that it was “letting all those feelings out” that made this girl feel “a lot happier” suggested to me the “act” of expression was important for her, and in itself an outcome of the program important for her psychosocial wellbeing. This theme was named “letting it out” to reflect that this was an active process in which internalised thoughts and feelings were externalised.

184

Dealing with bereavement. Participants also reported that the program helped them to “deal” with feelings of loss and grief. While these comments often overlapped with descriptions of how the program helped participants in other ways, and which are described in other themes reported here, I made the decision to attribute them to a separate theme. This was based on my perception that, regardless of the process or outcome involved, these responses related specifically to building either resilience or coping skills to deal with bereavement. Comments from four of the five participants informed this theme, which I named “dealing with bereavement”. This included generalised statements, such as the acknowledgement that the program had “really helped”, or “helped a lot” when dealing with bereavement. For others, such as BG1, the description was more explicit: Interviewer: Can you just tell me a little bit about the Bereavement Group [?] BG1: Well, there’s six of us in a group and we sit down and talk about our loss and how to deal with our loss, and compare how we felt through it […] and we just do like, small activities to help us through it. Interviewer: Ok, overall, how did you find it? What was your experience like? BG1: Ahh, sort of helpful, like it’s helped me, both me and my sister through losing our mum. While still a fairly simple statement, what made this meaningful for me was that it was offered in response to a general question at the start of the interview. Firstly, the immediate description of the program as something directed towards dealing with loss signified to me that this girl clearly identified with this aspect of the program. Further, the second response suggests to me the experience of the program that was most powerful for her was that it helped her and her sister deal with losing their mother. This suggests a primary outcome of this program was coping with loss. Others, such as BG3, described this coping in terms of the music helping him “get back on track”, while BG2 reported it as follows: BG2: I listen to music more now [it] keeps me calm when I’m really upset, or angry, it just keeps me relaxed. It helps me go to sleep some nights when I can’t go to sleep. It just helps me through, like, the day.

185

Interviewer: [And] have you changed since participating in the program, do you think? BG2: I’ve become less angry with everyone, and it’s helped build my confidence up a bit [about my dad]. I felt these responses show that through her participation in the program, and the way she has learnt to appropriate music, BG2 has acquired coping skills beneficial to a number of areas in her life which have been affected by the loss of her father. I can remember the good things. Another outcome, which appeared to be related to coping, emerged through a number of unsolicited comments that made reference to the act of reflecting on loss. Such comments were made by three of the five participants, and mostly focused on the idea that the program had taught them how to “think about the happy parts” of the relationships with those they had lost. For some this was through songs or musical memories, while for others it was knowing they could “think of that person in happy ways, not sad ways”. Participant BG2 described it like this: Interviewer: How did you find [the program]? BG2: Kinda helpful, coz I was always upset about my dad passing away, and finally I have realised that like I don’t have to cry all about the sad parts, I can just be happy that he’s not in pain anymore, and I can think about happy parts. I interpreted this to mean participation in the program had led to this participant’s self-realisation that she could focus on happy feelings. For me this indicated she had learnt to negotiate between good and bad feelings related to the memory of her father. This suggests she had reached a new level of maturity in which she was able to, at least to some degree, manage her own emotions – a resource closely linked to the management of psychosocial wellbeing. This link to psychosocial wellbeing is further evidenced in the above comment, “I can just be happy that he’s not in pain”. I took this as an acknowledgement that, not only can she focus on the positive things, but she can also gain strength from them. Such comments were placed under the theme “I can remember the good things”. “The good things” aims to represent the fact participants could identify good feelings associated with the memory of those they had

186

lost, while the “I can” signifies the element of self-efficacy required for this acknowledgement. Moving on. A pattern also emerged relating to letting go of painful memories, and progressing to the next stage of the bereavement process. This was reported by two of the five participants; sometimes in terms of being able to “finally move on” themselves, and sometimes in terms of being able to seek advice and hear about other people’s strategies for, or experiences of moving on. For participant BG4, this idea appeared to figure prominently in her experience of the music program: Interviewer: Ok, cool. And, having participated, what would you say has been the best bit? BG2: Probably making friends with everybody, and finding new ways to, sort of, move forward. While only a brief statement, this girl made reference to finding new ways to move forward several times throughout the interview. That this girl identified finding new ways of moving “forward” as one of the most valuable parts of the program further indicates its importance. It also suggests that she was ready to move on to the next stage of her life – one that wasn’t dominated by the loss of her father. These response were grouped under the theme “moving on”, as this is exactly what these students were describing. Affect. Participants also described the how the program had affected their mood or feelings. For some, it was the program itself that made them feel “a lot happier”, or “become less angry”. For others, it was the way they had come to use music through participating in the program that helped them to stay “relaxed”, get in the “mood”, feel “better”, or feel “happier”. This indicated that, whether it was directly related to the music or not, the program had an identifiable impact on their affect. That all five participants reported this impact suggested to me it was a significant outcome. These codes were grouped together to form a theme named “affect”. This name aimed to capture the range of emotions, moods, and feelings participants reported, and the fact they “changed” as result of their participation. Participant BG2 explained it as follows:

187

Interviewer: Has your relationship with music changed since you became involved in the program? BG2: Its become more like…I listen to music more now […] it keeps me calm when I’m really upset, or angry, it just keeps me relaxed. Interviewer: Ok, cool. So would you say you have changed since you participated in the program? BG2: I’ve become less angry with everyone. Conditional outcomes. While all participants reported that the program did influence their affect in a positive way, three participants suggested that this impact was not absolute. Two suggested participation did not make the “pain” or bad feelings disappear. Others, such as BG5, admitted they were still not “coping” well, and qualified reported changes in affect with the acknowledgment they were sometimes only partial: Interviewer: Would you say your relationship with music has changed since you became involved in the program? BG5: Yeah, sort of. The songs like, they change your feelings, when you’re feeling down, you listen to music and you feel at least a bit better. Interviewer: Was this the same before you started the program? BG5: No, when we started the program. Interviewer: So, since you began the program you have been using [music] for that? BG5: Yeah. Here BG5 acknowledges that, since participating in the project, he had started to use music to cope with “feeling down”. However, this is qualified with the statement that it made him “feel at least a bit better”, suggesting the change in affect was not complete; that some painful feelings remained. Statements such as these seemed important to me in that, although changes in affect were reported, to assume the program had removed all feelings of grief and loss for these students would be unrealistic. I interpreted these comments to mean that, while psychosocial wellbeing outcomes from musical participation are possible, we must be careful in assuming such

188

outcomes will be absolute. As such, these and similar comments were grouped under the theme “conditional outcomes”. In total, eight themes were identified in the “Psychosocial wellbeing outcomes” category, all of which have been summarised here. An example of how the raw data and codes were combined to create themes in this category is presented in Table 8.2. Table 8.2 Raw Data and Codes Used to Form the “Social Support” Theme Category Theme Code Data

Someone to talk to

Psychosocial wellbeing outcomes Social support Not alone anymore Helping others

BG1: [..] there’s people I can BG1: […] because I talk to if I need to. know I’m not the only one going BG2: I was […] relieved that through it I could tell someone what I BG5: […] Because was going through normally I wouldn’t BG2: Just talking to someone have anyone to speak about what’s happened in my to about my feelings life BG4: I know that I can talk to them and they’ll understand BG4: I’m a lot happier after having so many people I can talk to BG5: It was good, it just told us you always have someone to express your feelings to, and your problems. BG5: Whenever you have like a bad problem, you always know that you have someone to go to to express them, and there is always someone to tell.

BG1: Coz you’re helping others out while your helping yourself out at the same time.

How they deal with it BG2: […] just to see if they have any advice for me to cope with it better.

BG1: […] coz my sister has trouble talking to people, so it was sorta to help her as well to, so she knew that she had other people there to talk to.

BG4: To sort of find out how others were feeling after losing someone, and how long they’ve been dealing with it, and if I needed some help I BG2: […] to see about could go to them and my friends, what they ask them […] is it were in there for, see if I going to get easier, could help them with any […] am I going to move forward soon or advice. whatever? BG2: […] he was coping with it kinda bad, and he wouldn’t talk to anyone, and when he found out that I lost my dad, me and him just talked a lot about happy stuff.

BG5: […] knowing that there is always somebody there to help you out.

Perceptions of the program. Codes within this category revealed several program elements that were perceived by participants to assist in realising the psychosocial wellbeing outcomes identified above. This included elements that made the program enjoyable, some that were practical, some that provided them with the space in which to deal with their grief and loss, and some that were sometimes difficult but “worth it” in the end.

189

How music helped. Perhaps the most obvious element of the program noted by participants was its focus on music. I say obvious both because it was a music therapy program, and because all participants made repeated references to this element. Some participants noted music activities helped them “tell” others how they were “feeling”, “share” stories, or just have “fun” during sessions. Others noted the program helped them to learn how to use music to help them in their everyday lives, and to understand how this use of music helped achieve several psychosocial wellbeing outcomes noted above. Participant BG4 articulated this as follows: Interviewer: Do you think you have changed since participating in the program? BG4: Yep, I’m a lot happier after having so many people I can talk to, and letting all those feelings out. Interviewer: Do you think music played a big role in that? BG4: Yep, when we were writing the song, I got to say a lot about how I was feeling, and I got to hear how others were feeling, and how they found a way to move forward, and sort of get happier in life. My interpretation of these statements was that participation had resulted in some significant outcomes for this girl, typified here as being “a lot happier”, which I related to the “affect” outcome theme above. Further, she was able to clearly articulate music’s role in achieving these outcomes. The fact all five participants had data coded into this area signified music was a beneficial element of the program. These codes were placed under the theme “how music helped”. Musicality not required. Comments from two participants indicated musicality was not required for them to benefit from musical activities. For one participant, while playing instruments was something that made the program enjoyable for them, it did not matter that they had never done it before. The other, BG3, explained it like this: BG3: It was really fun…I really don’t like drums, but because we did different beats, it was enjoyable. It helped with the program because it got me back on track.

190

For me this suggested that the program was facilitated in a way that made the benefits of musical participation accessible to everyone, regardless of their musical background. I considered this an important element of the program when considering its relationship to wellbeing, as it signified interest in music and musical proficiency were not barriers to participation or outcomes. As such, this theme was named “musicality not required”. Change in musical relationship. All five participants noted their relationship with music had changed as a result of the program. For some I identified this new relationship as a new appreciation for musical aesthetics, such as the “different beats of the songs”. For others it appeared to be a new understanding of how music could help with expression, coping, or affect. Four of five participants also said they “listen to more music now”, and some explained how music had taken on a new “meaning” for them. This idea of meaning was described by BG4: BG4: It was different to find out different ways of playing different feelings, I never thought that different sounds could mean different things, I used to just think they were sounds, but now I know that they can mean different things. Interviewer: Did that help? BG4: Yeah. Interviewer: In what way? BG4: It sort of like, if I wanted to get my feelings out, I could play the drums and play the feeling that I’m feeling This passage seemed to articulate how the program helped this girl gain a new appreciation of music. Not only did music take on a new meaning, she had learnt to appropriate it in way that helped her with things such as expression. This indicated an important element of the program for this girl was that it had taught her more about the inherent qualities of music that can be appropriated for therapeutic benefits. While this example refers specifically to the “meaning” of musical elements, a common pattern in these responses was that they described how their use and or understanding of music had changed. This theme was named “change in musical relationship”, with “musical relationship” used to capture the range of interactions with music that participants

191

described. “Change” was used to represent the fact participants reported this relationship to be different as a result of their participation. Gave us the place and space. Participants also identified non-musical program elements that were considered beneficial. For three participants, this related to the practical issue of being provided the physical space, and time, to focus on themselves and their experiences of bereavement. For some this was because it gave them a forum to voice their feelings of loss and grief, while for others it provided them with the chance to connect with others who were going through similar experiences. This was articulated for me in BG1’s following responses: BG1: […] it was interesting to hear how everyone else, what happened with everyone else, and who they lost, and how. Interviewer: And did you feel like this was a good setting, or a good opportunity for you to be able to talk about that kind of thing? BG1: It was coz then I know because I know I’m not the only one going through it, and there’s people I can talk to if I need to. This indicated that this girl appreciated the opportunity to hear others’ stories of bereavement, and meant she knew she was not alone in her own experiences and had a support network there if she needed it. Most importantly for this theme, however, was her acknowledgment that the program provided both the space and opportunity for these things to take place. For this, reason, I named this theme “gave us the place and space”. Private group setting. Another element related to space which all participants noted as positive was the private group setting. Some participants noted this setting worked because they were all there for the same thing, and that they knew each other, were “close”, and could “trust” each other with their stories and problems. Some also indicated they got to know each other “better” through the group, and were able to learn what their schoolmates were going through. Again, this figured prominently for participant BG1, who suggested this represented the element of the program she appreciated most: Interviewer: So, having participated, what would you say has been the best bit for you?

192

BG1: Just listening to each other and knowing that it’s confidential, they won’t go out and blab or laugh at you because you cried about it, and we’ve all been through the same thing. This suggested the participant saw the group as having a unified focus that, when given the safety of a confidential setting, meant they could open up and express themselves without fear of reprisal or ridicule outside the group. As such, codes that expressed an appreciation of the small confidential nature of the group were grouped under the theme “private group setting”. We also had fun. Another element of the program that appeared beneficial for participants was that sessions were sometimes “enjoyable” or “funny”. While some suggested the program was “fun” overall, this was mostly articulated in relation to the use of musical instruments: Interviewer: Now, the next session, I believe you guys got to use instruments… BG2: Yeah, we used drums. Interviewer: Yeah? And how was that? BG2: It was kinda funny, because everyone was doing different drumbeats. We tried to do like words; like “sad” we did on the drums, and then “confused” and “upset” and “depressed”, and all that, like feeling words on the drums. And it was really funny because, now I know how to play it. This response suggested that, even though these students were dealing with serious issues, the program was designed and facilitated in a way that sometimes also made it enjoyable. I do not interpret this to suggest the program trivialised their experiences of bereavement. Rather, I believe responses such as this suggest the program was able to address these experiences in ways that made them accessible without always being confronting. For me, the idea that the program was not always serious seemed to have an important impact on students, as it provided opportunities for fun, and perhaps even relief from experiences of loss and grief. Further, this felt to me as if it made the program accessible for participants in a way that it was easier for them to engage in it. Four participants offered comments that contributed to this theme, which

193

was named; “we also had fun”. “Also” is used to honour student recognition that the program did have a serious focus, while “fun” is used to indicate that students did find elements of the program enjoyable. Hard but worth it. As suggested above, students also acknowledged that the program was often challenging. All five reported it was sometimes “difficult”, four did so in particular reference to the sessions in which they “told” their “stories”, and two noted song sharing sessions were particularly confronting. What I found interesting, however, was that despite the widespread recognition sessions were often challenging, this was also identified as a rewarding factor or process of participation. This gave me the feeling that this confrontation was beneficial, perhaps even necessary, for these young people to deal with bereavement. As such, I named this theme “hard but worth it”. This idea was expressed in statements such as the following by BG2: Interviewer: How was that [song sharing] session? BG2: Ahh, I was upset, but then also kinda relieved that I could tell someone what I was going through. Interviewer: Ok, and what did you see as music’s role in that BG2: It was good because, like most of my friends that are in the group knew that song, and they knew that I would play it, because it would remind me of my dad. This statement illustrates that even though the experience of sharing her song – and going through the emotions it evoked about the loss of her father – was difficult, it was valuable. This value was in both being able to express her experiences of grief, and communicate them to her friends. This suggests that elements of the program that encouraged students to confront their feelings was worth it for them, as it enabled them to experience the benefits of going through those experiences in an appropriate environment. A total of seven themes were identified in this category. An example of how the raw data and codes were combined to create themes in this category is presented in Table 8.3.

194

Table 8.3 Raw Data and Codes Used to Form the “Private Group Setting” Theme Category Theme Code Data

Perceptions of the program Private group setting Because I knew and trusted Got to know each Because they're them other better like me BG2: […] I saw that some BG1: All six of us are BG1: […] of my friends were going like really close. knowing […] there, so I just wanted to go we’ve all been BG2: I didn’t know through the BG2: One girl is in my that she lost some of same thing. her friends, and then I class, and another girl in finally found that out. another class and we’re Then another boy, he pretty close lost his mum, and I BG2: […] it was good knew that, but I because, most of my didn’t know how he friends in the group knew was coping with it, that song, and they knew and then I finally that I would play it, realised that he was because it would remind coping with it kinda me of my dad bad, and he wouldn’t talk to anyone, and BG3: […] yeah [the program helped], because I when he found out that I lost my dad, me knew I could trust people and him just talked a […] because I already lot about happy stuff knew them. BG3: […] yeah it was actually [easy to commit], because I knew the people.

Because it’s confidential BG1: Just listening to each other and knowing that it’s confidential, they won’t go out and blab or laugh at you because you cried about it BG5: […] there was good friends that you can trust and won’t say nothing to other people.

Everybody could have their say BG4: I reckon I would do the same thing coz it was pretty good, getting along with everybody, and being involved – everybody had a chance to get involved, it wasn’t always one person getting chosen, it was always, everybody had their option to have their say in whatever it was

BG3: I’ve changed by, now I know the people even better, now we all know each other’s stories

BG4: [in the group was] my sister, two other girls [who] were my best friends BG4: [my best bit was] Probably BG5: […] there was good making friends with friends that you can trust everybody and won’t say nothing to BG5: […] it was nice other people. knowing what their BG5: [I participated] story was, everyone because I just noticed some telling what happened people that were in the in their life. program.

Emergent themes. Interviews primarily produced data related to the “Psychosocial wellbeing outcomes” and “Perceptions of the program” categories. This was not entirely surprising given the nature of the program and the focus of the semi-structured interview (see Table 6.4, Chapter 6). That said, given the philosophical and theoretical assumptions underpinning this research (i.e. the importance of valuing the participants’ voice), actively providing a space within the analysis process to allow new themes to emerge was considered important. For me, the premeditated creation of the “Emergent

195

themes” category was one way of doing this, and the fact little data was coded into it does not diminish its importance. The survey was confusing. Data coded into this category, did however, prove to be important for the study in a way that extended past a commitment to honouring the participant’s role in knowledge creation. While this data did not always relate directly to either the nature of the music program or its psychosocial wellbeing outcomes, I felt it did provide valuable insight into this area of research. For example, one response spoke to the methodological approach of the BG study, and suggested the need to consider more carefully the use of surveys in this area. This particular comment was offered by BG4 when asked what she thought was the most challenging element or experience of participating in the program: BG4: The only time when I was really confused, was like answering all those questions in the, the question things that we got in the start and the finish – some of the questions were basically repeated, and some of them I didn’t really understand. This indicated that not only was the content of the survey confusing for this girl, but she was also unsure of its purpose – so much so she identified it as being a part of the musical program. This caused me to seriously question the validity of the survey results in this study. It also alerted me to the need to consider what survey tools are used in future studies, and how they are administered. For me, even though this can be seen as an isolated piece of data, offered from only one participant, the fact it was an unsolicited response makes it powerful. This response was placed in the theme “the survey was confusing”. We missed class, but it was worth it. Further themes that arose from the “Emergent themes” category were partly related to the “Perceptions of the program” category. These responses, however, were not categorised into the later as, while they contained data relative to how participants perceived elements of the program, they did not appear to be related to psychosocial wellbeing outcomes. These include responses to an interview question which asked about barriers to program participation. This question was originally aimed to explore the sustainability of providing programs such as these in mainstream schools, and assess 196

the impact of taking students out of regular classes. In relation to this, two participants suggested it was not a problem because they just missed “boring classes”, such as “maths and English”. BG1 suggested: BG1: I was just missing out on classes, a couple of classes, but we catch up easy. It was a different class each week. Further, BG2 suggest that, while it was not always ideal, it was worth it: BG2: I think one of our sessions they pulled us out of P.E, and I really wanted to do P.E because I’m in a class with bike, and I really wanted to learn more about bikes, But I saw that some of my friends were going there, so I just wanted to go to see how I would cope with it. I do not for any moment suggest missing classes will have no impact on students, or that classes such as maths and English are “boring” or unimportant. Rather, what this data suggests to me is that sometimes it may be advantageous for the overall wellbeing of students to find ways to work within the parameters of mainstream school systems to deliver directed programs like this one. Furthermore, given school seems to be an appropriate place to provide such programs, for it is a key point of access for young people, sometimes it may be worth negotiating perceived barriers to this kind of participation. To represent this idea, I have named this theme “we missed class, but it was worth it”. Maybe next time. The two final themes related to the last interview question, which asked participants to identify whether anything in the program needed to be changed. Three participants took this opportunity to suggest what may have made the BG program better. BG3 claimed he “didn’t get enough time”, suggesting more or longer sessions. He also said he would like to see more people in future groups: Interviewer: Do you have any other ideas about what you would like to do in a future program like this? BG3: Same, but more people. Mix of new people and people knew already, everybody.

197

However, these program elements were not reported as having a direct influence on reported outcomes. Therefore, these responses were coded as “maybe next time” to indicate they were things identified as being of possible benefit in future programs. I’d recommend it to anyone. The final theme was made up of the remaining responses to the same interview question to which BG3 responds above. In these cases, however, all participants indicated a general level of satisfaction with the program. For example, several participants noted they would readily “recommend” the program to anyone who was suffering from bereavement, while others suggested it did not “need to be changed”. For example, BG4 reflected: BG4: I don’t reckon I would have done it any other way. To me these represent a level of overall satisfaction for the design, facilitation, and outcomes of the program. For this reason, this last theme was named “I’d recommend it to anyone”. While these last themes do somewhat relate to participants’ perceptions of the program, they have been classified here as emergent themes given they were not reported as having a direct impact any psychosocial wellbeing outcomes. Even so, I perceived them as important to participants. Given my own philosophical and theoretical stance, it was important to give voice to these perceptions in order to keep with a participatory approach to knowledge generation. In total, four themes were identified in this category. An example of how the raw data and codes were combined to create these themes is presented in Table 8.4. Table 8.4 Raw Data and Codes Used to Form the “We Missed Class, But it was Worth it” Theme Category Theme Code Data

Emergent themes We missed class, but it was worth it It was just maths It was easy to catch up I wanted to go anyway BG3: I did have to miss BG1: I was just missing BG2: I think one of our sessions they pulled us out out on classes, but boring out on classes, a couple of P.E, and I really wanted to do P.E because I’m classes like maths and of classes, but we catch in a class with bike, and I really wanted to learn English. up easy. It was a more about bikes, But I saw that some of my different class each friends were going there, so I just wanted to go to BG5: […] it was just week. see how I would cope with it. maths.

Qualitative results summary. All students reported experiencing a number of psychosocial wellbeing benefits from the music program. These mostly related to dealing with loss and bereavement, 198

including the facilitation of social support and helping to move on. However, students also noted these outcomes were sometimes conditional. Students also reported that while music was an important element of the program, musicality was not a pre-requisite. Some noted that it had changed their relationship with music. Other elements of the program reported as important included the confidential group setting. Additional themes that emerged included the idea that missing class was worth it to attend the program, and that, while there were some things they would like to see done differently next time, on the whole students would recommend the program to anyone. One student also reported some dissatisfaction with the survey. Observational Reports Survey sessions. For this study, the music therapist undertook survey data collection during program sessions. After collecting follow-up survey responses, the music therapist came to me and related her experience of this process. She had encountered audible resistance to the request of filling out the follow-up survey from several of the participants, who had to be coaxed to complete it. This is despite the fact they had been engaged during all program sessions, including the one in which they filled out the last survey. It seems possible it was the survey itself they were reacting against. This idea was supported during interviews, in which I also experienced unsolicited feedback which both expressed dissatisfaction with the survey process, and challenged its relevance to the music program (see “the survey was confusing” theme above). Interview sessions. Despite initial reservation from one boy, all participants appeared quite open during interviews when talking about feelings of grief and loss, and how the music program had affected them. They also appeared open when identifying elements of the program and research process that they felt did and did not work. However, despite prompting on my part, interviews were brief and most participants offered short responses, while others did not respond to some questions. This appeared to be a combination of both the age and comprehension levels of the respondents, and the fact that they were talking to a stranger.

199

Chapter 9: Critical Reflection Analysis 1 This chapter reports the results of the first critical reflection analysis (CRA). This analysis critically reflected upon the research design and methods used in the Keys to Success (KTS) and Bereavement Group (BG) studies (see Chapter 6), as well as the subsequent data and results of each (see Chapters 7 and 8 respectively). Using this information, and observational reports from both studies, the CRA aimed to identify challenges to investigating the psychosocial wellbeing benefits of school-based musical participation related to the research method and design used in these studies. The potential for school-based musical participation (defined here as involvement in any musical activity facilitated in an educational context) to enhance the psychosocial wellbeing of students in mainstream settings has been widely reported in policy literature (Australian Government, 2005). Research aiming to confirm this link, however, has reported mixed (Gooding, 2011), and sometimes contrary findings (Rickard, Bambrick, et al., 2012). This lack of consistent or positive results has been attributed to several methodological challenges, including design issues (Currie & Startup, 2012), student engagement in research activities (Knox-Anderson & Rickard, 2007), and student understanding of academic or subjective concepts (Kim et al., 2006). During collection and analysis of data from both the KTS and BG studies, similar challenges related to methodology were experienced. This prompted the decision to critically reflect on both the procedures and results of each study to analyse how these challenges related to and affected the process of conducting school-based research in this area. Several perspectives were considered as part of this rigorous review, including that of students, teachers, researchers, and the school principal. Challenges were identified as being related to distinct aspects of design, method, and other factors commonly associated with research in school settings. These aspects were grouped into six overarching areas: 1. Inconsistency in reported benefits. 2. The challenges of comprehension. 3. Engagement in the research process. 4. The challenges of quantitative reporting methods. 5. The challenges of qualitative reporting methods. 6. The challenges of intervention studies in schools. 200

This chapter provides a structured presentation (observable in section headings) of the individual challenges and sub-challenges identified in each of these areas. This includes an explanation of how each were seen to manifest in the two empirical studies, and where possible, discusses their relationship to the scant existing literature. Given the central role of critical reflection and interpretation in the CRA process, results are often presented using the first-person to reflect the presence and voice of the researcher. All challenges were perceived to have important implications for research in this field, thus this structured account aims to provide a reference point for future work and discussion in this area. In this thesis, they have been used to inform recommendations for prospective research (see Chapter 11) and policy (see Chapter 12). Inconsistency in Reported Benefits Discrepancy between reporting methods. The first challenge identified through this CRA was the inconsistency between data sets. Several discrepancies were observed between what was reported in qualitative and quantitative results. These occurred at a general level, in that benefits were only observable in interview data, and at a more specific level, in that results from both methods were sometimes contradictory. Benefits only reported via qualitative methods. In both KTS and BG studies, the psychosocial wellbeing benefits of musical participation were only reported clearly in qualitative data. For KTS participants, this included benefits such as improved affect, engagement with learning and school, and social connection (see Table 7.3, Chapter 7). BG students reported increases in social connection, social support, affect, and bereavement-specific benefits such as letting go and moving on (see Table 8.1, Chapter 8). No significant benefits were reported for either program in quantitative data. While improvement was evident on some scales, no mean change scores showing an increase in psychosocial wellbeing were statistically significant. This suggests psychosocial wellbeing benefits in this area may be more readily reported via qualitative methods. This is consistent with previous claims that qualitative methods may be better suited to capturing the psychosocial wellbeing benefits of musical participation in schools (Grimmett et al., 2010; Knox-Anderson & Rickard, 2007; Shields, 2001). It also supports claims that qualitative methods are better suited to

201

capturing the benefits of any arts participation in schools (Eisner, 1999; Ewing, 2010; Hunter, 2005; O'Toole, 2010; Winner & Hetland, 2000), and investigating the psychosocial wellbeing of young people more generally (AIHW, 2012). Inconsistencies between qualitative and quantitative data sets. In some cases, benefits reported in interviews were contradicted by survey results. For example, all participants BG students reported an increase in social support and connection during interviews. However, data collected using the two scales in the Social support domain of the survey showed social support decreased during the program. Similarly, all five BG participants reported increased affect, emotional control, or general levels of psychological wellbeing during interviews. These increases were not reflected in survey data, with scores on scales such as Emotional control showing a decrease at follow-up. Similar contradictions were evident in the KTS group. Increased engagement with school and learning was evident in eight of nine participant interviews, and figured prominently in the teacher interview. Yet, data from one scale in the School domain showed a decrease at follow-up, while the other showed an insignificant increase. Similar inconsistencies were observed in areas of social connection and affect. Inconsistencies between qualitative and quantitative reports within a single study have been reported widely (Darrow et al., 2009; Kim et al., 2006; McFerran et al., 2010; Rickard, Bambrick, et al., 2012; Shields, 2001). In all cases, reported benefits were more evident in qualitative than quantitative data, further suggesting quantitative reporting methods are unsuited to capturing benefits in this area. Similar observations are reported by Grimmet, Rickard and Gill (2010). Using qualitative methods, they retrospectively assessed the self-esteem benefits of a school music program reported by Rickard, Appelman, et al. (2012) 8 using quantitative methods. Their findings not only repudiated Rickard, Appelman, et al.’s positive findings, but also identified the use of quantitative methods as potentially problematic in this area. In comparing the results from both qualitative and quantitative methods used in the BG and KTS studies, qualitative methods emerged as being better suited to

8

Despite being released later, Rickard, Appelman, et al.’s, 2012 research was

completed prior to Grimmet, Rickard, and Gill’ 2010 paper. 202

capturing the psychosocial wellbeing benefits of school-based music programs. This suggests quantitative methods present a significant challenge to this kind of research. Discrepancy between reporters. Between students. Discrepancy in the psychosocial wellbeing benefits reported by different participants in the KTS group during interviews was also observed. While four of the nine students reported the program increased social connection, the remaining five reported that it had no impact in this area. Similar patterns were observed for school engagement and engagement with learning. This discrepancy indicates that the benefits of the music program differed from person to person: an important outcome for one participant was absent for another. This reinforces reports that the benefits of school music programs are experienced differently by different students (McFerran & Teggelove, 2011). This presents a challenge for researchers attempting to capture such benefits using quantitative methods. If only a small proportion of a group experiences a certain benefit, it is unlikely that benefit will show up as a statistically significant outcome for a group, particularly when using group mean scores as an indicator of outcomes. That a certain benefit does not present a statistically significant group outcome does not necessarily mean it did not exist, or was not an important outcome for participants who experienced it. Rather, given a critical mass of participants must report a benefit strongly enough to raise the groups’ mean score (by a statistically significant amount), it means this benefit is unlikely to be detected using standard quantitative methods. Between students and teachers. Inconsistency was also observed between benefits reported by students and the teacher in KTS interviews. This was most evident in the area of student-teacher relationships. While the teacher reported the program had positive impact on their relationship with students, students did not report the same experience. Six students reported the program had no effect on their feelings toward their teacher, two were unresponsive and one suggested minimal impact. Kim (2015) relates a similar observation, noting teachers in her study reported markedly better psychosocial outcomes from a music therapy program than students. She connects this to the teachers’ enthusiasm for having students participate in a music

203

program, and goes on to insinuate results were likely a reflection of their “eager[ness] to report positive outcomes” (Kim, 2015). This suggests observer (i.e. teacher) reports may not always present an accurate reflection of students’ subjective experiences, a premise corroborated by Michel and Farrell (1973). Rickson and Watkins (2003) also reported conflict between teacher and non-teacher reports as a challenge in their study, and concur with Rapoport and Ismond’s (1996) claim that teachers are not reliable reporters for student outcomes. Alternatively, the present incongruity may suggest the teacher was better placed to perceive change in the student-teacher relationship, thus questioning the validity of student reports and advancing the value of teacher report. Either way, this inconsistency indicated to me that discrepancy between student and teacher (and or other third-party) reports presents a challenge when investigating a link between student musical participation and psychosocial wellbeing. The Challenges of Comprehension Students’ comprehension of the language and concepts used in data collection procedures presented one of the most persistent and obvious challenges to investigating benefits in these studies. I observed that students in both studies struggled with this issue significantly. Furthermore, this occurred both during surveys and interviews. Ability to understand the language in surveys. The difficulty experienced by students in understanding the language used in surveys was evident from several sources. Before and during survey sessions, teachers from both control and intervention classes in the KTS study noted that the language in the surveys would be difficult for their students to comprehend. This was confirmed by my own observations in these sessions. Students regularly sought clarification for the meaning of words and terms used in questions and response options. While more pronounced for some students, who needed extra assistance for the majority of survey items, this appeared representative of the groups in general. That students struggled to understand survey content was also evident within the KTS interview data. Two students explicitly told me they had struggled with question wording, citing that it did not always “make sense”. When asked, other students also gave responses that suggested to me they had struggled in this regard. Similarly one BG participant also complained that “some of the questions were basically repeated, and

204

some of them I didn’t really understand”. I interpreted this to mean that, not only were the questions themselves hard to understand, but the often subtle but important differences between items were not perceivable. The fact survey wording was problematic for students is curious given these scales have been used widely with this population (DEECD, 2010; Victorian Government, 2011), and many were specifically developed for use with students (Bond et al., 2000). However, this observation is vindicated by reports that some young people struggle to understand and engage with the content of self-report tools designed to measure psychosocial wellbeing (AIHW, 2012; Knox-Anderson & Rickard, 2007; Michel & Farrell, 1973). Two factors that appeared to account for this lack of understanding were students’ stage of development, and the cultural and linguistic diversity (CALD) within groups. Stage of development. Teachers attributed students’ difficulty in understanding survey items to their low proficiency in academic or intellectual tasks. They advised some words were too “big” for them, which I interpreted to mean; grammatically complex and hard to read; representative of unfamiliar concepts or ideas; or simply words to which they had not yet been exposed, or were unfamiliar with. This was upheld by my observations, both when answering student queries during survey activities, and listening to the way students articulated themselves during these sessions. These observations were also reinforced during interviews. Literature also suggests a respondents’ developmental stage can inhibit their ability to report psychosocial wellbeing, given the challenge it presents to comprehending language and concepts in research tools (Hamilton & Redmond, 2010), particularly for those with low academic proficiency (Knox-Anderson & Rickard, 2007). For example, the Australian Institute of Health and Welfare (AIHW) note the ability for young people to provide valid self-reports of psychosocial wellbeing “depends on their cognitive capacity to understand the question and communicate a response” (2012, p. 27). Further, they claim that while developmental stages are often associated with age, progression through stages varies between individuals (thus explaining why comprehension was not an issue for all students). They go on to state the importance of using appropriate question wording and response formats, suggesting further

205

investigation is needed in this area to establish suitable ways to assess the subjective wellbeing of young people. CALD. In my view, the proportion of students from CALD backgrounds presented two important challenges for survey comprehension. The most obvious of these was my perception during sessions that students who spoke English as an additional language (EAL) experienced heightened difficulty in understanding and responding to questions. The other related to the idea that there are different culturally-based conceptions of wellbeing. This was less obvious to me (recognised only through reviewing literature), but something I acknowledged as presenting a potential challenge nonetheless. Language comprehension. From my experience in classrooms, I deduced at least 10 to 20 per cent of KTS participants were EAL students. This was backed by the principal’s comment that students in the school were “significantly from non-English speaking backgrounds”, 9 and is also consistent with a larger trend in the state of Victoria (DEECD, 2014; Department of Health, 2014). This added dimension of ESL appeared to make responding to surveys even more challenging for these participants. Approximately a third of students who asked for help with surveys had thick accents, and needed basic terms or language explained. This suggested many only had a basic to intermediate grasp on English. Teachers also noted this challenge. For example, one asked me to lower my expectations regarding the responses from one girl given she had emigrated from India only two weeks earlier. Cultural conceptions of wellbeing. During survey sessions, I also estimated that roughly 75 per cent of students in KTS classes were from cultural backgrounds not consistent with the white Anglo-Celtic culture of Australian mainstream society. Again this was supported by information from the principal, who claimed over 40 different cultural groups were represented in the school community, and is consistent with government data regarding the changing population of Victoria more generally (DEECD, 2014; Department of Health, 2014).

9

Quote taken from interview with school principal. 206

Reflecting upon claims that concepts of wellbeing are socially and culturally constructed (AIHW, 2012; White, 2008), I feel this diversity may have affected some students’ ability to engage with the measures of wellbeing used in the survey. The survey focused upon mainstream Anglo-Celtic understandings of wellbeing, which may be incongruent with concepts of wellbeing experienced by the myriad cultural groups represented in the studies. This feeling is consistent with previous reports that different cultural approaches to wellbeing can make capturing it through empirical means a problematic exercise (AIHW, 2012). Ability to understand interviewer’s questions. The issue of comprehension was not limited to quantitative data collection procedures. It also presented a considerable challenge in the collection of interview data. Throughout interview sessions conducted for both studies it became clear that students did not understand my questions. Some students explicitly told me this, and sought clarification of terms and concepts. Others gave responses indicating they had not grasped a question. These included generic, coverall or unspecific responses not directly related to the question; responses that were completely off-topic; short non-committal answers; one-word answers; or, in some cases, silence. In such situations, I often asked if participants understood my questions, to which invariably the answered they did not. Often I tried to clarify questions using different wording, or asking them a different way. While this worked in some instances, many times it did not. This suggests that even when given the opportunity to discuss questions with participants, explain difficult terms or concepts, or ask questions in alternate ways, there is no guarantee students will understand what we as researchers ask them. This has serious implications for all research relying on self-report data. Interestingly this does not appear to have been addressed in the literature to date. Ability to respond to interviewer’s questions. Some students also had difficulty responding to interview questions. While the challenges of understanding a question and responding to it are inextricably linked in many cases described above (regarding comprehension of questions), this was not always the case. In some instances students did understand questions but found it difficult to respond. Some students told me they could not “find the words” to explain their responses. Others, simply said they “didn’t know”, indicating to me they were

207

unsure how to respond to the given question. An illustration of the latter was related to the interview question; “How did [the music program] make you feel about your place in community and or society more generally”? After explaining the question to one boy, and confirming that he understood it, he responded that he “really didn't know”. It appeared he had never thought about his place in society or community before, and therefore did not know the answer. The challenges of both understanding and responding to interview questions were again related to stages of cognitive and intellectual development, as well as EAL (while the latter was less pronounced in interviews). What stood out here, however, was the difficulty students had in articulating answers related to abstract concepts. This challenge is also noted by Kim et al. (2006), who suggested students in their study struggled to articulate themselves in relation to the subjective topics investigated. Incongruence between lay and researcher’s vocabulary. The difference between students and researcher lexicons also manifested as a challenge during interviews. It became apparent terms commonly used and taken for granted by the latter, such as “wellbeing”, “social”, or even “connection” or “useful”, were not necessarily in the everyday vocabulary of participants. And those that were, were not always used in the same way. Thus, it often appeared that even when using the same words, students and I were talking about different things. Other times it became obvious that much of the vocabulary I habitually use to articulate a concept in writing, or discussion with colleagues (or even conversations with friends) was of little use when interviewing students. This was also supported by teacher reports, and observations made during quantitative data collection. Unfamiliarity with academic concepts. Concepts important to us as researchers, such as “social”, “wellbeing”, “community”, or “society”, were largely unknown by participants. While these concepts underpinned the collection of data from a research perspective, they appeared foreign to many participants. This presented an obvious challenge in collecting and analysing data: it was hard for students to articulate the impact of a program on their connection to the school community when they weren’t familiar with the idea of what a community was. While linked to challenges regarding cultural conceptions of wellbeing mentioned above, this presents a challenge at a more fundamental level. If a participant

208

is unfamiliar with or has not been exposed to a concept, it does not matter whether the researcher and participant come from congruent cultural or social backgrounds. The ability to communicate one’s point of view in regards to a particular concept demands some understanding of that concept. This challenge was observed during interview sessions, and was present within interview data. It was also observed in survey sessions, and reported in the interview with the teacher from the KTS study, who suggested the themes and concepts used in surveys were too abstract for the students to grasp. Again, this idea does not appear to be represented in the literature. Engagement in the Research Process Students’ level of engagement in research activities presented a significant challenge to the investigation of benefits. While more evident for the KTS group, this was also observed in the BG program. This was apparent across different data collection procedures, and presented in several different ways. Disengagement from survey sessions. Several authors note that student disengagement from survey data collection activities affected the ability for their analyses to reach statistical significant results. Knox-Anderson and Rickard (2007) suggest the “paper-pencil” method they used for surveys disengaged students, especially those with lower academic skills, and attribute missing data in their study to the use of this method. McFerran et al. (2010) also proposed self-report survey methods led to student disengagement, reporting vocal complaints and non-compliance behaviours during survey completion. These and several other disengagement behaviours were also observed in the present studies. “Playing up”. Around a quarter of students in each KTS class appeared to disengage from the survey at some stage during sessions. These students started other activities like talking with classmates about unrelated topics, quarrelling with other students, leaving their seats to walk around the room, drawing, or daydreaming. Such students needed prompting to attend to the survey to ensure they recorded their responses and finished in the allotted time.

209

Before sessions started, teachers from both KTS classes independently suggested disengagement might be an issue. This prediction was ratified by my observation that teachers disciplined students several times during sessions. Not taking it seriously. Students often read questions aloud, shared answers, and completed surveys in pairs or groups – despite my requests for them to do so privately. In at least five cases I heard students discussing their answers and joking in a way that suggested they were not taking the survey seriously. Such observations also led to my sincere doubt that these students were providing responses that gave a representative account of their experiences – particularly for questions on sensitive topics. These perceptions were verified during the KTS teacher’s interview. When asked what he thought of the survey, his immediate response was; “I have to be honest with you, I don’t think the students took that seriously”. This lack of serious or sincere engagement represents, for me, one of the most important challenges for investigations that use self-report survey methods in school settings. Yet interestingly, again this phenomenon has not been reported in the literature. Resistance to participation. Some KTS students also made comments that implied they did not want to complete surveys. These included questions such as, “Why do we have to do this?”, or “Do I have to do all of this?” I did not collect survey data in the BG study, but did seek commentary from my supervisor after she had done so. She also noted a general level of dissent from students when they were asked to fill out the follow-up survey. This was especially the case for one student who vocally challenged the need to fill out the survey a second time. Other students also complained loudly, and questioned the point and length of the survey. Student resistance to surveys was also evident in interviews. Some students in the KTS program said they did not like the survey, and questioned its purpose. One student in the BG program expressed similar disdain, noting it as the only thing about their participation in the program that they did not like. Rushed, careless, and missing responses. In some cases I observed students rushing through surveys, or responding carelessly. Sometimes students were either being pushed by teachers to finish in the 210

allotted time, or they were reacting to suggestions they would have to stay after class to finish the survey during their break time. Other times it appeared they did not understand questions, were bored, were resistant to participation, or just wanted to finish the survey as soon as possible. In these cases students did not seem to be reading questions and responses carefully, and were just responding. This was also evident when entering the data from the paper surveys into an online database; I observed rushed pen strokes, and responses that followed consistent patterns on scale response matrixes (i.e. ticking the middle option for each item on a scale). In many cases, it seemed this careless or rushed response approach was the cause for missing data on single items within a scale, and whole scales or pages within the survey. Disengagement from interview sessions. Some students also appeared to not engage with interview sessions. This was most notable in the KTS group, but also evident to a lesser degree with BG participants. Non-committal responses. In several cases, rather than taking the time to think about and offer responses that reflected their experiences, I perceived students offered either non-committal responses, or the minimum response needed to finish the interview. Students often appeared to just agree with me, repeat my words, or repeatedly use the same terms or responses throughout the interview. Several students also appeared to give me just enough of the “right” responses (i.e. ones they thought I wanted to hear) to get through the interview. While this may be equated with the “yea saying” effect (where participants give socially desirable responses, see Gergen & Back, 1966) I interpreted this behaviour differently. For me, such interactions were efforts to get through interviews as quickly as possible, and in a way that required the least amount of engagement on their part. Other students appeared to pick up on some of the key terms I was using, and just repeat them back to me. Again I identified this as a strategy for getting through the interview with minimal input. Such tactics appeared to be an attempt to give me just enough information so I would continue through the questions until the interview was finished and they could leave. These interpretations drew heavily from my observations of, and reflections on, interview sessions, including participants’ body language, tone of voice, and the overall

211

quality of responses. This was predominantly experienced in KTS interviews, and emerged as an unanticipated theme from the data. Short answers. While all students in both studies gave relatively short answers, this was most notable for those who seemed particularly disengaged during these sessions. Again this was mostly for those in the KTS study. These students were largely unresponsive to prompts to explore their responses or experiences in more depth, often using body language to indicate they were not prepared to do so. This was a substantial challenge for data collection and analysis, as it led to a significant dearth of data (an associated issue also discussed later in this chapter). While there has been discussion regarding the willingness for respondents to engage in interviews, this predominantly relates to issues of “truth telling”, anonymity, and intimacy (Plummer et al., 2004; Weinreb, Sana, & Stecklov, 2011; Youn, 1996). Rarely, if ever, has a young person’s resistance to interview participation at a more general level been reported in the research literature. Not understanding the purpose of the research. One factor that potentially explains the lack of student engagement observed in both quantitative and qualitative research activities is linked to students’ comprehension of the purpose of KTS and BG studies more generally. In their interview, the teacher from the KTS group proposed an explicit link between students’ lack of engagement in survey sessions and the fact they did not understand the purpose of the research. They claimed that for students to engage in such activities, they needed to have a better understanding of who we were as researchers, what the research was trying to do, and why it was important. He also suggested this required an extended period of engagement with the students of up to one year. Without an understanding of the purpose of the research, he argued students would not take it seriously or engage in it to a degree that would afford accurate results. This sentiment was also evident in students’ interview responses. Several students from both programs made it clear they did not understand the purpose or importance of surveys. Further, these statements were invariably coupled with comments that expressed an aversion to completing surveys. For me, this provided a

212

clear link between sufficient understanding of the value of research, and a lack of willingness to engage with, and subsequent disdain for, survey activities. I gained a similar perception from less verbal interactions experienced during interviews. Despite the care I took to explain myself and the reason for our interviews, it seemed apparent that some students were unsure why they were in a room with someone they had only seen two or three times, and being asked questions that were for them (it seemed) unclear in their purpose. The Challenges of Quantitative Reporting Methods Several challenges to investigating psychosocial wellbeing benefits observed in these studies related specifically to use of quantitative research methods. These included elements of design that are necessary in any research using these methods, and thus illustrate some of the inherent challenges of using quantitative approaches in this setting. Others related to design elements that are preferable for the facilitation of rigorous research, and others still relate to the methods themselves. Distribution of data. Variability of reported data. Significant variability was observed amongst individually reported scores for all groups at both baseline and follow-up. This was likely linked to the small sample sizes in both studies, and the fact students in the KTS groups varied in terms of academic ability, CALD background, and developmental stages. It may also represent random response variation, again potentially attributable to issues such as comprehension and disengagement from the research process. Ultimately this variation made achieving statistical significance problematic, even when a positive change was reported, thus presenting a significant challenge to the ability to reach conclusions based on this data. Ceiling and floor effects. Several ceiling and floor effects were apparent in scale data for both studies. This was most evident for the KTS study. Intervention group data showed potential floor and ceiling effects in baseline scores for the Community domain on Low neighbourhood attachment and Opportunities for prosocial involvement scales respectively, and ceiling effects on all scales in the Psychological wellbeing and Psychological needs domains. There was also evidence of ceiling effects for both groups on all scales in the Peer and individual, Emotional control and Social support 213

domains. Subsequently, the intervention group reported high levels of psychosocial wellbeing on 12 of the 18 scales at baseline. For the BG study these effects were most notable on the Educational expectation, Opportunities for prosocial involvement, and Emotional control scales. Such high baseline scores present a challenge to capturing any improvement. The minimised amount of increase possible on scale ranges meant it was unlikely any improvement would be sufficient to achieve statistical significance. These scores also suggest the decline on some scales may reflect a regression to the mean or random response variation rather than a decrease in reported wellbeing. The challenge of high baseline scores are noted by Rickard, Appelman, et al. (2012) and Rickard, Bambrick, et al. (2012). In both cases they suggest this prevented the ability to capture a significant increase in psychosocial wellbeing constructs. Nature of scales. That benefits reported in interviews were not evident in quantitative data suggests the ability for the chosen scales to capture any reported benefits of musical participation was limited. This appeared to be due to the purpose of scales, the constructs they measured, and the wording used. Purpose of scales. Many scales used were designed for assessment or screening purposes, rather than detecting change over time. While created for use with students in school environments, CTC scales (see “Materials”, Chapter 6) are intended for epidemiological research, and designed to detect the prevalence of a risk or protective factor in a given population (Arthur et al., 2002). The same can be said for the Emotional control (Prior, Sanson, Smart, & Oberklaid, 2000) and Trusted adult in life (DEECD, 2010) scales. This potentially explains the floor and ceiling effects observed with these scales, and why they did not detect statistically significant change on constructs measured. Other scales used in the present studies, such as the MSPSS-SO (Zimet, Dahlem, Zimet, & Farley, 1988) and MHI-PWB (Veit & Ware, 1983, p. 730), have also predominately been used for epidemiological or correlation study designs, yet their developers suggest they are appropriate for detecting change over time. The K10 is primarily used as a screening tool (Andrews & Slade, 2001), yet is used in intervention research with a variety of populations (Crockett, Taylor, Grabham, & Stanford, 2006; Richmond et al., 2006), including randomised control designs (Griffiths & Christensen,

214

2006) with young people (Hides et al., 2012; Hides, Lubman, Elkins, Catania, & Rogers, 2007). The BPN tool has also been used for detecting change by those who developed it (Niemiec, Ryan, & Deci, 2009). While such scales have been used to measure change, it seems possible they were not designed with sufficient range to detect modest increases in psychosocial wellbeing, especially when used with small sample sizes. This is problematic for this kind of research as even marginal increases may be important for those that experience them (especially on subjective constructs), and because large sample sizes are not always possible or appropriate (a point further expanded later in this chapter). Scale wording used to articulate a construct. The ability to control emotions was a significant outcome reported by BG participants during interviews. Yet, these students’ scores on the Emotional control scale decreased at follow-up. Perplexed by this discontinuity, I decided to interrogate the language used to express this construct in both reporting methods. On inspection, the way emotional control was explained in the scale differed in important ways to the experiences of emotional control described by students. Students explained their increased ability to control emotions in terms of “understanding” their emotions, and being able to “remember” or “focus” on the “good things”. Such concepts are not mentioned in the scale items, which instead focus on knowing “how to relax”, “calm down if I am feeling nervous”, or “control my temper when people are angry with me”. Neither being able to relax or calm nerves were articulated as issues for BG participants. Increased ability to deal with temper was reported in interviews, but this was expressed in terms of reducing anger directed at the world due to their loss or bereavement – not in terms of controlling their temper when others were angry with them. The Emotional control scale also uses the item “I am always able to keep my feelings under control” to indicate high functioning on this construct. The idea of keeping “feelings under control” conflicts with one of the main benefits reported by BG participants: “Letting it out”. In this theme, they articulate that “Letting the feelings out” and knowing that “It’s ok to cry” was one of the most powerful coping strategies that helped them address and deal with negative emotions. This example shows how the wording of a scale may misrepresent a participants’ experience of the construct it aims to measure, thus impeding its ability to report benefits. There were several other examples of this incongruence in both studies. 215

Kim et al. (2006) report a similar challenge in their study, suggesting an alternate self-control scale may have been better suited to picking up participant experiences of this construct. Darrow et al. (2009) make the same suggestion in regards to the self-esteem scale used in their study. Devroop (2012) designed his own scales to capture benefits targeted by his musical intervention, and used language he felt would represent participants’ experiences of these constructs. His success in capturing reported benefits supports the need to use appropriate scale wording to capture a construct. Measuring the “right” constructs. Being able to match scales to the constructs of wellbeing affected by a music program presents another challenge. This involves finding scales suited to the purpose of a musical intervention, and having sufficient breadth in measures to capture the range of potential program benefits. Fitting the scale to the intended benefit. Matching scales to the constructs targeted by the particular music program presented a particular challenge in the BG study. While the program was specifically designed to address bereavement, no specific bereavement, loss, or grief scale was used. Consequently, while all participants reported the program beneficial in addressing issues of bereavement, this benefit was not captured in quantitative data. This mirrors claims made by McFerran et al. (2010) that coping, self-perception and self-worth scales used in their study were not capable of capturing grief related benefits, suggesting a grief-specific scale was needed. Anticipating or capturing program benefits. The benefits of musical participation can vary between participants and programs. This makes anticipating and capturing all (or any) potential program benefits a challenge when using scales. Theoretically this may be addressed by using a multitude of scales, with the hope they will pick up or correspond to at least one benefit, or perhaps capture them all. However, using a “battery of tests” not only adds to response burden (discussed below), but can also create Type 1 error (Sheskin, 2003) 10 and thus represents an ethically questionable approach to research (Shrader-Frechette, 1994).

10

Type 1 error arises when many statistical calculations or analyses are

undertaken on a large amount of data from a single experiment or study until a 216

The subsequent need to anticipate or estimate what benefits will be experienced by which participants (and thus predict which scale[s] to use) can present a significant challenge. As mentioned above, benefits reported in KTS interviews varied significantly between participants. Further, while some participants claimed their engagement with learning was improved through the use of music as a learning tool, others claimed it was the fact musical activities made learning more fun. This shows that it is not only the benefits themselves that vary between participants, but also a participant’s particular experience of a benefit. Benefits also vary from program to program, something clearly evident in the difference between benefits reported in BG and KTS programs. Given differences between programs this is not surprising, yet it illustrates the challenge of pre-empting the expected benefits of a program, and including the relevant scales needed to capture all, or any of the benefits experienced by participants. The challenge of choosing scales which capture all the benefits of a program is also reported by Rickard, Bambrick, et al. (2012). They suggest there may have been several psychosocial benefits experienced by their participants that were not captured by their scales. Response burden. Response burden was also observed as a challenge in both studies. It presented an overt barrier to engagement, and was seen to lead to missing survey data. Participants from both studies complained about the length of the survey when completing them, and many were observed to disengage towards the end of sessions, either exhibiting distracted behaviour or rushing their responses. Comments made during student interviews also indicated they felt over-burdened from survey activities. While participants’ explicit reports of burden primarily related to filling out the survey a second time at follow-up, evidence suggested survey length was another important, if not the central, factor. This was apparent from my observations during manual data entry, and the data itself. For example, the BG survey contained 79 questions asked over 12 pages (see Appendix N), and while full responses were

statistically significant result is found. “Error” occurs when said result is a product of the statistical probability of finding a positive result, rather than a reflection of an actual event or interaction (thus being a false finding). See Sheskin (2003), particularly page 226, for further discussion. 217

collected for the first eight scales to appear in the survey, missing data was observed for six of the last seven scales (see Appendix O). Similarly, the KTS survey consisted of 13 pages and 105 questions, and while response rates were more spread across the survey, there was a recognisable decrease in response rates for the last few questions. Further, during manual data entry I observed that the rushed, careless, or disengaged responses mentioned in the “Disengagement from survey sessions” section above were most noticible in latter pages. Authors have noted the impact survey length can have on response rates (Burchell & Marsh, 1992; Galesic & Bosnjak, 2009; Hoerger, 2010), and suggest data quality often decreases towards the end of long surveys (Rathod & LaBruna, 2005). Interestingly, however, the “battery of tests” approach continues to be mainstream practice in psychological studies, and similar research fields. Response hesitation. Another potential challenge was the reluctance for some participants to answer “strongly”, or towards the top end of item response scales. For example, participants in the BG study claimed during interviews that while they did experience some important benefits, these were conditional. This is represented by codes such as “I'm still not coping with it, but it’s helped”, “It’s not over” and “It’s made me at least a bit better”. These suggests that while benefits were reported in areas such as affect, the fact they were still experiencing issues may have made them reluctant to answer “strongly” on related scales. This is perceivably because wording for scale items measuring affect, such as the “Has living been a wonderful adventure for you?” item in the MHI-PWB scale (see Appendix N), refer to a level of wellbeing they had not reached. This does not mean the program did not have an effect, but that the scales do not have the subtlety necessary to capture this kind of change. Response hesitation (Toepoel, Das, & Van Soest, 2008) and item wording (Schwarz, 1999) are recognised issues in scale measurement research, yet are more often related to a participant’s hesitation to assign themselves a negative score (Schwarz, Knäuper, Hippler, Noelle-Neumann, & Clark, 1991). I argue that in the context of studies such as the BG, participants may also hesitate in assigning themselves to positive values when increases in wellbeing are only marginal, and not represented by response option wording. I further argue that, as qualitative codes from the BG study noted above show, even marginal improvements may be significant for those that 218

experience them. As such, response hesitation was seen to represent another challenge to capturing subjective benefits using scales in these studies. The Challenges of Qualitative Reporting Methods Lack of data acquired. There was a distinct lack of data acquired during the interview sessions in both studies. On average, interviews for the BG study lasted approximately 11 minutes. The average length for KTS interviews was approximately eight minutes, with some as short as five minutes. This presented a substantial challenge to capturing reported psychosocial wellbeing benefits, as there were limited participant responses upon which to conduct an analysis. This seemed linked to several factors noted above, including comprehension and engagement in research. However, there also appeared to be factors endemic to the qualitative research design used. Self-expression in adolescents. The perceived ability or willingness for students to express themselves presented a notable challenge during interviews, and meant students often offered short or inexpressive answers. This inexpressiveness appeared related to the use of face-to-face interviews (FFIs) with this age group in the school context. Nature of material. I perceived the nature of material discussed in interviews was a challenge. Some students appeared uncomfortable talking about certain issues, such as their feelings towards their school, teacher, and classmates, and seemed unsure how much information they should divulge. Others were open to discussing most issues, but appeared unsure of how to express themselves in relation to subject matter (see comprehension challenges reported above). The first of these factors is supported by claims young people often under-report sensitive information in FFIs (Gutiérrez & Torres-Pereda, 2009; Youn, 1996), and the second by claims adolescents can struggle expressing themselves in relation to abstract concepts (McFerran-Skewes, 2005). Talking to someone unfamiliar. Interview sessions were the first time I met BG participants, and my only prior contact with KTS participants was during survey sessions. Some participants appeared bemused by the fact they were being interviewed by a stranger, which in turn seemed to

219

affect their expressiveness. Traditionally the “stranger interviewer norm” (Stecklov, Weinreb, & Sana, 2013, p. 1), where interviewers are not known to participants, has been preferred in data collection activities given the belief it provides more objective data, and reduces response bias (Stecklov et al., 2013). However, recent research suggests the stranger-interviewer method presents a challenge when collecting sensitive information from young people (Weinreb, 2006). This was my experience in several interviews. Despite efforts to build rapport, some students seemed less engaged than they may have been with someone who was familiar to them, and held back from providing detailed information. Setting. Interview sessions were held in meeting rooms in the main administration block of the school. These rooms were adjacent or next to the school reception and the offices of school coordinators and principals. While doors were always closed, conversations in the reception and in the hallways of the administration block were audible. Potentially, this limited the willingness of students to divulge personal information, or information that may be seen as getting them in trouble with school authorities. Literature supports this, stating young people are less forthcoming in interviews where settings or environments lack confidentially, or represent a setting in which they would typically hide certain behaviours (Gutiérrez & Torres-Pereda, 2009; Leigh, Gillmore, & Morrison, 1998; Youn, 1996). The Challenges of Intervention Studies in Schools Sampling. Carrying out the present studies in a school environment led to several sampling challenges. These included accessing sufficient participants, conducting rigorous designs, and equal variance of reported scale scores between groups. Sample size. Both studies had small sample sizes. The BG had only five participants, while the KTS study had 14 in the intervention group, and 17 in the control. Previous quantitative investigations using groups as small as eight (McFerran et al., 2010) and 17 (Knox-Anderson & Rickard, 2007) have not specifically noted sample size as a limitation in this area. Further, De Winter (2013) claims t-tests are appropriate for use with samples of six or less. He qualifies this however, in that they are only appropriate 220

when effect sizes are large, something “uncommon in behavioural/psychological sciences” (2013, p. 7) and not observed in either study here. De Winter finishes by agreeing with Siegel (cited in De Winter, 2013) in that larger sample sizes are always preferable when aiming for statistical significance in behavioural science research. Accessing large samples was problematic in this case both due to the school context, and nature of musical interventions. Barriers to a school’s participation. Studies were carried out in a government school that had pre-existing commitments (i.e. curriculum and other mandated school programs) that prevented their ability to facilitate large intervention designs. This was combined with staff resistance to participation in both research and musical activities above and beyond standard duties. 11 Therefore, it was neither possible nor practical to ask the school to commit to studies which where likely to have a high impact on their already limited time and resources, thus making large sample sizes unobtainable. Literature suggests the expectation that schools have sufficient time and resources required for “rigorous” research designs is unreasonable, and that sample size in school-based research is more often set by pragmatic factors than design preferences (Allensworth, Lawson, Nicholson, & Wyche, 1997). Schools have both curriculum and extra-curricula commitments which are often mandated by policy, and necessary for securing ongoing funding and enrolments. These include the need to maintain standards of academic proficiency among students, measured annually by national testing procedures (Polesel, Dulfer, & Turnbull, 2012). Given overloaded curriculums (Thomas, 2009) and the pressure for schools and students to preform on these tests, taking time away from core academic subjects to focus on non-essential, creative, or group learning activities is often considered a burden on schools (Crooke & McFerran, 2015; McFerran & Crooke, 2014; Polesel et al., 2012). This burden, along with resistance from staff and students to be involved in research (Allensworth et al., 1997; Thomas, 2009), and pressures to fit sport and other extracurricular activities into an six-and-a-half-hour day, limits the ability for schools to allocate large amounts of students to music programs or research projects.

11

Information taken from Music Matters project consultation records. 221

Low participation rates outside school hours. To address the sample size issue noted above, it is possible to conduct programs and or data collection outside school hours. Several studies looking at the impact of musical participation on student psychosocial wellbeing have taken this approach (Costa-Giomi, 2004; Schellenberg, 2004). Yet this requires extra commitment from parents – something reported to be in decline in recent years (Thomas, 2009) – to allow or facilitate a child’s participation, and may be met with significant resistance from students as it would consume their free time. Securing the commitment of students and families to participate in research conducted within schools can require a budget unavailable to many research organisations (Thomas, 2009). Conducting effective research with students outside of school hours may require even more resources, making this setting even less feasible for most researchers. Further, it may also fail to capture the experience of musical participation in school contexts. Challenges in the therapeutic context. Large sample sizes can also be counterintuitive for musical programs aiming to achieve psychosocial wellbeing benefits. Smaller group sizes have long been recognised as important in social work, and are said to be necessary for the impact of a group intervention to be experienced by participants (Hartford, 1971). Later psychotherapy literature has suggested “the ideal size of an interactional therapy group is approximately seven or eight” in general (Yalom, 1995, p. 276), and five to nine when working with adolescents (Malekoff, 2014). A similar trend can be seen in music therapy interventions, which often rely on the reflexive ability of programs to adapt to the strengths and needs (Rolvsjord, 2010) of participants, or the cultural and social contexts within which a group is run (Stige & Aarø, 2012) – suggesting reduced group sizes are more suitable for facilitating these processes. For example, Bullard (2011) reports the small group size of four to six participants in her music therapy study was one of the critical factors which led to program outcomes, going on to suggest that any groups of 12 participants or more are too large. While the implication of group size for the design and facilitation music programs is discussed further in Chapter 10, of importance here is the relationship between program size and study sample size. The above literature suggests that music groups should be kept small to achieve psychosocial benefits. Yet this becomes

222

inherently problematic when attempting to recruit a sample size large enough to reach statistically significant results. In statistical terms, a large sample size equates to a higher possibility of positive results; yet in therapeutic terms, a large sample size equates to a reduced likelihood of wellbeing benefits. Accessing a control group. Gaining access to appropriate control groups was difficult in the school context. Using a control group for the BG study was not only inappropriate given its focus on bereaved students, it would have necessitated dividing the original group in two. This would have been both impractical design-wise, given the small size of the original group, and inappropriate given the potential increase on burden on the school (due to the ethical requirement of running a second waitlisted group for the control). Despite claims that the practicalities of school-based research mean control groups are often forgone when investigating school health programs (Allensworth et al., 1997), methodological literature state them a necessary element of any well-designed research project (Tilley, 1999). The need for an appropriate control group has also been noted in literature investigating the impact of musical participation in schools (Currie & Startup, 2012; Knox-Anderson & Rickard, 2007; Rickard, Bambrick, et al., 2012). Randomisation. While the KTS study did use a control group, neither the selection of participants nor their allocation to study groups was random. Groups were randomised at the class-group level, as these were the groups the school allocated to participate in the project. Dividing these classes using random allocation procedures was not appropriate given the disruption it would have caused to class-groupings and timetables. This lack of randomisation may partially explain the lack of statistical significance achieved. Previous authors attribute their significant results to achieving randomisation (Costa-Giomi, 2004; Kirschner & Tomasello, 2010), while other say their non-significant results were likely due to fact true randomisation was not achieved (Knox-Anderson & Rickard, 2007; Michel & Farrell, 1973). However, studies that claim randomisation was effective conducted their research outside of mainstream school contexts, either using private lessons (Costa-Giomi, 2004), or controlled settings (Kirschner & Tomasello, 2010).

223

Other authors suggests random allocation is not widely used in school health research given practical limitations of conducting research in this context (Allensworth et al., 1997). Rickard, Bambrick, et al. (2012) attempted to randomly allocate participants to groups, yet claim schools provided students the opportunity to influence which group they would be allocated to, leading to a “pseudo-random” design and insignificant results. Currie and Startup (2012) also reported attempts to randomise samples in their study were complicated by school factors, again suggesting this as a reason for their non-significant results. This indicates that even when schools do acquiesce to randomisation procedures, it is not guaranteed they will facilitate it to level of methodological rigour preferred by researchers. Comparability of data between samples. The inclusion of a control group in the KTS study created another challenge. Mean reports at baseline differed significantly between groups on several scales, meaning equal variances could not be assumed. Therefore, in many cases it was difficult to undertake a comparison between the mean change scores for each group. For example, the intervention group’s mean baseline score was approximately eight points higher than that of the control group on the BPN-SF scale (see Figure 7.15, Chapter 7), thus making a comparison of mean change between groups problematic for this measure. Similar baseline discrepancies were observed on several other scales. Ensuring comparability of a control group’s data has been noted as a significant challenge when investigating the impact of school health programs (Allensworth et al., 1997). Several authors investigating the impact of musical participation in this area have also claimed this may have affected the validity of their results (Montello & Coons, 1998; Rickard, Appelman, et al., 2012; Schellenberg, 2004). Missing data. Missing quantitative data was a significant challenge in both studies. Results for the BG on five of the 14 scales (including the MHI-PWB) were based on responses from only four participants (see Appendix O). For the KTS intervention group, results for only seven of the 18 scales were based on data from all 14 participants, and for two scales the n was only 11. This was experienced in both studies as missing responses within a data set, and as complete data sets in the KTS study due to student absence at either baseline or follow-up sessions.

224

Missing responses within a data set. Data missing within a participant’s response set (i.e. where a participant has skipped or missed items or scales) is attributable to factors noted above such as engagement in research, comprehension, and response burden. This type of missing data has been partially credited for the lack of statistical significance reached in previous studies. Some suggest it presents a particular problem when linked to low academic ability, given such students may experience the most observable psychosocial wellbeing benefits of musical participation (Knox-Anderson & Rickard, 2007). Missing full baseline or follow-up. Students absence from class on one or both of the days KTS surveys were administered meant whole baseline or follow-up data sets were often missing, rendering all data from that student unusable in the analysis. Again this represents a challenge of conducting research in schools, as it is hard to predict which students will be in class due to issues such as illness, truancy, and school transfer (Ellickson, Bianca, & Schoeff, 1988; Kano, Franke, Afifi, & Bourque, 2008; McGuigan, Ellickson, Hays, & Bell, 1997). This was particularly problematic in these studies given the small sample sizes. Confidentiality in classroom data collection. While the design and content of many survey items dictate they be completed individually, or in a controlled or private setting, all KTS surveys were carried out as a whole-class activity. This meant that, despite the teachers’ best efforts, and mine, students often completed surveys in pairs or groups, sharing answers with friends or classmates. Moreover, student misbehaviour observed in survey sessions (see disengagement challenges above) included loud reading of questions, often using mocking or sarcastic tones, and the vocal sharing of one’s, and sometimes one’s neighbour’s, answers with the rest of the class. Not only did this suggest some students were not taking questions seriously, it caused me to earnestly question whether all students were prepared to answer truthfully to sensitive, personal, or embarrassing questions. I challenge the idea that any young person would be prepared to commit to this activity with classmates looking on, and potentially sharing their answers with the entire class. Given the dynamics of a Year 7 classroom, and the developmental stage of early adolescence (i.e. dealing with issues of bullying, identity, and forging social

225

connections; McFerran, 2010a; McFerran-Skewes, 2005), it seems logical students would avoid divulging personal information in this situation. For me, this fundamentally challenges the legitimacy and or validity of data collected in these sessions, and I argue, the data and results of any surveys conducted in similar conditions. Indeed, privacy and confidentiality in school-based survey delivery have been noted challenges in the literature for over 15 years (Beebe, Harrison, Mcrae, Anderson, & Fulkerson, 1998), and students continue to report insufficient privacy when completing surveys in classroom setting (De Looij‐Jansen, Petra, & De Wilde, 2008). Yet, intriguingly, this remains mainstream practice in many fields investigating young people’s health and wellbeing. This includes both recent and current population health studies undertaken by the Victorian Government’s Department of Education and Training (http://www.education.vic.gov.au/about/research/pages/newdata.aspx). Conclusion The methodological challenges of investigating the psychosocial wellbeing benefits of school-based musical participation are both numerous and varied (see Table 9.1). While many are acknowledged in existing literature, it is interesting that several challenges identified in this analysis have not been reported more frequently, particularly given how apparent they were in these studies. By presenting the challenges identified here in a clear and comprehensive fashion, this chapter has aimed to both communicate how problematic current approaches to research in this area can be, and provide a reference point for future discussion and research. The current lack of forum for this kind of discussion is a concern, especially if research results that may be based on dubious data sets are taken seriously and used to inform policy or interventions. This analysis does not claim to be an exhaustive account of the challenges in this area. However it does serve to acknowledge the inherently problematic and often paradoxical nature of conducting research in this field. It also challenges the assumption that current approaches to investigating psychosocial wellbeing outcomes in schools are appropriate, and illustrates the impact these assumptions can have on results. It also identifies some elements of current approaches we can work to improve.

226

Table 9.1 CRA 1 Challenges and Sub-challenges Challenge Area Inconsistency in reported benefits

Challenge Discrepancy Between Reporting Methods Discrepancy between reporters Ability to understand language in surveys

The challenges of comprehension

Engagement in the research process

Ability to understand interviewer’s questions Ability to respond to interviewer’s questions Incongruence between lay & researcher vocab Disengagement from survey sessions

Disengagement from interview sessions

227

Not understanding the purpose of the research Distribution of data The challenges of quantitative reporting methods

The challenges of qualitative reporting methods

The challenges of intervention studies in schools

Nature of scales Response burden

Sub-challenge Level 1 - Benefits only reported via qualitative methods - Inconsistencies between qualitative & quantitative data sets - Between students - Between students & teachers - Stage of development - Cultural & linguistic diversity (CALD)

- Variability of reported data - Ceiling & Floor Effects - Purpose of scales - Scale wording used to articulate a construct - Measuring the “right” constructs

- Self-expression in adolescents

Sampling

- Sample size

- Accessing a control group - Randomisation

Confidentiality in classroom data collection

- Language comprehension - Cultural conceptions of wellbeing

- Unfamiliarity with academic concepts - “Playing up” - Not taking it seriously - Resistance to participation - Rushed, careless, & missing responses - Non-committal responses. - Short answers

Response hesitation Lack of data acquired

Comparability of data between samples Missing data

Sub-challenge Level 2

- Missing responses within a data set - Missing full baseline or follow-up

- Fitting the scale to the intended benefit - Pre-empting the benefits experienced - Nature of material - Talking to someone unfamiliar - Setting - Barriers to a school’s participation - Low participation rates outside school - Challenges in the therapeutic context

Chapter 10: Critical Reflection Analysis 2 This chapter presents a critical reflection analysis (CRA) on how aspects of musical participation affect the investigation of psychosocial wellbeing in mainstream schools. It explores both the practical and reported differences between two music programs delivered within the same school and how these differences related to the psychosocial wellbeing benefits reported for each program. This analysis seeks to identify how certain elements of participation affect or link to the reporting of particular benefits. Identifying these links aims to foster an understanding of how the nature of musical participation may affect reported benefits. To undertake this investigation, a CRA was conducted on data collected for the Keys to Success and Bereavement Group studies (see Chapters 7 and 8 respectively). Both studies collected self-reports of psychosocial wellbeing using qualitative and quantitative methods. An overview of this data is presented first to identify the difference in benefits reported for each program. These differences are then discussed in relation to program elements, starting with the particular attributes of participant groups, followed by the attributes of the programs including their design and their facilitation. Qualitative data regarding participant perceptions of the programs are also utilised to identify subjective experiences of participation and how they related to reported benefits. A full list of music program attributes identified as affecting investigations of psychosocial wellbeing, and which are discussed in this chapter, is provided at the end of the chapter in Table 10.2. The Programs The Keys to Success (KTS) program was a classroom-based music program delivered during class-time, facilitated by a trained music therapist and co-facilitated by the regular classroom teacher. This program aimed to explore the potential of using music to deliver curriculum content. The curriculum content in this case related to a preexisting school initiative, called Keys to Success, run during a single 50-minute class period each week with all Year 7 classes. This initiative was designed as a curriculum subject to educate students about five keys to living a successful life, including confidence, resilience, respect, persistence, and organisation. The music therapist was invited to run a musical version of the program for four weeks (delivered as part of, or embedded within, the existing KTS program). Sessions included group songwriting and 228

instrumental improvisation, and focused on the creation of a music video (written, performed, directed and acted by students) that aimed to address issues such as bullying. The Bereavement Group (BG) program was a small music therapy program run on school grounds during school hours in a semi-private setting. Group members were recruited by the school wellbeing coordinator, and were students known to be experiencing bereavement, grief or loss. The program was designed and delivered by the same music therapist who delivered the KTS program, who is experienced in running music therapy groups with bereaved teenagers. There were six weekly sessions that lasted for approximately 50 minutes and focused on addressing experiences of grief and loss. Sessions included song sharing, instrumental improvisation, group songwriting and group discussions. The group also wrote their own song, which was recorded by the music therapist and provided to students at the end of the program. Data was collected for both programs using questionnaires, interviews, and observational reports (see Chapter 6). The analysis presented in this chapter compares reported psychosocial benefits from each program against particular program elements. Comparison of Reported Outcomes A comparative breakdown of benefits reported for each program is presented here as a reference point for the analysis presented throughout the chapter. Comparison of reported benefits is limited to the qualitative data as neither study reported benefits in the quantitative data. These benefits were identified through the psychosocial wellbeing benefit themes identified during qualitative analysis (see Table 7.3, Chapter 7, and Table 8.1, Chapter 8). The first observable difference between programs is the type of benefits reported for each. While both programs reported “affect” and “social connection”, remaining themes centred on certain topics. Benefits reported in the KTS program relate mostly to school experiences, including “school engagement”, “engagement with learning” and “outcomes extended to other classes”. Further, reported benefits such as “prosocial behaviour” and “self-efficacy” (often articulated as “confidence”) were outcomes targeted by the curriculum-based program, and can be seen as learning outcomes. Benefits reported in the BG contained little or no connection to school experiences, centring instead on dealing with bereavement. These included themes such as “dealing with bereavement”, “social support”, “letting it out”, “moving on”, and “I can remember the good things”. 229

The level to which participants endorsed reported benefits also differed for each study. The fact seven psychosocial wellbeing themes were reported for each program gives the initial impression each group experienced a similar magnitude of benefits. Yet, as shown in Table 10.1, benefits were reported more consistently, or at a higher rate, among students in the BG study. The reported rate of endorsement in this table does not aim to suggest benefits with low endorsement were less important or meaningful. Nor is Table 10.1 an attempt to quantify qualitative data: from a qualitative perspective, any reported benefit was considered a significant outcome. Rather, it provides an overview of which benefits were reported for which study, and how many people reported them. This provides insight into the universality of benefits reported for each group. This overview of reporting endorsement illustrates that themes reported by the BG were endorsed by a higher percentage of participants than those reported in the KTS study, suggesting they were experienced more consistently for the BG. This is supported by the observation that all nine participants in the KTS group described an absence of at least three benefits reported by the group (see “Absence of outcomes” theme, Chapter 7). Further, a number of benefits reported by KTS participants were based on questionable responses (see “Emergent themes”, Chapter 7), something not experienced in the BG data. Combined, these factors suggest benefits reported by the BG were more consistent, and more representative of the overall group’s experience of musical participation. To summarise, reported benefits for each study differed in two ways; the type of benefits reported, and the level to which groups endorsed them. The rest of this chapter explores how aspects of musical participation shaped or solicited these differences. Given the central role of interpretation in this exploration, the first-person voice is used to indicate where explication and reflection are used as a tool of analysis.

230

Table 10.1 Reported Psychosocial Wellbeing Themes with Rate of Endorsement Psychosocial wellbeing outcome themes

Number of participants who reported this theme

Percentage of sample who reported this theme

Keys to Success (9 Participants) Social connection Outcomes extended beyond KTS class School engagement Engagement with learning Affect Prosocial behaviour Self-efficacy Average rate of endorsement

4 2 8 4 3 2 5

44.44% 22.22% 88.89% 44.44% 33.33% 22.22% 55.56%

4.00

44.44%

Bereavement Group (5 Participants) Social Connection Social Support Letting it out Dealing with bereavement I can remember the good things Moving on Affect Average rate of endorsement

5 4 2 4 3 2 5

100.00% 80.00% 40.00% 80.00% 60.00% 40.00% 100.00%

3.57

71.43%

Group Attributes The attributes of participant groups in each program seemed to have a significant impact on the benefits reported. These included the amount of participants in a group, whether or not they were considered as being “at-risk”, and the different ways in which they engaged with the program. Group size. Smaller participant numbers appeared to equate to more consistently reported benefits. On face value, I deduced this by comparing group size against the endorsement level of benefits reported by each group. The BG program had a total of six participating members, and reported a number of consistent benefits across the group. The intervention class in the KTS program had a total of 20 participants (in the program, not the study), and reported benefits less consistently (see Table 10.1). At a correlational level, this suggests smaller groups report benefits more consistently. Evidence for a causal relationship between these factors is evident in BG interview data. Codes in the “Perceptions of the program” category (see Table 8.1, Chapter 8) included responses referring to the advantages of a small group. These 231

included feelings of being in a private space, and having equal levels of program involvement. Interview data also suggested having a smaller group facilitated increased social connection and social support, the sense of having a shared goal, and increased program engagement. For me, these codes and interview responses suggested that many of the benefits experienced by the group were due to social and therapeutic processes enabled by the smaller group size. That the same levels of social connection and support were not reported in the larger KTS group supports this notion. The idea that a relationship exists between group size and the benefits reported by BG participants is reinforced by literature. Previous researchers have noted that working with bigger groups can be problematic (Rickard, Bambrick, et al., 2012), and programs with less participants may be more effective (Rickson & Watkins, 2003). This notion is also supported from a theoretical standpoint. Music therapy literature claims effective group programs should recognise and adapt to the strengths of each member (Rolvsjord, 2010), while also addressing individual (Oldfield, 2006) and group needs (Carter & Oldfield, 2002; Nicholls, 2002), thus indicating the benefits of smaller groups. Hartford supports this idea from a social work perspective, suggesting “group[s] must be small enough for each person to be heard and to contribute, and also to feel the impact of the group upon his beliefs and behaviors [sic]” (1971, p. 162). Offering a practical example, Bullard (2011) observes that participants in her music therapy study reported the small group size – of between four and six members – played a central role in facilitating psychosocial benefits. She also suggests groups including 12 or more participants are unlikely to promote the same outcomes. This observation is buttressed by psychotherapy literature, which suggests the optimal size of therapeutic groups should range between seven and eight when working with general populations (Yalom, 1995) and between five and nine when working with young people (Malekoff, 2014). This literature supports my interpretation that the larger size of the KTS group impeded participants’ ability to experience psychosocial wellbeing benefits, which in turn limited the ability for the KTS study to report related outcomes. Subsequently, the large KTS group size presented a challenge to investigating the psychosocial wellbeing benefits of musical participation. At-risk status of participants. The fact that BG participants were “at-risk” also appeared to impact reported benefits. At-risk youth are often defined in educational terms, including low academic 232

performance or school engagement (Baker, 1999; Wang & Gordon, 1994). Other definitions incorporate behavioural problems, low socioeconomic status (Sagor & Cox, 2004), or experiences of conflict at school or home (Darrow et al., 2009). A wider definition is used here to include young people experiencing difficulties and or disadvantages that may affect any facet of development and or future quality of life (Batten & Russell, 1995; Robinson, 2004). Given BG participants were experiencing loss or grief, a reading of related literature suggests they were at-risk in several areas. This includes psychological factors such as depression, anxiety, and poor global functioning (Kaplow, Saunders, Angold, & Costello, 2010), as well as areas of adolescent development, including identity formation (McFerran, 2010b) and peer relationships (McFerran & Hunt, 2008). Interestingly, it was in these general areas that BG participants reported the most benefits (see Table 10.1). This suggests BG members reported benefits that related primarily to their at-risk status. The KTS group was a “mainstream” class. Following the Australian Bureau of Statistics (ABS, 2013) definition, this indicates they were not identified as having mental or physical disabilities or impairments, difficulties in learning, or problems in emotional, social, or behavioural areas. Upon reflection, this questioned the expectation they would report observable improvements in wellbeing. For example, when one student was asked if they had changed as a result of the program, they responded “no”, and went on to suggest they were still as happy as they were when they started the program. This was a clear indicator for me that many of these students were already experiencing relatively high levels of psychosocial wellbeing prior to participation; which is perhaps unsurprising given these students were in a “mainstream” class. This suggested to me that, given they were classified as at-risk, the BG participants (as a group) were more likely to both experience and report outcomes in this area. That at-risk students are more likely to report positive outcomes is also visible from differences in benefits reported within the KTS group. Those who described benefits in areas of school engagement or social connection also indicated these were problematic areas for them. Conversely, those who reported no benefits in areas such as social connection claimed high levels of social connection before the program started. These comparisons imply the at-risk status of participants was a prerequisite for the benefits reported. This is consistent with previous claims that at-risk participants are more likely to experience psychosocial wellbeing benefits from musical participation 233

than mainstream populations (Shields, 2001; Teachout, 2005). Further support comes from prior research in this area which credit their statistically insignificant results to students’ high baseline scores on psychosocial wellbeing constructs, indicting they were not at-risk (Rickard, Appelman, et al., 2012; Rickard, Bambrick, et al., 2012). Participant engagement. The way students engaged with music programs appeared to play an important role in how and when benefits were reported. This was observed in both commitment to and understanding of programs. Participant understanding of program and its purpose. Participant understandings of their respective music programs varied significantly between studies. This included understandings of program purpose, intended outcomes, and the role of music in achieving these outcomes. Participants in the BG had a clear idea of what the program was for. All interviewees quickly and easily articulated that sessions were designed and facilitated to help them deal with issues of loss and bereavement. They indicated the program had achieved its goal of helping them find ways to manage or move beyond feelings of frustration, loss or anger. They also explained that musical activities and processes helped in achieving these goals by providing a safe and sometimes fun way to approach confronting issues, helping express hard feelings, and connecting with others. The same understanding was not evident in KTS interviews. While some students linked the program to facilitators’ objectives (i.e. learning about “the five Keys to Success”) most claimed it had a different purpose. This included music education, musical mastery, increasing academic achievement in other curriculum areas, or just fun. Others struggled to articulate any purpose, or offered one unrelated to any element of the program. These participants also struggled to articulate the role of music in their program. While some did identify its intended role (i.e. to help deliver learning material related to “the five Keys”), others suggested it was purely recreational; a substitute for regular work; to get them out of class; to make class more fun; or that it played no role. I interpreted the difference in understanding between these groups as being connected to their level of engagement in the program. Further, it appeared that these factors influenced the benefits reported by each. The lack of understanding among many KTS students may explain why they didn't engage with the program, and why it didn't

234

have an impact on them in the intended areas. The understanding reported by BG participants, however, seemed to indicate a higher level of engagement in the program, and account for the fact reported benefits were closely linked to program goals. Researchers note the need to facilitate participant understanding and engagement in music programs, both at the start and throughout the life of a program. This requires consultation with students before commencing music programs (Grimmett et al., 2010). It also entails listening to and recognising student needs to establish a collaborative understanding and set of goals for the program which are negotiated throughout its life (Bolger, 2013). Failure to do so is said to result in a lack of engagement, and a lack of reported outcomes (Grimmett et al., 2010). Participant commitment to program. Participants’ level of commitment also differed between groups. During interviews, BG responses suggested high levels of commitment to both the goals of the program and participation in general. Participants expressed the importance of supporting their fellow group members during sessions, and several explained they had come to understand how music could be used to support their wellbeing. Appropriation of these mechanisms to address group goals outside of sessions was also reported, including supporting fellow group members, or using music for emotional regulation in everyday life. Some members also reported making sacrifices to attend the group, such as missing classes they wanted to attend. For me, these reports revealed that these group members not only understood the goals and content of the program, they were committed to and actively engaged in achieving them. The main indicators of commitment in the KTS group were reports that students looked forward to sessions. One student claimed he was more likely to come to school on program days, and others stated the program made them want to be in class more. However, students suggested this was primarily because it made class fun, not because they were committed to engaging in the program goals or content. In fact, many students struggled to articulate what the goals were, suggesting to me they were not highly committed to addressing them. With the exception of a small percentage of students, it emerged that the main goal of participation was a distraction from regular classwork. Further, some students reported that they “had” to participate in music program activities (see Table 7.3, Chapter 7). This led me to the conclusion that some students saw participation as a class requirement (much like tests or assignments), not 235

something they wanted or chose to do. 12 Reflecting on these responses, it seemed to me these students were not committed to participation, but rather felt obliged to do so. The difference in commitment described here appeared to influence the types of benefits reported by each group, and their level of endorsement within said groups. Benefits reported by BG participants related to the purpose of their program, and were endorsed strongly among the group. Benefits reported by KTS participants were less consistent with program goals, and endorsed less consistently within the group (see Table 10.1). This seems to suggest that lower engagement with and commitment to a program leads to decreased reporting of related outcomes. These observations align with previous studies that suggest music programs must be sufficiently engaging in order for intended benefits to be reported (Grimmett et al., 2010; Hallam, 2010; Rickard, Bambrick, et al., 2012). Other literature also suggests students more highly engaged with musical activities are more likely to psychosocial wellbeing benefits from their participation (Chin & Rickard, 2012). Program Attributes The elements of the program, or intervention, itself also appeared to affect the benefits reported in each study. These elements included how and why programs were delivered, who delivered them, and the different activities undertaken in each. Setting. The BG participants linked the private setting of their program to many reported benefits. As indicated by codes in the “private group setting” and “gave us the place and space” themes (see Table 8.1, Chapter 8) students suggested this setting gave them a safe and confidential environment where they could express themselves and connect with others like them. They also reported it provided an environment in which everyone could have their voice heard. This setting seemed particularly important given program content, and appeared to facilitate many of the benefits they reported (see Table 10.1). The classroom setting of the KTS group appeared to restrict the reporting of psychosocial wellbeing benefits. The more emotionally based benefits reported by the BG were either less endorsed or not reported by KTS participants (see Table 10.1).

12

This is possible as students were not consulted about the implementation of

music in their class. This decision was made by staff, and enacted on their behalf. 236

Instead, reported benefits were largely education related, and not consistently endorsed. I interpreted this to mean the classroom was not an optimal setting for addressing sensitive psychosocial wellbeing issues. Potentially, this is because activities delivered in this setting are not sufficiently removed from the routine, structure or regimented nature of regular classes to inspire engagement with issues of psychosocial wellbeing. While several students made reference to the novel relief the program provided from regular class activities, it was largely referred to as a curriculum activity, where participants “had” to do things related to “learning”. This suggests participants did not make a clear distinction between the program and regular classes, with some suggesting it made no difference to how they felt about attending school or class. This sentiment is best summed up by one boy who, when asked whether the program changed how he felt about being in class, answered that it made no difference because it was “still class”. The lack of confidentiality in the classroom setting may also explain the difference in reported benefits. As noted by one BG participant, the ability to divulge personal information and address sensitive issues with the confidence other group members “won’t go out and blab or laugh at you” was important. Perceivably, this same level of confidence is not easily engendered in classrooms, a setting associated with bullying both in Australia (Cross et al., 2011) and internationally (Smith, 2010). For me this shows that, while music programs delivered in classrooms may aid in education related benefits, the setting does not necessarily promote program engagement, nor provide an appropriate setting for addressing psychosocial wellbeing. A preference for private settings when targeting psychosocial wellbeing outcomes is also expressed in the literature. Some authors argue catering programs to the needs of each individual in a classroom setting is problematic (Pitts & Davidson, 2000; Rickard, Bambrick, et al., 2012). Others suggest classroom sessions become disruptive, and advocate for more controlled (Shields, 2001) or structured settings (Montello & Coons, 1998), such as private groups. Rickard, Bambrick, et al. (2012) suggest music programs delivered in classrooms are perceived as curriculum activities, and should instead be offered as extra-curricula activities to achieve observable psychosocial wellbeing benefits. These claims are further supported by the benefits observed in studies that did deliver programs both as private groups, and out of the classroom (Baker & Jones, 2006; Cheong-Clinch, 2009; Choi, 2010; Devroop, 2012; Gooding, 2011; Jones et al., 2004; Kim et al., 2006; McFerran & Teggelove, 2011; McFerran et al., 2010; Rickson

237

& Watkins, 2003). In summation, it would appear that unsuitable settings (i.e. classrooms) pose a serious challenge to investigation in this area. Program length. Length of program was reported as having a potential impact in both studies. One student in the BG program suggested the program was too short, and that he would have liked to see more sessions in future programs. This is reflected in reports from other participants that, while they did experience benefits, they were not absolute (as shown in the codes “I'm still not coping with it, but it’s helped”, “it’s made me at least a bit better”, “it’s not over”). The teacher in the KTS program also noted the program was not long enough. He claimed some benefits need time to be fostered and or emerge in a way that is observable, particularly for disengaged students. He also suggested that more time was needed to explain to students what the program was trying to achieve, what the purpose of the music was, and why it was being investigated. He went on to claim that without sufficient time spent in these areas, students would find it hard to engage in the project. Several authors make arguments for longer programs (Baker & Jones, 2006). These include claims that programs of seven (Michel & Farrell, 1973) and 10 weeks (Knox-Anderson & Rickard, 2007) are too short to bring about significant changes in measures of psychosocial wellbeing. Montello and Coons (1998) note some benefits reported in their study were only evident in the last week of a 12-week program, suggesting shortening the program by one week would have led to an absence of these reported outcomes. This is consistent with music therapy literature. Investigators assessing the impact of program length, or “dose” of treatment, in work with disabled children suggest longer “doses” of a music program are linked to more observable benefits (Gold, Solli, Krüger, & Lie, 2009). Some go so far as to suggest five months is the optimal dose for detectable improvements in social functioning (Geretsegger, Holck, & Gold, 2012). Consequently, it seems investigating music programs of short length likely prevents investigators from attaining positive findings. Program purpose. Another element that appeared important was the underlying purpose, goal, or intention of the two programs. Again this was seen to impact both the type of benefits reported, and their endorsement by the group.

238

Wellbeing versus curriculum focus. The two music programs had quite different purposes. The BG had a targeted wellbeing focus, centred on addressing bereavement related issues through therapeutic processes. The KTS, however, had a quasi-wellbeing focus. Less concerned with directly targeting wellbeing in a therapeutic sense, it focused more on teaching about the five “Keys” through the delivery of curriculum content. Further, although the facilitator was a music therapist, the aim of music program was to explore how effective musical processes were in delivering this primarily scholastic content. The first impact of this difference can be seen in the types of benefits reported. As mentioned, the BG reported mostly emotional, psychological and social benefits, while benefits reported by the KTS were largely focused on education. The second impact was the rate of benefit endorsement. As Table 10.1 shows, psychosocial wellbeing benefits were reported more consistently for the BG’s wellbeing program than the education or curriculum-based program used in the KTS. This observation is supported when comparing outcomes reported in previous studies. Those investigating the outcomes of therapeutic or wellbeing focused programs often observed reports of increased psychosocial wellbeing (Baker & Jones, 2006; Cheong-Clinch, 2009; Choi, 2010; Gooding, 2011; Jones et al., 2004; Kim et al., 2006; McFerran, 2010b; McKegg et al., 2012; Vaughan et al., 2011). Those investigating education-focused programs, however, have found less consistent reports of psychosocial wellbeing outcomes (Grimmett et al., 2010; Knox-Anderson & Rickard, 2007; Rickard, Appelman, et al., 2012; Rickard, Bambrick, et al., 2012). This divide is further reflected in Schellenberg’s (2004) investigation regarding the effects of participation in music lessons. While no psychosocial wellbeing benefits were reported, improved academic or cognitive development (IQ) was. Targeted to the needs of a coherent group. A further dimension of program purpose was the relationship between program intentions and the group to whom programs were delivered. The BG program was facilitated for the express purpose of addressing bereavement, and directly targeted a group of bereaved students. The KTS, on the other hand, focused on a wide range of wellbeing or developmental constructs and delivered them to a heterogeneous group of Year 7 students brought together through school-administration procedures. As the principal and teachers reported, the sample was culturally diverse, at different levels of

239

development, and from my observations, had little in common apart from being in the same class-group. While it is perceivable that individuals within this class had a range of psychosocial wellbeing needs, these were likely to be as diverse as the group itself. Further, without identifying any of these needs, it was impossible to know if they would be addressed by the content of the music program. Again, this divide showed in the data. BG participants reported benefits consistent with their shared needs as a bereaved group of teenagers and the focus of the program. KTS students, however, reported more diverse benefits. While more or less related to learning, or the education environment, reported benefits were also sporadically endorsed. This is consistent with he KTS group teacher’s suggestion that programs need to target explicit or desired outcomes, or the needs of a specific group. This suggests programs that focus on the needs of a specific group, and deliver content designed to meet these needs, are more likely to see reported benefits. The need to target a program to the shared needs of a coherent group is also supported in the literature (Gooding, 2011; Rickson & Watkins, 2003). Facilitation. The facilitation or delivery of a program also appeared to have an impact on the reporting of outcomes. This included who facilitated the program, and the delivery method. Facilitator. Participants in each program noted the importance of having an experienced facilitator. Students sang the praises of the music therapist. The KTS teacher also made specific reference to her skills and competencies, such as being able to engage with the students on their level, and linked these factors to benefits reported for that group. The importance of having a quality facilitator is noted in the literature. Grimmett et al. (2010) suggest the lack of quality facilitators was a limitation in their programs, and led to a lack of engagement and observable outcomes. Rusinek (2008) suggests that while the experience and skills of the facilitator in their study were not sufficient in themselves to achieve the program benefits observed, they were a necessary element.

240

Participatory and democratic approach. The participatory and democratic approach taken in programs appeared to foster benefits reported in both studies. However, while evident in both programs, these elements were more prevalent in the BG. While the BG program followed a predetermined structure, participants were given the opportunity to have their say and influence program structure through ongoing consultation. This included the activities undertaken, and the focus and direction of group discussions. This follows a participatory approach to musical participation often described in community music therapy (Stige & Aarø, 2012) as empowering participants to be active agents in their own learning and pursuit for health and wellbeing (Bolger, 2013). Evidence of this approach is reflected in BG participant responses. Students noted that all participants had an equal opportunity to have their voice heard and have a say in how they participated. Also, students reported that activities such as songwriting were group led, including the choice of backing track, lyrics and themes addressed. This suggested to me a sense of ownership of both the program, and the musical processes among students. It also appeared to foster program engagement and the appropriation of musical processes as wellbeing strategies. Both participatory and democratic elements were also reported in KTS data. Students reported the fact everyone had a role to play in the music video was important, as evidenced by the “I had a part to play theme”. They also made consistent reference to choosing the songs to remix for their video, that they wrote and sang the new lyrics, and that they made the video as a group (as described in the “we chose the music”, “we did it”, and “we did it together” themes). The teacher also claimed the fact that students “created” and sang the music was important. However, students also reported, “we had to do it”, indicating the sense of democracy was less prevalent than in the BG group. For me, this linked to the limited choice offered to them; they were instructed to make a music video about bullying, and given only two backing tracks to choose from. The influence of this reduced choice was, for me, further reflected in reports that the musical elements of the program had no beneficial impacts for some students. The need to provide young people opportunities for input in music programs when targeting wellbeing is supported in music therapy literature. McFerran (2010a) claims that group facilitation which allows and actively encourages participatory and democratic involvement is important for fostering engagement, ownership and a sense of control for young people in music programs (as well as engendering subsequent 241

wellbeing benefits). She also notes this model works better when participation is voluntary, and less focus is placed on structuring sessions. Fostering musical appropriation. One aspect that appeared critical to the reporting of benefits was how programs affected a participants’ relationship to music, and their perceptions of the role it can play in their everyday life. While KTS participants largely saw music as something that was for fun, a distraction, or a skilled activity, BG participants appeared to gain an indepth understanding of how they could appropriate music for use in their daily lives. I interpreted that this appreciation of how music can be appropriated was linked to the way music was presented and used in each program. In the KTS program, the focus remained on using music as a curriculum tool, or vehicle to present and engage with certain ideas or curriculum content. As such, perceptions of music in this group related mostly to education, making school fun, or creating a distraction from the monotony of regular class. There were, however, some cases in which students reported an increased understanding of how music can be used to regulate emotion and engage in learning. These instances appeared related to some of the most remarkable reported benefits, such as “music calms me down” and “music is my way of learning”. That such reports were not more widespread appeared connected to the focus of the program and its reduced emphasis on helping students learn how to appropriate music for wellbeing. Participants from the BG offered consistent reports of how the program and the facilitator had taught them new ways to think about and use music in their lives. This led to new and enhanced relationships with music and was linked to some of the most noteworthy benefits reported by the group. This was reflected in themes such as “how music helped” (including the codes “music can make you feel better” and “music can help you deal with things) and “change in musical relationship” (including the themes “now I know how music can help me” “I understand music better now”). In turn, these led to reported benefits that were linked to music, such as “got me back on track”, “helps me deal with problems”, “helps me get through the day” and “helps me sleep”. Here we can see that delivering programs in a way that both capitalises on known uses of music to address psychosocial wellbeing, and empowers young people to use music in this way, is linked to the reporting of significant benefits. This is consistent with DeNora’s suggestion that the effect music has on a person depends on how they

242

appropriate it; “This process of appropriation is what consolidates and specifies music’s force” (2000, p. 43). Activities. Activities undertaken in sessions also appeared to influence reported benefits. These included activities that were linked to more consistent reporting of benefits generally, and specific activities that I interpreted to be linked to particular benefits. Group activities. Social connection. Social connection, a benefit reported in both studies, emerged as being closely related to group activities. In the KTS study, responses that made up the “social connection” theme constantly referenced the communal nature of activities undertaken in the program, embodied in codes such as “we did more together”. Some students described how group musical activities enabled them to engage more with classmates, including increased “communication”, “getting along” with others, and initiating friendships. The teacher also reported group activities enabled particularly disengaged students to become more involved with the class, and helped students learn to cooperate and come together as a group more generally Participants in the BG study also noted group activities as important for social connection. This included expressing feelings, emotions, and experiences with the group through song sharing, group discussion, and instrumental improvisation. They suggested such activities meant “we know each other better now”, and had the “chance to connect with others” on a deeper level through sharing their own experiences and learning about what others were going through. The more personal nature of these group activities also appeared to lead to increased “social support”. The ability for musical participation to foster social connections through both musical and social processes, including group activities, has long been recognised (Schütz, 1951). More recently this has been linked specifically to participatory musicmaking activities facilitated with a given community or group, which in turn foster a sense of belonging or connectedness among that group (Ansdell, 2002). This premise is supported in work with students, with reports group music activities can foster connection between group members in school settings (Jones et al., 2004).

243

Communication. Increased communication also related to group activities. For KTS students, the fact group activities brought students together in a setting that necessitated cooperation was reported to increase communication between members. Likewise BG participants suggested that song sharing and song creation activities undertaken as a group also facilitated communication between members, which in some cases led to increased social interaction and support outside of the group setting. There did, however, appear to be important differences between types of communication reported in each study. KTS participants appeared to be describing increased communication through increased social contact, and the necessity of cooperation (i.e. the need to communicate to undertake music related tasks such as planning, production, and performance). BG participants reported this type of communication too, yet also reported another form that related to the communication of feelings and ideas through musical performance and song sharing. This second form can be described as the non-verbal communication of thoughts and emotion through the use of musical elements such as tone, rhythm, tempo and timbre, as well as the lyrical content of a recording. The link between group activities and communication described by KTS group is supported by Jones et al. (2004), who suggested communal music activities increased communication among a group of boys. The non-verbal communication reported by the BG group has been reported in psychology (Juslin & Laukka, 2003; Olteţeanu, 2010) and music therapy (Bunt & Stige, 2014; Gfeller, 2002) literature, both of which state the ability for music to act as a vehicle for expression among group members. Self-efficacy. Group activities were also credited for the self-efficacy benefits reported by some KTS participants. Those who reported this benefit suggested the process of getting up in front of the class to either perform or show their videos to others increased their confidence. This was supported in the teacher’s interview. The link between musical participation and increased self-regard is supported by Clendenon-Wallen (1991) who found music activities increased both self-confidence and self-esteem in sexually abused adolescents. Further links have been made between self-esteem and musical participation, yet these have related primarily to private music instruction (Costa-Giomi, 2004).

244

Instrumental activities. For BG participants, instrumental activities were reportedly linked to having fun and being able to communicate emotions – as captured in codes such as “instrumental work was fun”, and “music for expression and sharing”. For KTS participants, instrumental activities were also reportedly linked to having fun, and giving students an opportunity to play a role in the group through a chosen or assigned instrument – evident in codes such as “there was something for everyone” and “I got to do my part”. The link between using instruments and having fun for bereaved teenagers has been reported previously, and noted as a valuable tool when working in this area. It has been described as providing both respite from dealing with confronting content, and a safe platform to engage with this content (McFerran, 2000). Importantly, musical instruments were used in the BG for improvisation, rather than developing instrumental skill or a musical product. McFerran (2010a) notes this an important distinction as it takes the pressure off trying to create music that is “significant” in technical skill or aesthetic value, allowing instead for the fun and freedom of free musical expression. Instrument use described by KTS participants also referred to similar improvisational sessions. In this case, however, it also provided a way for students to participate in musical activities that were directed towards the creation of a musical artefact. Using instruments in this way has also been noted by McFerran (2010a) as a less confronting way to participate without having to focus on singing. These reports suggest using musical instruments in sessions does not necessarily need to focus on musicianship. Rather, they can be used for fun, while also expressing emotions and offering opportunities for non-threatening participation in musical activities. Creating a musical product. The creation of a musical product was seen as important for the KTS group. This was articulated by students in terms of ownership and self-confidence, represented by codes such as, “we wrote it”, “we made it”, and “no one else did it, we did”. The teacher also reported the relationship between these factors through codes such as “we did that” and “they can feel they succeeded in something”. This suggests both creativity and ownership were important feelings or experiences engendered through song creation. This is consistent with reports that making a recording of songs written during music sessions can be a positive experience,

245

which is linked to a sense of ownership of the musical process and pride in the creativity involved (Grocke, Bloch, & Castle, 2009). Fun activities. The idea of having fun was consistently linked to reported benefits for each group. This is supported by previous research that suggests musical activities need to sufficiently fun for benefits to be reported in mainstream school settings (Hallam, 2010; Rickard, Bambrick, et al., 2012). However, while this link could arguably be related to all benefits, some outcomes received more endorsement for this link than others. Engagement with school, class and learning. For the KTS group, fun experienced through musical participation was credited for reports of engagement with school, class and learning. This was evident in codes such as “it’s more fun than just writing”, “being in class was better”, “class was more fun”, “music makes studies more fun and easy”, and “enjoyed being at school more”. These observations reflect author claims that engagement in musical (Rusinek, 2008) or other artistic (Ewing, 2010) activities can facilitate enjoyment in school and class activities. Such authors also suggest experiences of fun and enjoyment can be an effective way of motivating students to engage in educational contexts. Dealing with bereavement. Having activities that were fun were seen as important for the BG group when dealing with issues of bereavement. This was informed by my interpretation that fun involved in some music activities provided a way for students to address issues of grief and loss that were less direct and confronting than simply talking about it. I also perceived that fun music activities were able to help take the focus off the negative or challenging aspects of bereavement. Such interpretations were underpinned by data in codes such as “music provides a safe setting”, “the music was fun”, and “trying not to make it all sad”. The benefits of using music in this way were especially evident in instrumental and songwriting sessions. The idea that music can produce a fun and safe environment in which to address issues of grief and bereavement is supported by previous investigations of grief focused music programs. The findings of these investigations suggest music can provide a safe basis from which to explore challenging concepts such as loss (McFerran et al., 2010).

246

Challenging activities. Interview data suggested many benefits reported by the BG necessitated a certain level of challenge or discomfort. Participants often made reference to the fact that sessions and session activities were challenging. One student even noted being apprehensive about attending the group, as he knew it meant confronting difficult feelings. The same boy, however, said this never stopped him from attending, and that he knew it would be worth it. This was a common feeling reported by these students. Participation was not always seen as easy, and while many reported challenging experiences, they acknowledged that it was these challenging experiences that often led to benefits. This is reflected in the theme “hard but worth it” and codes such as “letting the feelings out”, “its ok to cry”, “some parts were hard but worth it”, “sharing songs can be hard”, “talking about feelings was difficult”. This accords with McFerran’s (2010b) suggestion that working with groups of bereaved teenagers may necessarily involve sustained group engagement with individual experiences of loss. While challenging, these more difficult activities can be seen as a necessary stage of participation through which more honest, constructive and finally comfortable understandings of loss are achieved. Non-musical activities. Other aspects of the music program noted as important by some participants included those that weren’t musical at all. This was most notable in the KTS group, in which some students reported the activities most rewarding or beneficial for them included directing and acting in the music video. This was supported by the teacher, who also noted that some students appeared to benefit from the fact they did not have to perform a musical task, but had the chance to undertake activities such as filming. He noted this was important for these students, as they were not confident or comfortable assuming musical roles. However the idea they could participate through non-musical activities was powerful for them. This suggested to me that sometimes it is not the “music” of a musical program that affects an individual, but the opportunities they afford for recognition and participation in group activities. This is an important distinction as it places equal focus on the “processes” and “content” of musical programs, rather than musical “content” alone. Importantly, this does not mean the acting, directing or filming were not considered musical participation. Following Small’s (1998) concept of “musicking”,

247

musical participation constitutes any activity associated with music, from passive listening to selling tickets for a concert. Musical skills not required. Despite acknowledging the benefits of musical activities, some students noted that being musical was not a prerequisite for participation in the program, or reaping the benefits reported. This was most evident for BG participants, who noted a complete lack of experience or skill did not influence their experience of instrumental activities. Specifically, one student stated they didn't like the drums, and another that they had never played them. Both, however, reported their experiences in drumming sessions were thoroughly enjoyable and rewarding. They claimed the process was both fun, and that it showed them how musical improvisation can be used to express emotion. It appeared the facilitator’s ability to shift focus away from musical mastery or skill, and towards the use of instruments for therapeutic purposes, enabled these students to both engage in and benefit from musical participation. This mirrors a music therapy approach described and advocated by McFerran (2010a) in which musical activities do not focus on musical skill. She suggests goals such as musical perfection or mastery can create unnecessary pressure, which distracts from program goals or other musical processes such as authentic expression which are beneficial for facilitating wellbeing outcomes. Conclusion A fitting summary of the above findings is offered by Pavlicevic, in that the “power” of music is not a “given”, but something that needs to be “evoked, invoked, crafted, shared, and worked with in a very particular way” (2010, p. 99). When aiming to “invoke” reports of increased psychosocial wellbeing, it appears school-based music programs require a number of elements. Not only is it important to consider the characteristics of participant groups, but also the focus of programs, their intent, activates, their facilitator, and how they are delivered. Consciously fostering student engagement, and focusing on the appropriation of music as a resource rather than a skill, also appears vital in this area. There also appear to be links between different musical activities and different experiences and or benefits reported by students. These elements, and how they varied between the KTS and BG programs, were shown to account for the difference in benefits reported for the two music programs.

248

This was shown for both the type of benefits reported, and the consistency with which they were reported within groups. Several shared elements were also identified between programs. Some, such as program length, were reported to restrict the reporting of benefits in both programs, while others, such as facilitator, were shown to support reporting for both. This analysis also suggests different musical activities or experiences are suited for different purposes. For example, challenging experiences appear to be useful, or even necessary for addressing experiences of grief and loss, while creating a musical product can create feelings of pride and self-confidence. Other elements appeared to be important for several reported outcomes, including fun and group activities. Importantly, however, the presence of one or more of these elements or activities does not guarantee benefits will be reported – as was seen by the inconsistent reports of class engagement for the KTS group. Instead, it would appear that a combination of different factors is needed to affect such reports. This evidence necessarily challenges previous assumptions that any and all school-based musical participation will lead to reported psychosocial wellbeing benefits (Australian Government, 2005). Specifically, it suggests programs that lack a wellbeing focus; are insufficiently engaging; have an academic or educational focus; are delivered in classrooms; are too big; or do not encourage the appropriation of music as a resource, are less likely to promote psychosocial wellbeing. This offers a potential explanation for the lack of existing evidence supporting the link between class-based music instruction and increased psychosocial wellbeing (Knox-Anderson & Rickard, 2007; Schellenberg, 2004), and the inconsistent evidence linking psychosocial wellbeing and school-based musical participation more generally (Gill & Rickard, 2012). The presentation of different elements of school-based musical programs, and their relationship to reported psychosocial wellbeing, intends to provide a point of reference for future investigations in this area. Importantly, while many of the elements described illustrate how they can pose challenges to such investigations, they also serve to illustrate which elements can be beneficial. It is not assumed the elements of musical participation identified here are a comprehensive list of those that affect research in this area. Yet, they do provide an important foundation upon which a deeper understanding and recognition of the importance of such factors can be built. The findings presented in this chapter have implications for the design and implementation of future school-based 249

music programs, as well as education and arts-based policies aimed at supporting psychosocial wellbeing through musical participation in mainstream schools. Table 10.2 Music Program Elements that Affect Investigations of Psychosocial Wellbeing Area Group attributes

Program attributes

Element Group size At-risk status of participants Participant engagement Setting Program length Program purpose Facilitation

Activities

Sub element

- Participant understanding of program and its purpose - Participant commitment to program

- Wellbeing versus curriculum focus - Targeted to the needs of a coherent group - Facilitator - Participatory and democratic approach - Fostering musical appropriation - Group activities - Instrumental activities - Creating a musical product - Fun activities - Challenging activities - Non-musical activities - Musical skills not required

250

Chapter 11: Recommendations Article This chapter presents the recommendations that arose from the results of the two critical reflection analyses. These recommendations have been published, and therefore the PDF file of the published article has been inserted directly into the dissertation, and starts from the next page. This includes the entire article as it appears in the journal, as per the University of Melbourne’s thesis with publication guidelines.

Article reference: Crooke, A. H. D., & McFerran, K. S. (2014). Recommendations for the investigation and delivery of music programs aimed at achieving psychosocial wellbeing benefits in mainstream schools. Australian Journal of Music Education(1), 2749.

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

Chapter 12: Discussion This chapter provides a broad discussion of the primary findings from this project. The initial aim of this PhD was to provide clear empirical evidence for how music could be used in mainstream schools to address specific policy goals. As the research progressed, however, it became clear that before this initial goal could be reached, there needed to be a critical review of the research approaches used in this area. Subsequently, the main focus of the project turned to identifying the challenges of using established research approaches to investigate a link between psychosocial wellbeing and musical participation in schools, and making recommendations to meet these challenges. The implications of these challenges and recommendations are discussed here, with specific reference to this area of research, and social science more generally. Despite this methodological focus, however, the initial policy dimension has remained an important part of the thesis. Therefore, a wider discussion is undertaken towards the end of this chapter that considers the implications of the present findings for policy. The Challenges of Existing Research Approaches The challenges of investigating the psychosocial wellbeing benefits of musical participation in mainstream schools identified in this thesis are both numerous and complex. As outlined in the results of the CRA presented in Chapter 9, many of these relate to the study designs and research methods used in this area. As the CRA presented in Chapter 10 shows, particular aspects of the music programs investigated may also pose several complications for reporting consistent or positive results. Importantly, it is not proposed that the challenges reported in this dissertation are an exhaustive account of all those which may be faced by researchers in this field. Rather, they represent those that were observed in the two studies critically analysed in this project. However, it is argued that failure to consider and address these challenges in any future study will undermine the ability for investigators to provide positive, consistent, or legitimate evidence for a link between psychosocial wellbeing and schoolbased musical participation. It is also proposed challenges identified in this dissertation illustrate the presence of several important assumptions that have underpinned existing research in this area. These assumptions include the suitability of using certain quantitative reporting methods, and the idea that class-based music education will achieve psychosocial 275

wellbeing. Seemingly, these assumptions not only account for the inconsistency of existing findings in this field, but also undermine the legitimacy of several studies that have reported positive results. For example, the results of this PhD can be used to contest the legitimacy of findings reported by authors such as Vaughan et al. (2011). Vaughan et al. asked Year 5 and 6 students (an approximate age-range of nine to 11 years old) to complete a survey with 53 statements, including: “During the past six months, I have felt so hopeless and down almost every day for one week that I have stopped doing my usual activities” (2011, p. 23). Reported results for this item were both positive and statistically significant. However, the challenge of student comprehension reported in Chapter 9 (see “The Challenges of Comprehension”) seriously questions whether students understood the language in this item. Yet, authors maintain results from this item were valid, thus signifying an assumption that students of this age are capable of understanding and responding legitimately to such a question. This thesis has deconstructed this, and other assumptions, showing them to be unwarranted and counterproductive in the context of school-based research. The validity of Vaughn et al.’s findings can also be questioned given the length of their survey, the timing of its administration and the location of the program delivery. That positive results like those reported by Vaughn et al. are derived from research approaches that contain many of the challenges identified in this project, underlines the critical importance of addressing them. Given such approaches have been shown here as problematic, it cannot be assumed that the benefits they report will be reproduced by all school music programs. If policymakers implement school music programs based on such evidence, it seems unlikely such programs will always achieve the intended benefits. If this happens, policymakers may well become wary of claims regarding the benefits of music in schools. Given the tenuous relationship between policy and the Arts in Australian schools (Gattenhof, 2009), this has serious implications for ongoing advocacy in this area. At a point when we are closer than ever before to realising concrete policy backing for music in schools, it would seem a shame this realisation is undermined by spurious evidence. Now more than ever, it is critical to ensure results in this field are gathered using appropriate research approaches. At a more fundamental level, investigators in this field should also take the time to critically engage with the research process. As researchers in the social sciences, we have a responsibility not only to the participants from whom we collect our data, but also to the field or communities we hope to affect with our results. This responsibility 276

demands that we must aim to undertake best-practice research by considering the process in which knowledge is created, and the effect it is having on the world around us. Only then will we provide the best possible evidence, and thus give our young people the best possible chance to participate in, and receive the benefits from, schoolbased musical participation. Recommendations for Future Research Through critical engagement with existing research approaches, this dissertation formed recommendations (see Chapter 11) for the design of future studies. These focus on two key areas: research method and designs, and the design and delivery of music programs. It is argued here that researchers must follow recommendations in both of these broader areas if they hope to reach positive, consistent, and valid findings. Research method and design. Central to recommendations for research method and design is the need to ensure that quantitative methods (such as self-report surveys) are suitable for coping with the subtleties and complexities in this area of investigation. This includes using surveys with language suitable for student populations, and delivering them appropriately. Ideally, this requires undertaking exploratory qualitative research to identify the tools and modes of research that are best suited to the young people involved, and the school-based contexts in which this research takes place. Recommendations are also made for such qualitative studies. These include the need to consciously and purposefully engage young people in the research process. Underpinning this recommendation is the acknowledgement that we are asking students to provide us with data. It is their opinions and point of view that provide us with the information we use to make grand claims about their lives. Therefore, rather than treat them as objects of enquiry, significant time and effort should be dedicated to informing young people about the significance of their participation in research. Studies that fail to do this cannot assume participants will be sufficiently engaged to provide legitimate self-report data (regardless of the method used). In addition to quantitative and qualitative approaches, alternate perspectives from teachers or participant observers should be sought to gain broader insights into student experiences and research processes. Meaningful consultation with school communities is also critical to ensure research is carried out in a way that is appropriate

277

for the school context. The findings of this project suggest that studies that do not address these issues are in danger of contributing to the ongoing inconsistency of findings in this field. Type of musical participation investigated. Recommendations made in Chapter 11 also stress the need to consider the type of music program investigated. Vital to this group of recommendations is the argument that classroom-based music training and music education provide a very specific musical context for students that is not tailored to psychosocial wellbeing outcomes. Research suggests that these forms of participation may be suited to promoting some important benefits of school-based musical participation, such as intrinsic benefits, and extrinsic benefits related to cognitive and academic development (see Chapter 3). Yet, the findings of this project show that such forms of musical participation are not only unlikely to achieve psychosocial wellbeing benefits, but may actually be counterproductive to their acquisition. Put simply, if psychosocial wellbeing outcomes are the intended goal of a music program, it is critical that programs are designed and facilitated specifically for this end. It is not enough – and it is argued here, counterintuitive – to assume music programs designed for educational or strictly musical benefits will, by default, also achieve psychosocial wellbeing benefits. More specifically, it is recommended that programs should be designed and delivered to meet the shared wellbeing needs of discrete groups. They should also be delivered using democratic and participatory approaches, and by a skilled facilitator. Further, they should be delivered outside of the classroom, in a private or semi-private setting, and should ideally not be connected to curriculum content. It is argued that research that fails to adopt these strategies is in danger of wasting schools’, students’, and researchers’ time. Even if the latter use suitable research designs and methods, they are unlikely to demonstrate psychosocial benefits if they investigate programs not suited to achieving them. It seems continuing to investigate unsuitable music programs will also add to the inconclusive nature of existing findings in this area. Need for ongoing critical engagement with research. It is acknowledged that following each recommendation described in Chapter 11 can be difficult given the reality of conducting research in any setting. Issues such as budget, timing, staffing or ethics can all place limitations on what is achievable in any

278

study. Conducting research in schools adds another level of complexity due to established and often restrictive structures such as timetabling, school resourcing, and in some cases, resistance from school communities to musical participation (Crooke & McFerran, 2015). However, until the challenges to research identified here are addressed, we cannot claim the knowledge generated in this field is meaningful enough to make a conclusive statement about the link between music in schools and psychosocial wellbeing. Sociologist Ulrich Beck talks of “zombie categories” as “living dead categories, which blind the social sciences” (2002, p. 24). Going further, Ken Plumm describes them as “categories from the past that we continue to use even though the have long outlived their usefulness and even though they mask a different reality” (2011, p. 195). It is argued here that this provides an apt description of this area: the ongoing use of unsuitable research approaches has served to obscure the real potential for music to promote student psychosocial wellbeing in mainstream schools. It is critical we admit some of our modes, categories, or approaches have past their point of usefulness. Plumm goes on to suggest that in order to mitigate the impact of “zombie categories”, “our tools for theory and research [in the social sciences] need radical overhaul” (2011, p. 195). Nowhere is this statement more applicable or pertinent that in this field. It is precisely this “overhaul” that the challenges identified in this project demand, and it is precisely this “overhaul” that the recommendations articulated in this PhD aim to facilitate. It is crucial that, by critically appraising our “tools” (as this PhD has done), we move forward through future research to reach an informed understanding of the relationship between music and psychosocial wellbeing in schools. It is only then that we can claim to really “know” about the potential for music to benefit students in this area, and communicate this “knowing” with confidence to other academics, schools, and policymakers. Recommendations made in this PhD are not limited to this area. To ignore the challenges identified in this dissertation when conducting any research in schools (or any kind of humanistic enquiry) involves taking a giant leap of faith that the research approaches being used are appropriate. For researchers in any field to state with conviction that they are producing meaningful or legitimate knowledge, there is a need to critically appraise or revise how this knowledge is generated on an ongoing basis.

279

Policy Implications The findings of this PhD also have significant implications for the field of policy. The key contribution they hope to make in this area is to inspire and guide better quality research, which will in turn be better placed to produce evidence to inform policymakers. However, it also has more immediate implications. These relate to a critique of current policy documentation, and the suggestion that such documents should temper or revise the nature of their advocacy. Challenge to claims in current policy literature. As described in Chapter 2, recommendations that school-based music be supported by government funding and legislation have grown in strength and frequency within policy documentation over the last decade (Parliament of Victoria, 2013; Stevens & Stefanakis, 2014). While from a historical perspective this is a step forward in the struggle to secure a meaningful place for music in schools (Ewing, 2010), this thesis has shown some of the reasoning behind these recommendations are problematic. This can be understood on two levels. The first is the lack of evidence to support the basic claim that musical participation in schools will increase students’ psychosocial wellbeing. The second refers to the more specific recommendation, or assumption, that musical participation in the form of class-based music education will achieve this increase. This policy recommendation is not only unsupported by existing evidence, but appears to be in direct contradiction with it. Combined, the nature of these key messages suggest policy documentation in this area is both misguided and misinformed. Challenge to claims of psychosocial benefits. As it stands, there is not enough consistent evidence to support the basic claim made in key education (Parliament of Victoria, 2013) and arts (MCEETYA & CMC, 2007) policy documents that musical participation (in any form) will result in increased psychosocial wellbeing for students in mainstream schools. This is demonstrated by the inconsistency in published findings in this area, the questionable methods behind these findings, and the results reported in this dissertation. While the latter found some support for claims that musical participation can increase students’ psychosocial wellbeing (see Chapter 8), this was limited and overshadowed by evidence that, in many cases, it did not (see Chapter 7). Overall, these inconsistencies show that, on a general level, too little is known of the relationship between music in schools and student psychosocial wellbeing to make any firm conclusions about a connection between them. 280

Challenge to recommendations for music education. Claims within policy literature for the type of musical participation that will increase students’ psychosocial wellbeing are equally unfounded. Such documents recommend that legislators, funders, and schools seeking such benefits should engage students in music education or training programs (Australian Government, 2005; Parliament of Victoria, 2013) which are delivered in classrooms as curriculum activities (ACARA, 2011). There is evidence to support related claims that such programs will promote the intrinsic benefits of musical participation – including musical literacy (Lowe & Belcher, 2012) and creativity (Ewing, 2010) – and extrinsic benefits related to cognitive (Črnčec et al., 2006) and academic development (Biasutti & Concina, 2013). However, a review of existing research literature suggests that the claim this type of musical participation will also promote student’s subjective psychosocial wellbeing is dubious (see Chapter 3). The findings of the current project not only support the dubious nature of such recommendations, but go further to reject their credibility outright. As explained in Chapters 10 and 11, music programs must be delivered in a specific way to achieve psychosocial wellbeing outcomes for students. The classroom-based music education and curriculum programs recommended in policy literature not only lack the elements of musical participation necessary for promoting student psychosocial wellbeing (e.g. privacy, a focus on wellbeing, and small participant numbers), but they also involve elements that serve to prevent or impede such outcomes (e.g. lack of private setting, focus on musical mastery, education or curriculum content, large participant groups that lack a common need). Ongoing support for music in schools. The above challenges to policy documentation support fears from authors such as Stevens and Stefanakis that current policy-based advocacy for school-based music is based on “unsubstantiated assertion[s]” (2014, p. 12). This has potentially serious implications for the ongoing support of music in schools. This support, at a policy level, is said to be delicate (Gattenhof, 2009). Music is just one of many activities competing for a place in schools which already have limited resources (Stevens & Stefanakis, 2014; West, 2012), overcrowded curriculums (Lowe & Belcher, 2012), and are increasingly pushed to compete on standardised test scores (Polesel et al., 2012; West, 2012). Any activities taking up precious time and resources are therefore unlikely to

281

maintain support from policymakers, or schools, if they do not live up to their promised outcomes. As such, the fact that policy documents are making unsubstantiated claims about the psychosocial wellbeing benefits school-based music is of considerable concern. A primary source of concern is the ability for these claims to undermine the potential that presently exists in this area. That the level of support for school music articulated in policy documentation is unprecedented in Australian history (Ewing, 2010; O'Toole, 2010) shows great promise. Further, the other benefits they attribute to school-based musical participation, such as creativity, cultural awareness, as well as academic and cognitive development, (Australian Government, 2005; Parliament of Victoria, 2013) are indispensible to students of the 21st Century (Wyn, 2009a, 2009b). What’s more, there is considerable evidence to suggest the models of musical participation advocated in current policy are well suited to achieving these other benefits (Ewing, 2010). All of these points show that the overall direction of current policy in this area is propitious. Yet, despite its promise, policy support has not yet translated into legislation or the provision of the practical resources needed to implement music in schools (Crooke & McFerran, 2015; McFerran & Crooke, 2014). To capitalise on this current situation, and convert rhetoric into reality, it is critical to galvanise support by ensuring all policy claims are underpinned by a robust foundation of evidence (Ewing, 2010; Stevens & Stefanakis, 2014). At present, this foundation of evidence is incomplete. While there may be support for the intrinsic benefits, and extrinsic benefits related to cognitive and academic development, more work is needed to understand how and when benefits related to psychosocial wellbeing can be achieved. This project has shown the policy claim that “music education [emphasis added] uniquely contributes to the emotional [and] social […] growth of all [emphasis added] students” (Australian Government, 2005, p. v) is unrealistic. If government funding and resource support is provided for this end, it seems unlikely it will be sustainable if (or when) it becomes apparent that this claim does not reflect the reality of what is experienced in schools. Considering both the historical tendency for governments to retract policy support for music in schools (Ewing, 2010; Gattenhof, 2009), and the current financial climate (Kirby, 2014), such a realisation will surely endanger ongoing policy support. This not only threatens students’ access to the psychosocial wellbeing benefits of musical participation, but also 282

has the potential to undermine policy support for music in schools more generally. Conceivably, this would also jeopardise access to the other benefits of school-based musical participation (discussed above), and undo the work which has led to the historically significant level of policy backing we are now experiencing. Even without legislation, policy recommendations in this area have the potential to negatively impact school communities. If school leaders act independently on these recommendations, and find room in their own budgets and timetables to integrate music into their curriculum, it is unlikely they will observe all the benefits promised. This would likely affect their willingness to support music in their schools. Recent research has shown, in absence of practical government backing, support from school leaders is one of the most significant enablers of school music programs (Crooke & McFerran, 2015; McFerran & Crooke, 2014). Compromising this support will severely affect access to music in Australian schools, regardless of whether the practical implementation of political support occurs. Recommendations for policy. This PhD has argued that while policy recommendations regarding the provision of music in schools are promising at a general level, more work is needed to ensure that all claims within these recommendations are grounded on solid foundations. Specifically, it appears recommendations that class-based music instruction or education will achieve psychosocial wellbeing are not credible. Such claims seem to be based more on “subjective opinions” (Stevens & Stefanakis, 2014, p. 12), or presuppositions, than rigorous results and theory. Further, this PhD has argued that these claims are counterproductive for the ongoing support of music in schools. Based on these findings, two main recommendations are made here to address the potentially detrimental effect of maintaining these claims in forthcoming policy literature. Limit claims to musical benefits that are supported by evidence. The first recommendation is that ensuing policy (or any related advocacy literature) should cease to make unsubstantiated claims regarding the connection between school-based musical participation and psychosocial wellbeing. Until more is known about this connection, policies should limit justifications for increased resourcing for school music to claims that it will achieve intrinsic benefits, and academic and cognitive development. These benefits are important, and provide valid

283

arguments for supporting more music in schools. Attempting to claim psychosocial wellbeing benefits as an “added extra” without sufficiently considering how and when they will occur potentially endangers the current opportunity to provide students access to these intrinsic, academic and cognitive benefits – which in relative terms, have been demonstrated, are understood, and are suited to the forms of musical participation being promoted (Gill & Rickard, 2012; Lowe & Belcher, 2012). Critical engagement. The second option for addressing the shortcomings of current policy is for policymakers to engage with this area, and acknowledge that psychosocial wellbeing benefits cannot be considered as an expected side effect of all school-based music participation. This not only requires the recognition that this relationship is not yet understood, but also the recognition that for an understanding to be reached, a different approach is required; one that does not fit neatly into existing research models used in educational contexts. This is not to suggest policymakers themselves engage in investigation. It is job of researchers to provide evidence that is based on best-practice research models. However, it seems necessary that any policy is based on some kind of reliable evidence, and in this case, even a quick review of available research literature shows that there is little to no consensus in this area. Further, in light of this inconclusive knowledge base, and their growing commitment to supporting music in schools, it seems prudent policymakers commission a thorough evidence-based investigation of all proposed benefits related to student musical participation. This would contrast with more recent attempts, such as the Victorian Government’s Parliamentary Inquiry (Parliament of Victoria, 2013), which focuses instead on the type of public opinion and anecdotal reports which critics such as Stevens and Stefanakis (2014) warn against. A more rigorous inquiry would demonstrate perspicuity on the part of policymakers regarding the nature of existing evidence, while also showing a commitment to promoting school initiatives that are based on best-practice knowledge generation. Addressing policy goals. The second recommendation for policy is one that is strongly endorsed here. Literature from many fields show subjective wellbeing is increasingly recognised as critical for young people’s quality of life (Bradshaw, Martorano, Natali, & de Neubourg,

284

2013). This is particularly pertinent in the field of education. Both academics (Allen & Bowles, 2012), and policymakers (MCEETYA, 2008) argue that schools must place significant emphasis on supporting the subjective and social wellbeing of their students. It seems illogical, then, that we omit the psychosocial wellbeing benefits from any policy regarding music in schools. This would mean passing up a valuable opportunity to enhance the quality of life of our younger generations. From a policy perspective, it also means passing up an important opportunity to meet some of the primary goals set out in recent education policy literature (MCEETYA, 2008). Conclusion To really grasp whether or not there is truth in the claim that music in schools promotes psychosocial wellbeing, we must first acknowledge the shortcomings of existing approaches, and carefully adjust our investigations appropriately. It is argued that the challenges and recommendations identified in this dissertation will provide researchers with the information needed to make these adjustments. Only when recommendations such as these are put in place can we claim with conviction that our findings are plausible, and that either we, or others, can make decisions on the knowledge they generate. However, while the onus is on researchers to produce better evidence in this area, policymakers also have a responsibility to ensure policies are based on solid evidence. If policymakers make more informed claims regarding the potential benefits of music in schools, then any legislation or policy implementation in this area is more likely to achieve its intended outcomes, and be sustainable. So too will schools be better informed, and have more realistic expectations about music provision. This, in turn, has important implications for equal access to quality school music programs, and their benefits, for all of our students. The findings of this project also have important implications for other areas of research and theory. It invites researchers and theorists alike to reconsider the assumptions that underlie both the generation of knowledge and the theories of practice in any area of scientific investigation. For any field to really have the conviction that they are producing meaningful or legitimate knowledge, there needs to be ongoing critical engagement with how this knowledge is generated.

285

Chapter 13: Reflections In the introduction to this dissertation, I stated a number of factors that influenced my decision to undertake a PhD. This included personal beliefs, aspirations, and past experiences with music and research. In concluding this thesis I wish to take the opportunity to reflect on how the project has met these aspirations, how the knowledge it has generated sits with my prior experiences of research and music, and how the candidature as a whole has influenced my beliefs. Aspirations The aspiration of combining my musical and professional selves through undertaking this PhD has been successful, but in ways that I did not anticipate. While it did not mean I was able to play or create music, it did enable me to connect with it in a practical way through the process of research. Perhaps more importantly, however, I was able to connect with it theoretically, and sometimes philosophically. I was afforded the time to contemplate what it is about musical participation that makes it important as a social act, and as something that can support our wellbeing in institutional settings, such as schools. This in turn influenced, and sometimes challenged, my own assumptions about how and when music can be helpful. Namely, it made me reconsider my own presupposition that music, in and of itself, will always positively enhance the wellbeing of an individual or a group. I came to understand that, while music as an independent entity is both powerful and sometimes mysterious, like many things, it is by appropriating its processes and qualities in specific ways that music becomes an agent of social and psychological change. In this way, the bridging of my personal (musical) and professional (research) selves occurred at a deeper and more profound level than the practical one I had anticipated. Upon reflection, this was an unexpected, yet welcomed development. The other aspiration, or driving motivation, for undertaking this project was not realised. I did not achieve my goal of communicating the undeniable power of music to the wider fields of the social sciences, nor did I collect the evidence needed to convince policymakers that investment in more music will solve the great social dilemmas of the contemporary human experience. Yet, I believe what I did achieve was just as powerful, and when considering the larger picture, just as critical for achieving this longer-term goal. I discovered that, while music does have great potential in addressing many 286

aspects of wellbeing (especially in mainstream educational contexts), to assume that this potential is automatic is to risk its realisation. I have learnt that the relationship between music and wellbeing is somewhat reciprocal; that this potential has to be nurtured; and that it is just as important to recognise the limitations of musical participation than it is to be aware of the possibilities it affords. I believe that by becoming aware of these limitations, both in terms of what can be expected of musical participation, and how we can expect to demonstrate or capture benefits when they do occur, has led to a more informed, less naïve, and ultimately more powerful understanding. Further, I believe this more informed understanding has placed me in a firmer theoretical position to communicate, and make a more grounded argument for, how and when music can be appropriated to address issues in contemporary society. Prior Experiences My decision to undertake this dissertation was also influenced by my prior experiences in two main areas: my own musical background, and my involvement with research. My involvement with research included work as a professional researcher, and as a participant completing surveys in school. Through both of these experiences I had come to be critical of using surveys as a data collection method in schools. This criticism in part influenced the desire to undertake a PhD. I wanted the chance to reconnect with the qualitative methods of enquiry which were largely undervalued in my professional roles, but which I feel are often more appropriate in the social sciences. As described in the introduction chapter, this desire to pursue qualitative enquiry was sidelined in order to pursue my initial aspirations of collecting data appropriate for influencing policy. Yet, as I pursued these aspirations, I found my inclinations that the use of quantitative approaches in school settings can be fraught with limitations and challenges held true. In this way, I feel my prior experiences in this area had primed me to be open to the possibility that many limitations of research in this area could be related to the challenges of using quantitative approaches (particularly self-report surveys). Yet, while this project confirmed my scepticism of using surveys, it also enabled me to gain insight into how challenges related to their use in schools can be addressed. Therefore, I do not believe my prior experiences with surveys in schools settings caused me to dismiss them based on prior prejudice, but rather gave me the critical edge to interrogate their usefulness and reach a position from which to make informed recommendations for their future use. In this process, I also became aware that 287

it is important to both acknowledge and value the influence of prior experiences. Instead of constraining what can be pursued or achieved in a project, they can provide a source of embodied knowing, or an informed position from which to make interpretations of the research process and the subsequent knowledge it generates. My prior experiences with music also informed my decision to undertake a PhD. Initially this included my involvement with music outside of institutional contexts, either playing in bands, as a DJ, or putting on events in nightclubs. Through each of these activities I experienced music as catalyst for social connection, and something that gave me a general sense of wellbeing that was incredibly important for my day-to-day life. The decision to investigate this sense of connection and wellbeing in a school context was not the original intention, but as I progressed it became clear how important music was for me during my time as a student. Reflecting now on how the knowledge generated from this project sits with my prior experiences of musical participation in schools, the experiences of participation described and observed in this project not only mirror my own, but have also helped me to understand them at a deeper level. I had not fully contemplated the difference between my regular music classes and other music programs, which I now see were more akin to the type of musical participation recommended in this thesis. Like the students who participated in the class-based Keys to Success music program, I do not remember engaging in mainstream music classes differently from other curriculum subjects. Sometimes they were more fun than nonmusic classes, but I don't remember them ever having a profound impact on my psychosocial wellbeing. However, programs that were outside traditional classroom settings, had less members, and were delivered democratically by facilitators that took an interest in our individual and group wellbeing, had a remarkable impact on my psychosocial wellbeing that stays with me to this day. Further, to think about the musical experiences that have supported my wellbeing throughout the rest of my life, these too had many of the factors recommended in this thesis as necessary for supporting psychosocial wellbeing. This project and the knowledge it has generated has therefore both reinforced and deepened my insights into how music affected me during school, and how it continues to do so in life more generally. Beliefs I also started this PhD journey with a number of important beliefs that underpinned its initial focus. While these have been partly discussed in the above 288

paragraphs, I feel it is important to acknowledge that my understanding of how our beliefs and convictions influence what we think, do, and particularly how we undertake research, has evolved. I came to recognise research is always underpinned by beliefs, and that these often manifest as assumptions. Further, these assumptions can be counterproductive and sometimes detrimental to an area of investigation if ignored, not acknowledged, and not engaged with in a critical way. I came to recognise that there were several key assumptions underpinning much of the policy and research regarding the link between musical participation in schools and psychosocial wellbeing: that generic or education-focused forms of musical participation can promote wellbeing; and that existing methods of research are appropriate for capturing this interaction. Further, I discovered that these assumptions have largely contributed to a lack of understanding regarding this relationship, and have led to misguided recommendations for facilitating it. Through first recognising, and then critically engaging and testing these assumptions, I was able to demonstrate their limitations and offer ways in which they could be overcome, thus providing guidance for ways to move forward in this area. It is important to acknowledge, however, my own conviction that musical participation can promote wellbeing has underpinned this whole process. This is a belief that I have held for many years, and was instrumental in motivating me to start this project. It has also informed the decisions made throughout my candidature. It was this belief that pushed me to seek alternate explanations for the lack of benefits reported by participants through surveys, and in some cases, through interviews. My confidence in the inherent potential for musical participation to enrich our social and personal lives led me to interrogate our methods of enquiry rather than simply accepting and reporting inconclusive results. This has guided me to the conclusion that, while our beliefs or assumptions can sometimes hold us back – as I argue it has done for prior research in this field – they can also serve to guide us; to push us towards a commitment for undertaking research that looks at a given phenomenon in a more considered way. The key to wether our beliefs can be a help or a hindrance in our goals of enquiry is whether or not we are conscious of them, and their potential impact. I acknowledge that my belief about the potential of music was what drove me to investigate the benefits of a program (Keys to Success) that was not suited to wellbeing purposes. I was sure that music would achieve outcomes regardless of the form of participation. However, it was only when I experienced the impact of this belief that I

289

was able to recognise it, revise it, and thus become more pragmatic about when and how music can promote wellbeing, and how this relationship can best be investigated. Conclusion To summarise, my belief regarding the potential of musical participation in promoting psychosocial wellbeing is as strong as it ever has been. Yet, through this project I have come to be more realistic about this belief, and suggest that both researchers and policymakers must also take a more considered approach in this area, if this potential is to be realised. This includes both the need to consider the context in which musical participation occurs, and the methods that we use to explore and demonstrate its link to wellbeing. I am passionate in my hope that the recommendations I have made will guide researchers (and ultimately policymakers) to make more informed decision in this area. I know that these recommendations will have a lasting effect on how I undertake research, regardless of the field in which it is located. Lastly, I hope the outcomes of this project lead to research and policies that serve to enrich the lives of all students, and in the long term, society more generally.

290

References ABS. (2013). Children at School with Dissability. Profiles of Disability. Retrieved 4 June, 2014, from http://www.abs.gov.au/ausstats/[email protected]/Lookup/4429.0main+features1003020 09 ACARA. (2011). Shape of the Australian Curriculum: The Arts. Sydney, NSW: ACARA. Adamek, M., & Darrow, A. A. (2010). Music in Special Education (2nd ed.). USA: American Music Therapy Association. AIHW. (2009). A Picture of Australia’s Children 2009. (Cat. no. PHE 112). Canberra: AIHW. AIHW. (2010). Health and wellbeing of young Australians: technical paper on operational definitions and data issues for key national indicators. (Cat. no. WP 63). Canberra: AIHW. AIHW. (2011). Young Australians: their health and wellbeing 2011. (Cat. no. PHE 140). Canberra: AIHW. AIHW. (2012). Social and emotional wellbeing: development of a children’s headline indicator. (Cat. no. PHE 158). Canberra Allen, K. A., & Bowles, T. (2012). Belonging as a Guiding Principle in the Education of Adolescents. Australian Journal of Educational & Developmental Psychology, 12, 108-119. Allensworth, D., Lawson, E., Nicholson, L., & Wyche, J. (Eds.). (1997). Schools and Health: Our Nation's Investment. Washington, DC: National Academies Press. Andrews, G., & Slade, T. (2001). Interpreting scores on the Kessler Psychological Distress Scale (K10). Australian and New Zealand Journal of Public Health, 25(6), 494-497. Andrews, T. (2012). What is social constructionism. The Grounded Theory Review, 11(1), 39-46. Ansdell, G. (2002). Community Music Therapy & The Winds of Change. Voices: A World Forum for Music Therapy, 2(2). doi: 10.15845/voices.v2i2.83 Ansdell, G. (2004). Rethinking Music and Community: Theoretical Perspectives in Support of Community Music Therapy. In M. Pavlicevic & G. Ansdell (Eds.), Community Music Therapy (pp. 65-90). London, UK: Jessica Kingsley Publishing. 291

ARACY, & AIHW. (2012). Social and emotional wellbeing: development of a Children’s Headline Indicator. Canberra Arthur, M. W., Hawkins, J. D., Pollard, J. A., Catalano, R. F., & Baglioni, A. J. (2002). Measuring risk and protective factors for use, delinquency, and other adolescent problem behaviors: The Communities That Care Youth Survey. Evaluation Review, 26(6), 575-601. ASIB. (2009). A compendium of social inclusion indicators: How's Australia faring? Canberra: Commonwealth of Australia. Retrieved from http://www.socialinclusion.gov.au/AusGov/Board/Documents/Compendium.pdf ASIB. (2010). Social Inclusion in Australia: How Australia is Faring. Canberra: Commonwealth of Australia. ASIB. (2012). Social Inclusion in Australia: How Australia is Faring. Canberra: Commonwealth of Australia. Australia Office for the Arts. (2011). National Cultural Policy Discussion Paper. Canberra: Commonwealth of Australia. Australian Government, DEST. (2005). National Review of School Music Education: Augmenting the diminished. Canberra: Department of Education, Science and Training, Australian Government. Babo, G. D. (2004). The relationship between instrumental music participation and standardized assessment achievement of middle school students. Research Studies in Music Education, 22(1), 14-27. doi: 10.1177/1321103x040220010301 Bahri, S. (2006). Educating Through Art in Secondary Education. Paper presented at the UNESCO Expert Panel Meeting, Education Through Art Proceedings, France. Baker, F., & Jones, C. (2005). Holding a steady beat: The effects of a music therapy program on stabilising behaviours of newly arrived refugee students. British Journal of Music Therapy, 19(2), 67-74. Baker, F., & Jones, C. (2006). The effect of music therapy services on classroom behaviours of newly arrived refugee students in Australia—a pilot study. Emotional and Behavioural Difficulties, 11(4), 249-260. Baker, J. A. (1999). Teacher-student interaction in urban at-risk classrooms: Differential behavior, relationship quality, and student satisfaction with school. The Elementary School Journal, 100(1), 57-70.

292

Ball, S. J. (2006). Education policy and social class; the selected works of Stephen J. Ball. Abingdon, UK: Routledge. Bamford, A. (2006). The wow factor: Global research compendium on the impact of the arts in education. Berlin: Waxmann Verlag. Bandura, A. (1997). Self-Efficacy: The Exercise of Control. New York: Freeman. Barclay, J. R., & Doll, B. (2001). Early prospective studies of the high school dropout. School Psychology Quarterly, 16(4), 357-357. Barraket, J. (2005). Putting people in the picture? The role of the arts in social inclusion. (Social Policy Working Paper No. 4). Melbourne: Brotherhood of St Laurence and Centre for Public Policy, University of Melbourne. Retrieved from http://www.bsl.org.au/pdfs/barraket_arts_social_inclusion_1.pdf Batten, M., & Russell, J. (1995). Students at risk: A review of Australian literature 1980-1994. Melbourne: Australian Council for Educational Research. Beck, J., & Young, M. F. D. (2005). The assault on the professions and the restructuring of academic and professional identities: a Bernsteinian analysis. British Journal of Sociology of Education, 26(2), 183-197. Beck, U. (2002). The cosmopolitan society and its enemies. Theory, Culture & Society, 19(1-2), 17-44. Beebe, T. J., Harrison, P. A., Mcrae, J. A., Anderson, R. E., & Fulkerson, J. A. (1998). An evaluation of computer-assisted self-interviews in a school setting. Public Opinion Quarterly, 62(4), 623-632. Bell, W. (1957). Anomie, social isolation, and the class structure. Sociometry, 20(2), 105-116. doi: 10.2307/2785636 Bernstein, R. J. (2011). Beyond objectivism and relativism: Science, hermeneutics, and praxis. Philadelphia, Pennsylvania: University of Pennsylvania Press. Bessant, B. (1984). Free, compulsory, and secular education: The 1872 education act, Victoria Australia. Paedagogica Historica, 24(1), 5-25. Biasutti, M., & Concina, E. (2013). Music education and transfer of learning. Journal of Communications Research, 5(3), 397-413. Biesta, G. (2010). Pragmatism and the Philosophical Foundations of Mixed Methods Research. In A. Tashakkori & C. Teddlie (Eds.), The SAGE Handbook of Mixed Methods in Social & Behavioral Research (2 ed., pp. 95-118). Thousand Oaks, CA: SAGE Publications Inc.

293

Blandford, S., & Duarte, S. (2004). Inclusion in the community: a study of community music centres in England and Portugal, focusing on the development of musical and social skills within each centre. Westminster Studies in Education, 27(1), 725. Bleazby, J. (2011). Overcoming Relativism and Absolutism: Dewey's ideals of truth and meaning in philosophy for children. Educational Philosophy and Theory, 43(5), 453-466. doi: 10.1111/j.1469-5812.2009.00567.x Blum, R., W , & Libbey, H., P. (2004). School connectedness—strengthening health and education outcomes for teenagers. Journal of School Health, 74(7), 231-232. Boer, D., & Abubakar, A. (2014). Music listening in families and peer groups: benefits for young people's social cohesion and emotional well-being across four cultures. Frontiers in Psychology, 5(392), 1-15. doi: 10.3389/fpsyg.2014.00392 Bolger, L. (2013). Understanding and articulating the processa and meaning of collaboration in participatory music projects with marginalised young people and their supporting communities. (Doctoral dissertation), University of Melbourne. Retrieved from http://hdl.handle.net/11343/39742 Bond, L., Butler, H., Thomas, L., Carlin, J., Glover, S., Bowes, G., & Patton, G. (2007). Social and school connectedness in early secondary school as predictors of late teenage substance use, mental health, and academic outcomes. Journal of Adolescent Health, 40(4), 357 e359-357 e318. Bond, L., Carlin, J., B, , Thomas, L., Rubin, K., & Patton, G. (2001). Does bullying cause emotional problems? A prospective study of young teenagers. British Medical Journal, 323, 480-484. Bond, L., Thomas, L., Coffey, C., Glover, S., Butler, H., Carlin, J. B., & Patton, G. (2004). Long-term impact of the Gatehouse Project on the incidence of cannabis use in 16 year olds: A school-based cluster randomised trial. Journal of School Health, 74(1), 23-29. Bond, L., Thomas, L., Toumbourou, J., Patton, G., & Catalano, R. (2000). Improving the Lives of Young Victorians in Our Community: a survey of risk and protective factors. Melbourne: Centre for Adolescent Health. Bourne, P. A. (2010). A Conceptual Framework of Wellbeing in Some Western Nations. Current Research Journal of Social Sciences, 2(1), 15-23. Boyatzis, R. E. (1998). Transforming qualitative information: Thematic analysis and code development. Thousand Oaks, CA: Sage. 294

Bradshaw, J., Martorano, B., Natali, L., & de Neubourg, C. (2013). Children’s subjective well-being in rich countries. Child Indicators Research, 6(4), 619-635. Bresler, L. (2005). Music and the Intellect: Perspectives, Interpretations, and Implications for Education. Phi Delta Kappan, 87(1), 24-31. doi: 10.1177/003172170508700107 Breslow, L. (1972). A quantitative approach to the World Health Organisation definition of health: physical, mental and social well-being. International Journal of Epidemiology, 1(4), 347-355. doi: 10.1093/ije/1.4.347 Bryce, J., Harvey-Beavis, A., Livermore, J., & O’Toole, J. (1996). The Mayer Key Competencies and arts education. Melbourne: National Affliation of Arts Educators and the Australian Council for Educational Research. Bryce, J., Mendelovits, J., Beavis, A., McQueen, J., & Adams, I. (2004). Evaluation of school-based arts education programmes in Australian schools. Melbourne: Australian Council for Educational Research. Bullard, E. (2011). Music therapy as an intervention for inpatient treatment of suicidal ideation. Qualitative Inquiries in Music Therapy, 6, 75-121. Bunt, L., & Stige, B. (2014). Music therapy: An art beyond words (2nd ed.). New York, NY: Routledge. Burchardt, T., Le Grand, J., & Piachaud, D. (1999). Social Exclusion in Britain 19911995. Social Policy & Administration, 33(3), 227-244. Burchell, B., & Marsh, C. (1992). The effect of questionnaire length on survey response. Quality and Quantity, 26(3), 233-244. Burton, J., Horowitz, R., & Abeles, H. (1999). Learning in and through the Arts: Curriculum implications. Washington DC: The Arts Education Partnership and the President’s Committee on the Arts and Humanities. Buxton, R. (1981). Policy Making in Music Education. Paper presented at the Growing with Music: Report of the Fouth National Conference, Parkville, Vic. Carr, C., & Wigram, T. (2009). Music therapy with children and adolescents in mainstream schools: A systematic review. British Journal of Music Therapy, 23(1), 3-18. Carter, E., & Oldfield, A. (2002). A Music Therapy Group to Assist Clinical Diagnoses in Child and Family Psychiatry. In A. Davies & E. Richards (Eds.), Music therapy and group work: sound company (pp. 149-163). London, England: Jessica Kingsley Publishers. 295

Catterall, J. (2009). Doing well and doing good by doing art: The long term effects of sustained involvement in the visual and performing arts during high school. Los Angeles, CA: Los Angeles Imagination Group. Catterall, J. S., Capleau, R., & Iwanga, J. (1999). Involvement in the arts and human development: General involvement and intensive involvement in music and theater arts. In E. Fiske (Ed.), Champions of change: The impact of the arts on learning (pp. 1-18). Washington DC: The Arts Partnership and the President's Committee on the Arts and Humanities. Chambers, N. (1999). Close encounters: the use of critical reflective analysis as an evaluation tool in teaching and learning. Journal of Advanced Nursing, 29(4), 950-957. doi: 10.1046/j.1365-2648.1999.00975.x Cheong-Clinch, C. (2009). Music for engaging young people in education. Youth Studies Australia, 28(2), 50-57. Chin, T., & Rickard, N. S. (2012). The Music USE (MUSE) Questionnaire: An instrument to measure engagement in music. Music Perception: An Interdisciplinary Journal, 29(4), 429-446. doi: 10.1525/mp.2012.29.4.429 Choi, C. M. H. (2010). A pilot analysis of the psychological themes found during the CARING at Columbia–music therapy program with refugee adolescents from North Korea. Journal of Music Therapy, 47(4), 380-407. Clendenon-Wallen, J. (1991). The use of music therapy to influence the self-confidence and self-esteem of adolescents who are sexually abused. Music Therapy Perspectives, 9(1), 73-81. CMC, & MCEETYA. (2007). National education and the arts statement. Retrieved from http://www.mceecdya.edu.au/verve/_resources/National_Education_Arts_State ment.pdf. Coplan, R. J., Arbeau, K. A., & Armer, M. (2008). Don’t fret, be supportive! Maternal characteristics linking child shyness to psychosocial and school adjustment in kindergarten. Journal of Abnormal Child Psychology, 36(3), 359-371. Corrigall, K. A., Schellenberg, E. G., & Misura, N. M. (2013). Music training, cognition, and personality. Frontiers in Psychology, 4(222), 1-10. Costa-Giomi, E. (2004). Effects of three years of piano instruction on children’s academic achievement, school performance and self-esteem. Psychology of Music, 32(2), 139-152. 296

Crabtree, B. F., & Miller, W. L. (1999). Using codes and code manuals: A template organizing style of interpretation. In B. F. Crabtree & W. L. Miller (Eds.), Doing qualitative research (Vol. 2, pp. 163-177). Thousand Oaks, CA: Sage Publications. Cranston, N., Kimber, M., Mulford, B., Reid, A., & Keating, J. (2010). Politics and school education in Australia: a case of shifting purposes. Journal of Educational Administration, 48(2), 182-195. Creswell, J., W, & Plano Clark, V., L. (2011). Designing and Conducting Mixed Methods Research (2nd ed.). Thousand Oaks: SAGE Publications, Inc. Crisp, B., R. (2010). Belonging, connectedness and social exclusion. Journal of Social Inclusion, 1(2), 123-132. Črnčec, R., Wilson, S. J., & Prior, M. (2006). The cognitive and academic benefits of music to children: Facts and fiction. Educational Psychology, 26(4), 579-594. Crockett, J., Taylor, S., Grabham, A., & Stanford, P. (2006). Patient outcomes following an intervention involving community pharmacists in the management of depression. Australian Journal of Rural Health, 14(6), 263-269. Crooke, A. H. D., & McFerran, K. S. (2015). Barriers and enablers for implementing music in Australian schools: The perspective of four principals. British Journal of Education, Society & Behavioural Science, 7(1), 25-41. doi: 10.9734/BJESBS/2015/16090 Cross, D., Epstein, M., Hearn, L., Slee, P., Shaw, T., & Monks, H. (2011). National Safe Schools Framework: Policy and practice to reduce bullying in Australian schools. International Journal of Behavioral Development, 35(5), 398-404. doi: 10.1177/0165025411407456 Currie, M. (2004). Doing anger differently: A group percussion therapy for angry adolescent boys. International Journal of Group Psychotherapy, 54(3), 275-294. Currie, M., & Startup, M. (2012). Doing Anger Differently: Two controlled trials of percussion group psychotherapy for adolescent reactive aggression. Journal of Adolescence, 35(4), 843-853. Darrow, A. A., Novak, J., Swedberg, O., Horton, M., & Rice, B. (2009). The effect of participation in a music mentorship program on the self esteem and attitudes of at-risk students. Australian Journal of Music Education(2), 5-16.

297

De Looij‐Jansen, V., Petra, M., & De Wilde, E. J. (2008). Comparison of web‐based versus paper‐and‐pencil self‐administered questionnaire: effects on health indicators in Dutch adolescents. Health Services Research, 43(5p1), 1708-1721. De Winter, J. C. F. (2013). Using the Student’s t-test with extremely small sample sizes. Practical Assessment, Research & Evaluation, 18(10), 1-12. Deasy, R. J. (Ed.). (2002). Critical links: Learning in the arts and student academic and social development. Washington DC: Arts Education Partnership. DEECD. (2009). Partnerships in schools and the professional arts sector. Melbourne, Australia: Education Policy and Research Division Office for Policy, Research and Innovation, Department of Education and Early Childhood Development. DEECD. (2010). The State of Victoria's Children 2009: Aboriginal children and young people in Victoria. Victoria, Australia: Department of Education and Early Childhood Development. DEECD. (2014, March 5). Multicultural Education Retrieved 5 June, 2014, from http://www.education.vic.gov.au/about/programs/cultures/pages/multiabout.aspx DEECD. (2015, January 05). Secondary school program Retrieved February 16th, 2015, from http://www.study.vic.gov.au/deecd/index.cfm?BCF14DEF-826C-457AA7F0-476F0993111F Demo, D. H., & Acock, A. C. (1996). Family structure, family process, and adolescent well-being. Journal of Research on Adolescence, 6(4), 457-488. DeNora, T. (2000). Music in everyday life. Cambridge, UK: Cambridge University Press. Department of Education & Children’s Services. (2002). ARTSsmart. A strategy for Arts Education in South Australian Schools, 2003–06. Adelaide: South Australian Government. Department of Health. (2014, May 12). Diversity in Health: Cultural and linguistuc diversity Retrieved 19 June, 2014, from http://www.health.vic.gov.au/diversity/cald.htm Devroop, K. (2012). The social-emotional impact of instrumental music performance on economically disadvantaged South African students. Music Education Research, 14(4), 407-416. doi: 10.1080/14613808.2012.685456 Dewey, J. (1933). How We Think. Lexington, MA: D.C. Heath & Co.

298

Dobbs, J., Doctoroff, G. L., Fisher, P. H., & Arnold, D. H. (2006). The association between preschool children's socio-emotional functioning and their mathematical skills. Journal of Applied Developmental Psychology, 27(2), 97108. Doll, B., & Hess, R. S. (2001). Through a new lens: Contemporary psychological perspectives on school completion and dropping out of high school. School Psychology Quarterly, 16(4), 351-351. Doyle, J. L. (2014). Cultural relevance in urban music education: A synthesis of the literature. Update: Applications of Research in Music Education, 32(2), 44-51. doi: 10.1177/8755123314521037 Eisner, E. (1999). Does experience in the arts boost academic achievement? Arts Education Policy Review, 100(1), 32-38. Ellickson, P. L., Bianca, D., & Schoeff, D. C. (1988). Containing attrition in schoolbased research: An innovative approach. Evaluation Review, 12(4), 331-351. Ewing, R. (2010). The Arts and Australian Education: Realising potential. Camberwell, Victoria: ACER Press. Fereday, J., & Muir-Cochrane, E. (2008). Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. International journal of qualitative methods, 5(1), 80-92. Finlay, L. (2002a). Negotiating the swamp: the opportunity and challenge of reflexivity in research practice. Qualitative Research, 2(2), 209-230. Finlay, L. (2002b). ''Outing'' the researcher: The provenance, process, and practice of reflexivity. Qualitative Health Research, 12(4), 531-545. Finlay, L. (2012). Five lenses for the reflexive interviewer. In J. F. Gubrium, J. A. Holstein, A. Marvasti & K. McKinney (Eds.), The SAGE Handbook of interview research: The complexity of the craft (2nd ed., pp. 317-331). Los Angeles: SAGE. Fiske, E. (Ed.). (1999). Champions of change: The impact of the arts on learning. Washington, DC.: The Arts Education Partnership and the President’s Committee on the Arts and Humanities. Fowler, C. (2002). Finding the way to be basic: Music education in the 1990s and beyond. In M. L. Mark (Ed.), Music education: Source readings from ancient Greece to today (pp. 279-285). New York: Routledge.

299

Fox Piven, F. (2004). Neoliberal social policy and labor market discipline. In M. Zweig (Ed.), What's class got to do with it? American society in the twenty-first century (pp. 113-124). Ithaca, NY: Cornell University Press. Galesic, M., & Bosnjak, M. (2009). Effects of questionnaire length on participation and indicators of response quality in a web survey. Public Opinion Quarterly, 73(2), 349-360. Garrett, P. (2009). Arts in Australia’s national curriculum (Media Release). Retrieved 23/01/2013 http://www.environment.gov.au/minister/archive/env/2009/pubs/mr20090417b.p df Garrett, S. (2010). The role of community music in helping disadvantaged young people in South Wales to confront social exclusion. International Journal of Community Music, 3(3), 371-377. doi: 10.1386/ijcm.3.3.371_1 Gaspart, F. (1998). Objective measures of well-being and the cooperative production problem. Social Choice and Welfare, 15(1), 95-112. Gattenhof, S. J. (2009). The Arts and the National Curriculum for Australian Schools. Paper presented at the Queensland Independent Schools Curriculum Forum, Brisbane. http://eprints.qut.edu.au/29360/1/c29360.pdf General Assembly of the United Nations. (1948). Universal Declaration of Human Rights. Retrieved from http://www.un.org/en/documents/udhr/ Geretsegger, M., Holck, U., & Gold, C. (2012). Randomised controlled trial of improvisational music therapy's effectiveness for children with autism spectrum disorders (TIME-A): study protocol. BMC Pediatrics, 12(2). Retrieved from http://www.biomedcentral.com/1471-2431/12/2/ doi:10.1186/1471-2431-12-2 Gergen, K. J., & Back, K. W. (1966). Communication in the interview and the disengaged respondent. Public Opinion Quarterly, 30(3), 385-398. doi: 10.1086/267431 Gfeller, K., E. (2002). Music as Communication. In R. F. Unkefer (Ed.), Music therapy in the treatment of adults with mental disorders (pp. 50-62). New York: Schirmer Books. Gill, A., & Rickard, N. (2011). Non-musical Benefits of School Music Instruction. In S. N, Rickard & M. K (Eds.), Lifelong Engagement with Music (pp. 59-74): Nova Science Publishers, Inc.

300

Gill, A., & Rickard, N. (2012). Non-musical benefits of school-based music education and training. In S. N, Rickard & M. K (Eds.), Lifelong engagement with music: Benefits for mental health and well-being (pp. 57-72). Hauppauge NY, United States: Nova Science Publishers, Inc. Glover, S., Burns, J., Butler, H., & Patton, G. (1998). Social environments and the emotional wellbeing of young people. Family Matters, 49, 11-16. Gold, C., Solli, H. P., Krüger, V., & Lie, S. A. (2009). Dose–response relationship in music therapy for people with serious mental disorders: Systematic review and meta-analysis. Clinical Psychology Review, 29(3), 193-207. Gonski, D. (2011). Review of Funding for Schooling—Final Report. Canberra: Department of Education, Employment and Workplace Relations. Gooding, L. F. (2011). The effect of a music therapy social skills training program on improving social competence in children and adolescents with social skills deficits. Journal of Music Therapy, 48(4), 440-462. Gorard, S., Taylor, C., & Fitz, J. (2001). Social exclusion and public policy: The relationship between local school admission arrangements and segregation by poverty. The International Journal of Sociology and Social Policy, 21(4-6), 1036. Government of Victoria. (1872). The Victorian Education Act: Schedule 1. Melbourne: Government of Victoria. Retrieved from http://www.foundingdocs.gov.au/resources/transcripts/vic8_doc_1872.pdf Grierson, E. (2007). Difference: A critical investigation of the creative arts with attention to art as a site of knowledge. Educational Philosophy and Theory, 39(5), 531-142. Griffiths, K. M., & Christensen, H. (2006). Review of randomised controlled trials of Internet interventions for mental disorders and related conditions. Clinical Psychologist, 10(1), 16-29. Grimmett, H., Rickard, N. S., Gill, A., & Murphy, F. (2010). The perilous path from proposal to practice: A qualitative program evaluation of a regional music program. Australian Journal of Music Education(2), 52-65. Grocke, D., Bloch, S., & Castle, D. (2009). The effect of group music therapy on quality of life for participants living with a severe and enduring mental illness. Journal of Music Therapy, 46(2), 90-104.

301

Guest, G., Namey, E. E., & Mitchell, M. L. (2012). Collecting Qualitative Data: A Field Manual for Applied Research. Thousand Oaks, CA: SAGE Publications, Inc. Gutiérrez, J. P., & Torres-Pereda, P. (2009). Acceptability and reliability of an adolescent risk behavior questionnaire administered with audio and computer support. Revista Panamericana de Salud Pública, 25(5), 418-422. Hall, C., & Thomson, P. (2007). Creative partnerships? Cultural policy and inclusive arts practice in one primary school. British Educational Research Journal, 33(3), 315-329. doi: 10.1080/01411920701243586 Hallam, S. (2010). The power of music: Its impact on the intellectual, social and personal development of children and young people. International Journal of Music Education, 28(3), 269-289. doi: 10.1177/0255761410370658 Hamilton, M., & Redmond, G. (2010). Conceptualisation of social and emotional wellbeing for children and young people, and policy implications. Sydney: Social Policy Research Centre, University of New South Wales. Hampshire, K. R., & Matthijsse, M. (2010). Can arts projects improve young people’s wellbeing? A social capital approach. Social Science & Medicine, 71(4), 708716. doi: 10.1016/j.socscimed.2010.05.015 Hargreaves, D. J., & Aksentijevic, A. (2011). Music, IQ, and the executive function. British Journal of Psychology, 102(3), 306-308. Hartford, M. E. (1971). Groups in social work. New York: Columbia University Press. Haynes, F., & Chalk, B. (2004). The impact of arts education programmes on student motivation: Final report of a research project. University of Western Australia, West Australian Department of Education and Training and Western Australian Department for Culture and the Arts. Hemphill, S. A., Heerde, J. A., Herrenkohl, T. I., Patton, G. C., Toumbourou, J. W., & Catalano, R. F. (2011). Risk and protective factors for adolescent substance use in Washington State, the United States and Victoria, Australia: A longitudinal study. Journal of Adolescent Health, 49(3), 312-320. Henderson, D. (2005). What is education for? Situating history, cultural understandings and studies of society and environment against neo-conservative critiques of curriculum reform. Australian Journal of Education, 49(3), 306-319. Heron, J., & Reason, P. (1997). A participatory inquiry paradigm. Qualitative inquiry, 3(3), 274-294. 302

Hides, L., Carroll, S., Scott, R., Cotton, S., Baker, A., & Lubman, D. I. (2012). Quik fix: A randomized controlled trial of an enhanced brief motivational interviewing intervention for alcohol/cannabis and psychological distress in young people. Psychotherapy and Psychosomatics, 82(2), 122-124. Hides, L., Lubman, D. I., Elkins, K., Catania, L. S., & Rogers, N. (2007). Feasibility and acceptability of a mental health screening tool and training programme in the youth alcohol and other drug (AOD) sector. Drug and Alcohol Review, 26(5), 509-515. doi: 10.1080/09595230701499126 Hietolahti-Ansten, M., & Kalliopuska, M. (1990). Self-esteem and empathy among children actively involved in music. Perceptual and Motor Skills, 71(3f), 13641366. Ho, Y.-C., Cheung, M.-C., & Chan, A. S. (2003). Music training improves verbal but not visual memory: Cross-sectional and longitudinal explorations in children. Neuropsychology, 17(3), 439-450. Hoerger, M. (2010). Participant dropout as a function of survey length in Internetmediated university studies: Implications for study design and voluntary participation in psychological research. Cyberpsychology, Behavior, and Social Networking, 13(6), 697-700. Holdsworth, R., Cahill, H., & Smith, G. (2003). Student Action Teams 2001-2002: Phase 2: An Evaluation of Implementation and Impact. (Research Report 22). Melbourne: The Australian Youth Research Centre. Hooghe, M., & Vanhoutte, B. (2011). Subjective well-being and social capital in Belgian communities. The impact of community characteristics on subjective well-being indicators in Belgium. Social Indicators Research, 100(1), 17-36. doi: 10.1007/s11205-010-9600-0 Humpage, L. (2006). An ‘inclusive’ society: a ‘leap forward’ for Māori in New Zealand? Critical Social Policy, 26(1), 220-242. doi: 10.1177/0261018306059773 Hunter, M. (2005). Education and the arts research overview. Sydney, NSW: Australian Council for the Arts, Australian Government. Huppert, F. T. (2013). Flourishing across Europe: Application of a new conceptual framework for defining well-being. Social Indicators Research, 110(3), 837-861. doi: 10.1007/s11205-011-9966-7

303

Huxley, P., & Thornicroft, G. (2003). Social inclusion, social quality and mental illness. British Journal of Psychiatry, 182(4), 289-290. Hyde, M. (2010). Understanding Diversity and Inclusion. In M. Hyde, L. Carpenter & C. Conway (Eds.), Diverstiy and Inclusion in Australian Schools (pp. 3-13). South Melbourne, Victoria: Oxford University Press. Iadeluca, V., & Sangiorgio, A. (2008). Bambini al Centro: Music as a means to promote wellbeing. International Journal of Community Music, 1(3), 311-318. doi: 10.1386/ijcm.1.3.311/1 IBM Corporation. (2011). SPSS Statistics (Version 20.0.0). Chicago: SPSS Inc. Ivey, B. (2002). Remarks to the national music education summit, Washington, D.C, October 2, 1998. In M. L. Mark (Ed.), Music education: Source readings from ancient Greece to today (pp. 291-293). New York: Routledge. Jarman, J. (2001). Explaining social exclusion. The International Journal of Sociology and Social Policy, 21(4-6), 3-3. Jones, C., Baker, F., & Day, T. (2004). From healing rituals to music therapy: bridging the cultural divide between therapist and young Sudanese refugees. The Arts in Psychotherapy, 31(2), 89-100. Jones, P. M. (2005). Music education and the knowledge economy: Developing creativity, strengthening communities. Arts Education Policy Review, 106(4), 512. doi: 10.3200/AEPR.106.4.5-12 Jones, P. M. (2010). Developing social capital: A role for music education and community music in fostering civic engagement and intercultural understanding. International Journal of Community Music, 3(2), 291-302. doi: 10.1386/ijcm.3.2.291_1 Jonkman, H., Steketee, M., Tombourou, J. W., Cini, K., & Williams, J. (2014). Community variation in adolescent alcohol use in Australia and the Netherlands. Health Promotion International, 29(1), 109-117. Jorgensen, E. R. (1995). Justifying music instruction in American public schools: A historical perspective. Arts Education Policy Review, 96(6), 31-38. Joseph, D. (2011). Cultural diversity in Australia: promoting the teaching and learning of South African music. Australian Journal of Music Education(1), 42-56. Juslin, P. N., & Laukka, P. (2003). Communication of emotions in vocal expression and music performance: Different channels, same code? Psychological Bulletin, 129(5), 770. 304

Kang, H. J., & Williamson, V. J. (2014). Background music can aid second language learning. Psychology of Music, 42(5), 728-747. Kano, M., Franke, T., Afifi, A. A., & Bourque, L. B. (2008). Adequacy of reporting results of school surveys and nonresponse effects: A review of the literature and a case study. Educational Researcher, 37(8), 480-490. Kaplan, R. M., & Anderson, J. P. (1988). A general health policy model: update and applications. Health Services Research, 23(2), 203-235. Kaplow, J. B., Saunders, J., Angold, A., & Costello, E. J. (2010). Psychiatric symptoms in bereaved versus nonbereaved youth and young adults: a longitudinal epidemiological study. Journal of the American Academy of Child & Adolescent Psychiatry, 49(11), 1145-1154. Karkou, V., & Glasman, J. (2004). Arts, education and society: The role of the arts in promoting the emotional wellbeing and social inclusion of young people. Support for Learning, 19(2), 57-65. Kelstrom, J., M. (1998). The untapped power of music: Its role in the curriculum and its effect on academic achievement. NASSP Bulletin, 82(597), 34-43. Kennedy, R., & Scott, A. (2005). A pilot study: The effects of music therapy interventions on middle school students' ESL skills. Journal of Music Therapy, 12(4), 244-261. Keyes, C. L. M. (1998). Social Well-Being. Social Psychology Quarterly, 61(2), 121140. Keyes, C. L. M. (2002). The mental health continuum: from languishing to flourishing in life. Journal Of Health And Social Behavior, 43(2), 207-222. Keyes, C. L. M. (2005). The subjective well-being of America’s youth: Toward a comprehensive assessment. Adolescent and Family Health, 4(1), 3-11. Keyes, C. L. M. (2006). Subjective well-being in mental health and human development research worldwide: An introduction. Social Indicators Research, 77(1), 1-10. doi: 10.1007/s11205-005-5550-3 Keyes, C. L. M. (2007). Promoting and protecting mental health as flourishing: A complementary strategy for improving national mental health. American Psychologist, 62(2), 95-108. doi: 10.1037/0003-066x.50.12.965. Kilgore, D. W. (2001). Critical and postmodern perspectives on adult learning. New directions for adult and continuing education, 2001(89), 53-62.

305

Kim, J. (2015). Music therapy with children who have been exposed to ongoing child abuse and poverty: A pilot study. Nordic Journal of Music Therapy, 24(1), 2743. Kim, S., Kverno, K., Lee, E. M., Park, J. H., Lee, H. H., & Kim, H. L. (2006). Development of a music group psychotherapy intervention for the primary prevention of adjustment difficulties in Korean adolescent girls. Journal of child and adolescent psychiatric nursing, 19(3), 103-111. Kinder, K., & Harland, J. (2004). The arts and social inclusion: What's the evidence? Support for Learning, 19(2), 52-56. doi: 10.1111/j.0268-2141.2004.00320.x Kirby, T. (2014). Health and science suffer major cuts in Australia's budget. The Lancet, 383(9932), 1874-1876. Kirschner, S., & Tomasello, M. (2010). Joint music making promotes prosocial behavior in 4-year-old children. Evolution and Human Behavior, 31(5), 354-364. doi: 10.1016/j.evolhumbehav.2010.04.004 Knox-Anderson, H., & Rickard, N. S. (2007). A preliminary examination of short-term effects of an active class-based music program on young adolescents' selfesteem and anger expression. Australian Journal of Music Education, 5(1), 4-16. Kostenko, W., Scutella, R., & Wilkins, R. (2010). Estimates of Poverty and Social Exclusion in Australia: A Multidimensional Approach. Melbourne: Melbourne Institute of Applied Economic and Social Research. Laberee, D. F. (1997). Public goods, private goods: the American struggle over educational goals. American Educational Research Journal, 34(1), 39-81. Labonte, R. (2004). Social Inclusion/exclusion: Dancing the dialectic. Health Promotion International, 19(1), 115-121. Lareau, A., & Horvat, E. M. (1999). Moments of social inclusion and exclusion race, class, and cultural capital in family-school relationships. Sociology of Education, 72(1), 37-53. Larson, J. S. (1993). The measurement of social well-being. Social Indicators Research, 28(3), 285-296. doi: 10.1007/bf01079022 Larson, J. S. (1996). The World Health Organization's definition of health: Social versus spiritual health. Social Indicators Research, 38(2), 181-192. doi: 10.1007/bf00300458 Lebler, D. (2007). Student-as-master? Reflections on a learning innovation in popular music pedagogy. International Journal of Music Education, 25(3), 205-221. 306

Leigh, B. C., Gillmore, M. R., & Morrison, D. M. (1998). Comparison of diary and retrospective measures for recording alcohol consumption and sexual activity. Journal of Clinical Epidemiology, 51(2), 119-127. Levitas, R. (1996). The concept of social exclusion and the new Durkheimian hegemony. Critical Social Policy, 16(46), 5-20. Levitas, R., Pantazis, C., Fahmy, E., Gordon, D., Lloyd, E., & Patsios, D. (2007). The multi-dimensional analysis of social exclusion. Bristol: Department of Sociology and School for Social Policy, University of Bristol. Libbey, H. P. (2004). Measuring student relationships to school: Attachment, bonding, connectedness, and engagement. Journal of School Health, 74(7), 274-283. Lincoln, Y. S., Lynham, S. A., & Guba, E. G. (2011). Paradigmatic controversies, contradictions, and emerging confluences, revisited. In N. K. Denzin & Y. S. Lincoln (Eds.), The SAGE handbook of qualitative research (4th ed., pp. 97128). Thousand Oaks, CA: SAGE Publications Inc. Lowe, G., & Belcher, S. (2012). Direct instruction and music literacy: one approach to augmenting the diminishing? Australian Journal of Music Education(1), 3-13. Lynch, H., & Allan, J. (2006). Social Inclusion and the Arts: Final Report to the Scottish Arts Council. Stirling: University of Stirling. MacIntyre, S. (2008). Participation in the classroom, productivity in the workforce — unfulfilled expectations. Paper presented at the ACER Research Conference: Touching the Future: Building Skills for Life and Work, Brisbane, QLD. Malekoff, A. (2014). Group work with adolescents: Principles and practice (3rd ed.). New York, NY: The Guilford Press. Marcus, R. F., & Sanders-Reio, J. (2001). The influence of attachment on school completion. School Psychology Quarterly, 16(4), 427-427. McCarthy, K., Ondaatje, E., Zakaras, L., & Brooks, A. (2004). Gifts of the muse: Reframing the debate about the benefits of the Arts. Santa Monica, CA: RAND. MCEETYA. (1999). Adelaide Declaration of National Goals for Schooling in the Twenty-first Century. Ministerial Council on Education, Employment, Training and Youth Affairs. MCEETYA. (2008). Melbourne Declaration on Educational Goals for Young Australians. Melbourne: Ministerial Council on Education, Employment, Training and Youth Affairs. Retrieved from

307

http://www.nap.edu.au/_Documents/PDF/National Declaration on the Educational Goals for .pdf MCEETYA, & CMC. (2007). National Education and the Arts Statement. Canberra: Department of Communication, Information Technology and the Arts. Retrieved from http://www.mceecdya.edu.au/verve/_resources/National_Education_Arts_State ment.pdf McFerran, K., & Grocke, D. (2007). Understanding music therapy experiences through interviewing: A phenomenological microanalysis. In T. Wosch & T. Wigram (Eds.), Microanalysis in Music Therapy (pp. 273-284). London: Jessica Kingsley Publishers. McFerran, K., & Teggelove, K. (2011). Music Therapy with Young People in Schools: After the Black Saturday Fires. Voices: A World Forum for Music Therapy, 11(1). doi: 10.15845/voices.v11i1.285 McFerran, K. S. (2000). From the mouths of babes: The response of six younger bereaved teenagers to the experience of psychodynamic group music therapy. Australian Journal of Music Therapy, 11(2000), 3-22. McFerran, K. S. (2010a). Adolescents, Music and Music Therapy. London. UK: Jessica Kingsley Publishing. McFerran, K. S. (2010b). Tipping the scales: A substantive theory on the value of group music therapy for supporting grieving teenagers. Qualitative Inquiries in Music Therapy (A Monograph Series), 5, 2-49. McFerran, K. S., & Crooke, A. H. D. (2014). Enabling Tailored Music Programs in Elementary Schools: An Australian Exemplar. Journal of Education and Training Studies, 2(4), 138-147. doi: 10.11114/jets.v2i4.536 McFerran, K. S., & Crooke, A. H. D. (In press). Music Therapy and Mourning. In J. Davidson & S. Garrido (Eds.), Music and Mourning: Interdisciplinary Investigations. Surrey, England: Ashgate Publishing Ltd. McFerran, K. S., & Hunt, M. (2008). Learning from experiences in action: Music in schools for coping with grief and loss. Educational Action Research, 16(1), 4353. McFerran, K. S., Roberts, M., & O'Grady, L. (2010). Music therapy with bereaved teenagers: A mixed methods perspective. Death Studies, 34(6), 541-565.

308

McFerran, K. S., & Stephenson, J. (2010). Music therapy and evidence based practice in special education. In V. Karkou (Ed.), Arts therapies in Special Education (pp. 259-270). London: Jessica Kingsley Publishers. McFerran-Skewes, K. (2005). Using songs with groups of teenagers: How does it work? Social Work with Groups, 27(2-3), 143-157. McGuigan, K. A., Ellickson, P. L., Hays, R. D., & Bell, R. M. (1997). Adjusting for Attrition in School-Based Samples Bias, Precision, and Cost Trade-Offs of Three Methods. Evaluation Review, 21(5), 554-567. McKegg, K., Wilson, D., Goodwin D., Black, X., Sauni, P., Toumu’a, R., . . . Hanna, K. (2012). Evaluation of Sistema Aotearoa. Auckland: Institute of Public Policy, AUT University. McLoyd, V. C., Jayaratne, T. E., Ceballo, R., & Borquez, J. (1994). Unemployment and work interruption among African American single mothers: Effects on parenting and adolescent socioemotional functioning. Child Development, 65(2), 562-589. Mehr, S. A., Schachner, A., Katz, R. C., & Spelke, E. S. (2013). Two randomized trials provide no consistent evidence for nonmusical cognitive benefits of brief preschool music enrichment. PloS one, 8(12), e82007. doi: 10.1371/journal.pone.0082007 Mezirow, J. (1990). How critical reflection triggers transformative learning. In J. Mezirow (Ed.), Fostering critical reflection in adulthood: A guide to transformative and emancipatory learning (pp. 1-20). San Francisco: JosseyBass. Mezirow, J. (1998). On Critical Reflection. Adult Education Quarterly, 48(3), 185-198. doi: 10.1177/074171369804800305 Michel, D. E., & Farrell, D. M. (1973). Music and self-esteem: Disadvantaged problem boys in an all-black elementary school. Journal of Research in Music Education, 21(1), 80-84. doi: 10.2307/3343983 Montello, L., & Coons, E., E. (1998). Effects of active versus passive group music therapy on preadolescents with emotional, learning, and behavioral disorders. Music Therapy, 35(1), 49-67. Moreno, S., Marques, C., Santos, A., Santos, M., Castro, S. L., & Besson, M. (2009). Musical training influences linguistic abilities in 8 yr old children: More evidence for brain plasticity. Cerebral Cortex, 19(3), 712-723.

309

Morrison, Z. (2010). On dignity: Social inclusion and the politics of recognition. (Social Policy Working Paper no. 12). Melbourne: Brotherhood of St Laurence & University of Melbourne Centre for Public Policy. Retrieved from http://www.bsl.org.au/pdfs/Morrison_On_dignity_2010.pdf Moses, M. S. (2004). Social welfare, the neo-conservative turn and educational opportunity. Journal of Philosophy of Education, 38(2), 275-286. Nicholls, T. (2002). 'Could I Play a Different Role?' Group Music Therapy With Severly Learning Disabled Adolescents. In A. Davies & E. Richards (Eds.), Music therapy and group work: sound company (pp. 231-246). London, England: Jessica Kingsley Publishers. Niemiec, C. P., Ryan, R. M., & Deci, E. L. (2009). The path taken: Consequences of attaining intrinsic and extrinsic aspirations in post-college life. Journal of Research in Personality, 43(3), 291-306. Nordoff, P., & Robbins, C. (1971). Music therapy in special education. New York, NY: John Day. Nussbaum, M. (2001). The fragility of goodness. Luck and ethics in Greek tragedy and philosophy. Cambridge: Cambridge University Press. Nutbeam, D., Smith, C., Moore, L., & Bauman, A. (1993). Warning! Schools can damage your health: Alienation from school and its impact on health behaviour. Journal of Paediatrics and Child Health, 29, S25-S30. doi: 10.1111/j.14401754.1993.tb02256.x O'Connor, P. (2008). Drama for inclusion: A pedagogy hope. In M. Anderson, J. Hughes & J. Manuel (Eds.), Drama and English teaching: Imagination, action and engagement. Melbourne: Oxford University Press. O'Toole, J. (2009). The Arts and creativity: A manifesto for schools. In C. Sinclair, N. Jeanneret & J. O’Toole (Eds.), Education in the Arts: Teaching and Learning in the Contemporary Curriculum. South Melbourne: Oxford University Press. O'Toole, J. (2010). Forward. In C. Glascodine (Ed.), The Arts and Australian Education: Realising Potential. Camberwell, Victoria: ACER Press. Odena, O. (2007). Music as a way to address Social Inclusion and Respect for Diversity in early childhood. Belfast, Ireland: NFER at Queen's. Retrieved from http://www.schoolsworkingtogether.co.uk/documents/music-addressing-socialinclusion.pdf

310

Odena, O. (2014). Musical creativity as a tool for inclusion. In E. Shiu (Ed.), Creativity Research: An Inter-Disciplinary and Multi-Disciplinary Research Handbook (pp. 247-270). New York, NY: Routledge. Oldfield, A. (2006). Interactive music therapy: A positive approach: music therapy at a child development centre. London, UK: Jessica Kingsley Publishers. Olteţeanu, I. (2010). Vocal expression, music performance, and communication of emotions. Linguistic and Philosophical Investigations(9), 311-316. Osmond, J., & Darlington, Y. (2005). Reflective analysis: Techniques for facilitating reflection. Australian Social Work, 58(1), 3-14. doi: 10.1111/j.0312407X.2005.00179.x Osterman, K. F. (2000). Students' need for belonging in the school community. Review of Educational Research, 70(3), 323-367. Parliament of Victoria. (2013). Inquiry into the extent, benefits and potential of music education in Victorian schools. (Parliamentary paper No. 277). Education and Training Committee: Victorian Government Printer. Retrieved from http://www.parliament.vic.gov.au/file_uploads/Music_Education_Final_041113 _FJWsJhBy.pdf Parsons, B., Schaffner, M., Little, G., & Felton, H. (1984). Drama, language and learning. NADIE Paper No 1. Hobart: National Association for Drama in Education. Pavlicevic, M. (2010). Let the Music Work: Optimal Moments of Collaborative Musicing. In B. Stige, G. Ansdell, C. Elefant & M. Pavlicevic (Eds.), Where Music Helps: Community Music Therapy in Action and Reflection (pp. 99-112). Burlington, USA: Ashgate Publishing Company. Pitts, P. S. (2005). Valuing Musical Participation. Aldershot, UK: Ashgate. Pitts, S., & Davidson, J. (2000). Supporting musical development in the primary school: An English perspective on band programmes in Sydney, NSW. Research Studies in Music Education, 14(1), 76–84. Plummer, K. (2011). Critical humanism and queer theory. In N. K. Denzin & Y. S. Lincoln (Eds.), The SAGE handbook of qualitative research (4th ed., pp. 195207). Thousand Oaks, CA: SAGE Publications Inc. Plummer, M. L., Ross, D. A., Wight, D., Changalucha, J., Mshana, G., Wamoyi, J., . . . Obasi, A. I. (2004). “A bit more truthful”: the validity of adolescent sexual

311

behaviour data collected in rural northern Tanzania using five methods. Sexually Transmitted Infections, 80(suppl 2), ii49-ii56. Poland, B., & Pederson, A. (1998). Reading between the lines: Interpreting silences in qualitative research. Qualitative Inquiry, 4(2), 293-312. doi: 10.1177/107780049800400209 Polesel, J., Dulfer, N., & Turnbull, M. (2012). The experience of education: The impacts of high stakes testing on school students and their families. Sydney: Whitlam Institute, University of Western Sydney. Retrieved from http://www.whitlam.org/__data/assets/pdf_file/0008/276191/High_Stakes_Testi ng_Literature_Review.pdf Prior, M. R., Sanson, A., Smart, D., & Oberklaid, F. (2000). Pathways from infancy to adolescence: Australian Temperament Project 1983-2000. (0642394784). Melbourne, Victoria: Australian Institute of Family Studies. Rapoport, J. L., & Ismond, D. R. (1996). DSM-IV training guide for diagnosis of childhood disorders. New York: Brunner Mazel. Rathod, S., & LaBruna, A. (2005). Questionnaire length and fatigue effects-does size really matter? Paper presented at the ESOMAR, Conference on Panel Research, Budapest. Rauscher, F. H., & Zupan, M. A. (2000). Classroom keyboard instruction improves kindergarten children’s spatial-temporal performance: A field experiment. Early Childhood Research Quarterly, 15(2), 215-228. Ray, M. A. (1990). Critical reflective analysis of Parse's and Newman's research methodologies. Nursing Science Quarterly, 3(1), 44-46. doi: 10.1177/089431849000300111 Resnick, M. D. (2000). Protective factors, resiliency, and healthy youth development. Adolescent Medicine, 11(1), 157–164. Resnick, M. D., Bearman, P. S., Blum, R. W., Bauman, K. E., Harris, K. M., Jones, J., . . . Udry, J. R. (1997). Protecting adolescents from harm: Findings from the National Longitudinal Study on Adolescent Health. JAMA, 278(10), 823-832. doi: 10.1001/jama.1997.03550100049038 Richmond, R. L., Butler, T., Belcher, J. M., Wodak, A., Wilhelm, K. A., & Baxter, E. (2006). Promoting smoking cessation among prisoners: feasibility of a multi‐

312

component intervention. Australian and New Zealand Journal of Public Health, 30(5), 474-478. Rickard, N. S., Appelman, P., James, R., Murphy, F., Gill, A., & Bambrick, C. (2012). Orchestrating life skills: The effect of increased school-based music classes on children’s social competence and self-esteem. International Journal of Music Education, 31(3), 292-309. doi: 10.1177/0255761411434824 Rickard, N. S., Bambrick, C. J., & Gill, A. (2012). Absence of widespread psychosocial and cognitive effects of school-based music instruction in 10–13-year-old students. International Journal of Music Education, 30(1), 1-22. Rickson, D. J., & Watkins, W. G. (2003). Music therapy to promote prosocial behaviors in aggressive adolescent boys-a pilot study. Journal of Music Therapy, 40(4), 283-301. Ringen, S. (1995). Well-being, measurement, and preferences. Scandinavian Sociological Association, 38(1), 3-15. Robinson, N. R. (2004). Who Is "At Risk" in the Music Classroom? Music Educators Journal, 90(4), 38-43. Roden, I., Kreutz, G., & Bongard, S. (2012). Effects of a school-based instrumental music program on verbal and visual memory in primary school children: a longitudinal study. Frontiers in Psychology, 3, 1-9. doi: 10.3389/fpsyg.2012.00572 Rolvsjord, R. (2010). Resource-Oriented Music Therapy in Mental Health Care. Gilsum, NH: Barcelona Publishers. Rusinek, G. (2008). Disaffected learners and school musical culture: an opportunity for inclusion. Research Studies in Music Education, 30(1), 9-23. Ryan, C., & Sartbayeva, A. (2011). Young Australians and Social Inclusion. Australian Social Policy Journal, (10), 1-26. Retrieved from http://ssrn.com/abstract=2012859 Sagor, R., & Cox, J. (2004). At-risk students: Reaching and teaching them (2nd ed.). Larchmont, NY: Eye on Education, Inc. Samdal, O., Nutbeam, D., Wold, B., & Kannas, L. (1998). Achieving health and educational goals through schools—a study of the importance of the school climate and the students' satisfaction with school. Health Education Research, 13(3), 383-397. doi: 10.1093/her/13.3.383

313

Saraceno, C. (2002). Social exclusion: cultural roots and diversities of a popular concept. Paper presented at the Conference on social exclusion and children, New York. Saunders, P. (2011). Down and Out: Poverty and Exclusion in Australia. Bristol: The Policy Press. Saunders, P. (2013). Reflections on the Concept of Social Exclusion and the Australian Social Inclusion Agenda. Social Policy & Administration, 47(6), 692-708. doi: 10.1111/spol.12038 Saunders, P., Naidoo, Y., & Griffiths, M. (2008). Towards new indicators of disadvantage: deprivation and social exclusion in Australia. The Australian Journal of Social Issues, 43(2), 175-194. Schellenberg, E. G. (2004). Music lessons enhance IQ. Psychological Science, 15(4), 511–514. Schellenberg, E. G. (2005). Music and cognitive abilities. Current Directions in Psychological Science, 14(6), 317-320. Schellenberg, E. G. (2011a). Examining the association between music lessons and intelligence. British Journal of Psychology, 102(3), 283-302. doi: 10.1111/j.2044-8295.2010.02000.x Schellenberg, E. G. (2011b). Music lessons, emotional intelligence, and IQ. Music Perception, 29(2), 185-194. Schütz, A. (1951). Making music together: A study in social relationship. Social Research, 18(1), 76-97. doi: 10.2307/40969255 Schwarz, N. (1999). Self-reports: how the questions shape the answers. American Psychologist, 54(2), 93-105. Schwarz, N., Knäuper, B., Hippler, H.-J., Noelle-Neumann, E., & Clark, L. (1991). Rating scales numeric values may change the meaning of scale labels. Public Opinion Quarterly, 55(4), 570-582. Scutella, R., & Wilkins, R. (2010). Measuring social exclusion in Australia: Assessing existing data sources. [Article]. Australian Economic Review, 43(4), 449-463. doi: 10.1111/j.1467-8462.2010.00611.x Seddon, T. L. (2008). Remaking civic formation: Transforming politics and the cosmopolitan school. In B. Lingard, J. Nixon & S. Ranson (Eds.), Transforming Learning in Schools and Communities: The Remaking of Education for a Cosmopolitan Society (pp. 152-169). UK: Continuum International. 314

Seeman, M. (1983). Alienation motifs in contemporary theorizing: The hidden continuity of the classic themes. Social Psychology Quarterly, 46(3), 171-184. Sen, A. (1980). Equality of What? In A. Sen (Ed.), Choice, Welfare and Measurement. Cambridge, MA: MIT Press. Sen, A. (1988). Freedom of choice: Concept and content. European Economic Review, 32(2–3), 269-294. doi: 10.1016/0014-2921(88)90173-0 Sen, A. (1998). Mortality as an indicator of economic success and failure. Economic Journal, 108(446), 1-25. Shadish, W. R., & Luellen, J. K. (2005). Quasi-experimental Designs. In B. Everitt & D. C. Howell (Eds.), Encyclopedia of Statistics in Behavioral Science: John Wiley & Sons, Ltd. Shapiro, A., & Keyes, C. L. M. (2008). Marital status and social well-being: Are the married always better off? [Article]. Social Indicators Research, 88(2), 329-346. doi: 10.1007/s11205-007-9194-3 Shenton, A. K. (2004). Strategies for ensuring trustworthiness in qualitative research projects. Education for Information, 22(2), 63-75. Sheskin, D. J. (2003). Handbook of parametric and nonparametric statistical procedures (3rd ed.). New York, NY: Chapman & Hall/CRC. Shields, C. (2001). Music education and mentoring as intervention for at-risk urban adolescents: Their self-perceptions, opinions, and attitudes. Journal of Research in Music Education, 49(3), 273-286. doi: 10.2307/3345712 Shrader-Frechette, K. S. (1994). Ethics of scientific research. Maryland, USA: Rowman & Littlefield. Sinkovics, R. R., & Ghauri, P. N. (2008). Enhancing the trustworthiness of qualitative research in international business. Management International Review, 48(6), 689-714. Small, C. (1998). Musicking. Hanover, NH: Wesleyan University Press. Smith, B. (2004). ‘Boys Business’: an unusual northern Australian music program for boys in the middle years of schooling. International Journal of Music Education, 22(3), 230-236. doi: 10.1177/0255761404047399 Smith, P. K. (2010). Bullying in primary and secondary schools. In S. R. Jimerson, S. M. Swearer & D. L. Espelage (Eds.), International Perspectives on Bullying Prevention and Intervention (pp. 137-150). New York, NY: Routledge.

315

Smyth, P. (2010). In or Out? Building an Inclusive Nation. Albert Park, Victoria: The Brotherhood of St Laurence. Springer, A., Parcel, G., Baumler, E., & Ross, M. (2006). Supportive social relationships and adolescent health risk behavior among secondary school students in El Salvador. Social Science & Medicine, 62(7), 1628-1640. doi: 10.1016/j.socscimed.2005.08.018 Stanley, P. (2014). Writing the PhD Journey(s): An Autoethnography of Zine-Writing, Angst, Embodiment, and Backpacker Travels. Journal of Contemporary Ethnography. doi: 10.1177/0891241614528708 Stecklov, G., Weinreb, A., & Sana, M. (2013). Effects of Interviewer-Respondent Familiarity on Contraceptive Use and Abortion Data. Paper presented at the Population Association of America, New Orleans, LA. http://www.iussp.org/sites/default/files/event_call_for_papers/strangers_fp paper 6gs.pdf Stevens, R. S. (1978). Music in state-supported education in New South Wales and Victoria, 1848-1920. (Doctoral dissertation), The University of Melbourne. Retrieved from http://hdl.handle.net/11343/36510 Stevens, R. S., & Stefanakis, M. (2014). Filling the Gaps: What Research is Needed to Assist with Music Education Advocacy in Australia. Journal of Music Research Online, 5, 1-13. Stewart, P. (2007). The true intrinsic value of music study. American Music Teacher, 56(5), 4-5. Stige, B. (2010). Musical Participation, Social Space, and Everyday Ritual. In B. Stige, G. Ansdell, C. Elefant & M. Pavlicevic (Eds.), Where Music Helps: Community Music Therapy in Action and Reflection (pp. 125-147). Burlington, USA: Ashgate Publishing Company. Stige, B., & Aarø, L. E. (2012). Invitation to Community Music Therapy. New York, NY: Routledge. Stige, B., Ansdell, G., Elefant, C., & Pavlicevic, M. (2010). Where Music Helps: Community Music Therapy in Action and Reflection. Burlington, USA: Ashgate Publishing Company. Stige, B., Malterud, K., & Midtgarden, T. (2009). Toward an agenda for evaluation of qualitative research. Qualitative Health Research, 19(10), 1504-1516.

316

Tait, H. (2004). Well-being of the non-reserve Aboriginal population (Article). (72). Retrieved 10 June 2012, from Statistics Canada http://www5.statcan.gc.ca/bsolc/olc-cel/olc-cel?catno=11-008X20030046805&lang=eng Taket, A., Crisp, B. R., Nevill, A., Lamaro, G., Graham, M., & Barter-Godfrey, S. (Eds.). (2009). Theorising social exclusion. Abingdon: Routledge. Taylor, C. (2008). "Both sides, now”: Balancing the extrinsic and intrinsic benefits of music in music advocacy. Canadian Music Educator / Musicien Educateur au Canada, 49(3), 36-38. Teachout, D. J. (2005). The impact of music education on a child’s growth and development. In D. A. Hodges (Ed.), Sounds of learning. Carlsbad, CA: International Foundation for Music Research. Teddlie, C., & Tashakkori, A. (2006). A general typology of research designs featuring mixed methods. Research in the Schools, 13(1), 12-28. Teddlie, C., & Yu, F. (2007). Mixed methods sampling: A typology with examples. Journal of Mixed Methods Research, 1(1), 77-100. Teese, R., & Polesel, J. (2003). Undemocratic schooling: Equity and quality in mass secondary education in Australia. Melbourne, Victoria: Melbourne University Press. Thomas, L. (2009). New challenges in school-based research. Health Promotion Journal of Australia: Official Journal of Australian Association of Health Promotion Professionals, 20(1), 73-74. Tilley, A. (1999). An Introduction to Research Methodology and Report Writing in Psychology. Brisbane, Australia: Pinapple Press. Toepoel, V., Das, M., & Van Soest, A. (2008). Effects of design in web surveys comparing trained and fresh respondents. Public Opinion Quarterly, 72(5), 9851007. Tomyn, A. J. R. (2013). The subjective wellbeing of indigenous Australian adolescents: validating the personal wellbeing index-school children. Social Indicators Research, 110(3), 1013-1031. doi: 10.1007/s11205-011-9970-y Torrance, H. (2011). Qualitative research, science and government: Evidence, criteria, policy and politics. In N. Denzin & Y. Lincoln (Eds.), The SAGE handbook of qualitative research (pp. 569-580). Thousand Oaks, CA: SAGE Publications Inc.

317

Turner, R. J., & Roszell, P. (1994). Psychosocial resources and the stress process. In W. Avison, R. & I. Gotlib, H. (Eds.), Stress and mental health (pp. 179-210). New York: Plenum. UNESCO. (1999). International appeal for the Promotion of the Arts in education. Paper presented at the General UNESCO conference, Paris. UNESCO. (2006). Building creative capacities for the 21st century. Paper presented at the World Conference on Arts Education, Lisbon, Portugal. Vaughan, T., Harris, J., & Caldwell, B. J. (2011). Bridging the gap in school achievement through the Arts: Summary report. Melbourne: Song Room. Retrieved from http://www.songroom.org.au/index.php?option=com_content&view=article&id =149:publications-and-reports&catid=31&Itemid=309 VCAA. (2012). Progression points examples for AusVELS The Arts - Music Level 8–10. Victorian Curriculum and Assessment Authority, State Government of Victoria. Retrieved from http://www.vcaa.vic.edu.au/Pages/foundation10/curriculum/resources/arts.aspx Veenhoven, R. (2002). Why social policy needs subjective indicators. Social Indicators Research, 58(1-3), 33-46. Veit, C. T., & Ware, J. E. (1983). The structure of psychological distress and well-being in general populations. Journal of Consulting and Clinical Psychology, 51(5), 730. Vickers, M. (2005). In the common good: the need for a new approach to funding Australia's schools. Australian Journal of Education, 49(3), 264-277. Victorian Government. (2010). The state of Victoria's children 2009. Melbourne: Department of Education and Early Childhood Development. Retrieved from https://http://www.eduweb.vic.gov.au/edulibrary/public/govrel/Policy/children/s ovc-2009.pdf Victorian Government. (2011). The state of Victoria's children 2010. Melbourne: Department of Education and Early Childhood Development. Retrieved from http://www.education.vic.gov.au/Documents/about/research/sovc2010.pdf Vitale, J. L. (2011). Music makes you smarter: A new paradigm? Perceptions and perspectives from four groups of elementary education stakeholders. Canadian Journal of Education / Revue canadienne de l'éducation, 34(3), 317-343. VSN International. (2008). GenStat. Hemel Hempstead, UK: VSN International. 318

Wang, M. C., & Gordon, E. W. (1994). Educational resilience in inner-city America: Challenges and prospects. Hillsdale, N.J.: L. Erlbaum Associates. Weinreb, A. A. (2006). The limitations of stranger-interviewers in rural Kenya. American Sociological Review, 71(6), 1014-1039. Weinreb, A. A., Sana, M., & Stecklov, G. (2011). Do insiders or outsiders collect better data? Results from a methodological experiment. Paper presented at the Population Association of America, Washington, D.C. Welch, G. F., Himonides, E., Saunders, J., Papageorgi, I., Rinta, T., Preti, C., . . . Hill, J. (2011). Researching the first year of the National Singing Programme Sing Up in England: An initial impact evaluation. Psychomusicology, 21(1/2), 83-97. West, C. (2012). Teaching music in an era of high-stakes testing and budget reductions. [Article]. Arts Education Policy Review, 113(2), 75-79. doi: 10.1080/10632913.2012.656503 Wetz, J. (2004). Promoting inclusion in school through the arts: a case study. Support for Learning, 19(2), 66-70. White, S. C. (2008). But What Is Wellbeing? A Framework for Analysis in Social and Development Policy and Practice. Paper presented at the Conference on Regeneration and Wellbeing: Research into Practice, University of Bradford. Wigram, T. (2004). Improvisation: Methods and techniques for music therapy clinicians, educators, and students. New York, NY: Jesseca Kingsley Publications. Williams, R. (1985). Keywords: A vocabulary of culture and society. New York, NY: Oxford University Press. Wills, R. (2011). The magic of music: a study into the promotion of children’s well‐ being through singing. International Journal of Children's Spirituality, 16(1), 37-46. Wilson, L. (2006). Developing a model for the measurement of social inclusion and social capital in regional Australia. Social Indicators Research, 75(3), 335-360. Winner, E., & Hetland, L. (2000). The arts in education: Evaluating the evidence for a causal link. Journal of Aesthetic Education, 34(3), 3-10. Wyn, J. (2009a). The changing context of Australian youth and its implications for social inclusion. Youth Studies Australia, 28(1), 46-50.

319

Wyn, J. (2009b). Touching the future: Building skills for life and work. Australian Education Review. No. 55, Melbourne: Australian Council for Educational Research Press. Wyn, J. (2009c). Youth health and welfare: The cultural politics of education and wellbeing. Melbourne: Oxford University Press. Wyszomirski, M. (2004). Defining and developing creative sector initiatives. Columbus, OH: The Ohio State University. Yalom, I. D. (1995). The theory and practice of group psychotherapy (4th ed.). New York: Basic Books. Youn, G. (1996). Sexual activities and attitudes of adolescent Koreans. Archives of Sexual Behavior, 25(6), 629-643. Young, R. A., & Collin, A. (2004). Introduction: Constructivism and social constructionism in the career field. Journal of Vocational Behavior, 64(3), 373388. Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52(1), 30-41.

320

Appendix A: Theoretical Approach to Knowledge Accumulation This project adopted a theoretical approach to knowledge generation commensurate with both the participatory paradigm (in that it was generated through reflection) and critical theory (in that it employed the process of critique). The process in which knowledge was generated, however, comprised two main stages. While these stages were sequential, the decision to carry out the second stage emerged during the candidature, and was driven by the lack of meaning procured in the first stage (as described in the introduction chapter). The first stage involved the analysis of both quantitative and qualitative data collected for two separate studies (detailed descriptions of this process are provided in Chapter 6, 7 and 8). The express goal of this first stage of investigation was to demonstrate the psychosocial benefits of musical participation in mainstream schools. However, analyses showed inconsistent results. They also suggested the research approaches were problematic. This realisation, combined with a critical reading of existing literature in the area, prompted a reflexive evaluation of the original project focus. It became clear that reporting these inconsistent and insignificant results would add very little to the field, and that the best way to use this data was to challenge and critically investigate the limitations of dominant research approaches in this area. Subsequently, the decision was made that the second stage of the project would involve two critical reflection analyses of the research approaches used in the first stage. These critical analyses were undertaken in order to identify the challenges of investigating the psychosocial wellbeing benefits of musical participation in mainstream schools, with the goal of generating recommendations for future research in this area. It is these identified challenges, and subsequent recommendations, that represent the knowledge accumulated and communicated in this dissertation, and it is through the processes of reflection and critique that they were generated. Collecting data Data for the first stage of the project was collected using standard quantitative and qualitative methods (see Chapter 6). The initial purpose of collecting this data was to gather empirical evidence to demonstrate the benefits of two different school music programs. This approach to data collection is sometimes referred to as an empirical, or evidence-based model of knowledge generation (Denzin, 2009). In this case, data collection aimed to gather proof that music programs can be effective as school-based interventions for wellbeing and education outcomes. The intended use of such proof was to lobby schools and policymakers as to the potential of implementing music programs in this setting. Results from this first evidence-based stage of data collection were inconsistent, and did not provide conclusive evidence as to the potential of the music programs investigated. This inconclusive evidence seemed to be based on several research challenges observed during this study. It was at this point that the decision was made to undertake critical reflection analyses of these two studies in order to identify the

321

challenges of undertaking research in this field. These critical reflection analyses become the second stage of the PhD project. The method of data collection for this second stage was markedly different from the first stage. Data from the first stage, as well as information and observations about how and where it was collected, was organised to facilitate reflection and critique. It was this process of reflection and critique that constituted the second stage of analysis. The reappropriation of data from the first stage to inform the critical reflection analyses was considered necessary. The many challenges to investigation observed in the first stage were embodied both within the data, as well as in the procedures used to collect and analyse it. Further, the methods used in this first stage reflect the methods, data, and general approaches to research that dominate this field, thus providing valuable insight into the challenges of established research approaches in this field more generally. For example, quantitative data and methods have both dominated prior research in this area, and are most commonly linked to the reporting of inconsistent results (see Chapter 4). This links with the critical theory position, where established processes or beliefs are critiqued and revised. It also aligns with participatory notions of deconstruction, in which notions of what is right, normal or appropriate are challenged through research (Lincoln, Lynham, & Guba, 2011). The organisation of data and information (or data collection process) used to inform the second stage of analysis also focused on assembling qualitative and quantitative reports in a way that they could be connected and compared. This combined approach is coherent with the critical theory position which suggests researchers can use both qualitative and quantitative data in order to approach “research with different styles and methods that can produce multiple forms of data” (Lincoln et al., 2011, p. 111). Here, this approach partly aimed to provide different perspectives on the reported outcomes of the program. For example, it used both quantitative and qualitative data to gain deeper insight into how benefits such as emotional control were experienced by participants. By comparing reports from each method, it also enabled a critical analysis of how suitable the different reporting methods (i.e. interviews and surveys) were for capturing constructs. This combined use, or assemblage, of qualitative and quantitative methods also affiliates closely with notions such as triangulation, often attributed to the field of mixed methods research (Fielding, 2012). The combined use of qualitative and quantitative approaches to assess the suitability of data collection techniques was also used in a slightly different way. Instead of simply comparing reports from each method, participant experiences of the survey (reported during interviews) were used to critically examine their use. As such, subjective understandings were used to assess the suitability of a tool designed for generating objective knowledge. This process was underpinned by the recognition that while surveys produce objective data, they are understood and experienced subjectively by each participant. This position is congruent with the participatory notion of subjective-objective reality (Heron & Reason, 1997). In order to better understand the potential limitations of these objective approaches, it was considered important to acknowledge the potential for them to be perceived differently by different people (i.e.

322

the potential for competing mental constructions of their efficacy and the concepts contained within them). To carry out this interrogation in a way that fully accounted for varying versions of reality necessitated an acknowledgment of the lived and practical experience of the players (staff, students and researchers) in the community of interest (the school), rather than relying on the abstracted knowledge of experts. This explains the use of interview data collected from students and teachers in the field. This commitment was further fuelled by the desire to account for contextual elements, which in turn led to the decision to include observational reports from school staff and the researchers that had been recorded along the way. This collection and use of interview, observational and contextual data links to Lincoln et al. (2011) and Heron and Reason’s (1997) concept of practical or experiential “knowing” within the participatory paradigm, where “worldview is based on participation and participative realities” (Lincoln et al., 2011, p. 102). In summary, while the data used in this thesis was originally collected following an evidence-based research model (stage one), the reappropriation of this data was seen as appropriate for informing the critical reflection analyses (stage two). This is because data from stage one either revealed or provided insight into many of the research challenges that stage two aimed to uncover. When assembling data to inform the second stage, observational data was also included in order to provide different perspectives, as well as to provide contextual information. Working with data Following the two-stages of data collection previously outlined, there were also two stages of analysis in this project, each with its own style and objective. The first used empirical analysis methods: standard quantitative and qualitative techniques used for survey and interview data (see Chapter 6). The results of these initial analyses (see Chapters 7 and 8) are not the outcomes of the project, but provided an important part of the content which was used to inform the second stage of analysis. This second stage of analysis not only utilised the results of the empirical data analysis in stage one, but also observational data collected throughout the project. Together, these formed the basis of two systematic critical reflection analyses. It is through these secondary analyses that the challenges to investigating the psychosocial wellbeing benefits of school-based musical participation were identified. These challenges (and the recommendations they informed) are the primary results of this project. The theoretical rationale for using critical reflection analysis as the main method for generating knowledge is again linked to both participatory and critical theory paradigms. From a critical theory standpoint, critique is considered a tool or methodology capable of enacting social change and social justice through the process of scrutiny (Strydom, 2011) and revisionism (Bronner, 2011). This process includes identifying how systemic beliefs or assumptions can inhibit the maintenance of core human values such as flourishing or equity (Lincoln et al., 2011).

323

Within a participatory framework, Lincoln et al. (2011) suggest critical reflection is captured in Kilgore’s notion of deconstruction; a “powerful postmodern tool for questioning prevailing representations […] in the […] literature” (2001, p. 56). Kilgore states: The purpose of deconstruction is to identify and discredit the false binaries that structure a communication or “discourse”; that is, to challenge the assertions of what is to be included or excluded as normal, right, or good. Deconstruction does not conclude with a piece of emancipatory knowledge […] but rather finds meaning continually problematic (Lather, 1991). (2001, p. 56) Kilgore’s notion of deconstruction is paralleled in the “critically reflective analysis” process proposed by Campbell and Baikie (2013). They suggest this approach is “grounded in an epistemology that understands knowledge to be socially constructed and therefore open to deconstruction and reconstruction” (Campbell & Baikie, 2013, p. 453). They go on to claim that such deconstruction should seek “provisional” rather than “definitive” knowledge, which can be used to approach the goal of social justice. The above descriptions of critique, reflection and deconstruction reflect how these concepts and processes are used in this dissertation as tools to critically appraise existing approaches to investigating links between psychosocial wellbeing and school music. Used together, these tools aimed to identify how the individual elements of existing research approaches may potentially present challenges to such investigation. They also aimed to illuminate any systemic presuppositions that underlie such approaches. References Bronner, S. E. (2011). Critical theory: A very short introduction. New York, NY: Oxford University Press. Campbell, C., & Baikie, G. (2013). Teaching critically reflective analysis in the context of a social justice course. Reflective Practice, 14(4), 452-464. Denzin, K. N. (2009). The elephant in the living room: or extending the conversation about the politics of evidence. Qualitative Research, 9(139), 139-160. Fielding, N. G. (2012). Triangulation and Mixed Methods Designs: Data Integration With New Research Technologies. Journal of Mixed Methods Research, 6(2), 124-136. doi: 10.1177/1558689812437101 Heron, J., & Reason, P. (1997). A participatory inquiry paradigm. Qualitative Inquiry, 3(3), 274-294. Kilgore, D. W. (2001). Critical and postmodern perspectives on adult learning. New Directions for Adult and Continuing Education, 2001(89), 53-62. doi: 10.1002/ace.8 Lincoln, Y. S., Lynham, S. A., & Guba, E. G. (2011). Paradigmatic controversies, contradictions, and emerging confluences, revisited. In N. K. Denzin & Y. S. Lincoln (Eds.), The SAGE handbook of qualitative research (4th ed., pp. 97128). Thousand Oaks, CA: SAGE Publications Inc. Strydom, P. (2011). Contemporary Critical Theory and Methodology. New York. NY: Taylor & Francis.

324

Appendix B: Keyes to Success Program Protocol

325

Appendix C: Information Statement for Students

326

Appendix D: Information Statement for Families

327

328

Appendix E: Consent Form for Parents

329

Appendix F: University of Melbourne Ethics Approval letter

330

Appendix G: Department of Education and Early Childhood Development Ethics Approval letter

331

332

Appendix H: Keys to Success Survey

333

334

335

336

337

338

339

340

341

342

343

344

345

Appendix I: Keys to Success Scale Scoring Information Measures were used in the Keys to Success survey instrument came from number of different sources. As such, several scoring methods were used. For consistency, all protective factors were scored so that high scores were positive, and increased scores at follow-up indicated improvement. For all risk factors, low scores were positive, and decreased scores at follow-up indicated improvement. The first nine scales were taken from the CTC survey. The Cronbach alpha levels for these scales stated here were reported by Bond et al. (2000) using an Australian sample. Reliability and source information for the remaining nine measures were reported where possible. Community Low neighbourhood attachment. Psychosocial wellbeing in the community domain was measured using three CTC subscales designed to measure a student’s perception of their connection to, and engagement in, their neighbourhood or community. This included the Low neighbourhood attachment scale, a risk factor consisting of three items scored on a four-point response scale (α = 0.80). For analysis, the first item was reverse scored and the mean of all three items was calculated to create the scale score. Participants responding to less than two items received a missing value for this scale. Opportunities for prosocial involvement. The Opportunities for prosocial involvement in the community scale was a protective factor and consisted of five items. Four items used two-point response scales, while the fifth used a four-point scale (α = 0.68). The first four items were recoded to reverse scores for analysis, and place responses onto a four-point scale by making the first response option (yes) equal to a score of four, and the second option (no) equal to a score of one. The fifth scale item was omitted from the present tool due to an administration error and therefore calculating the mean of the four remaining items created the scale score. Participants responding to two or less items received a missing value for this scale. Rewards for prosocial involvement. The last measure in the community domain was the Rewards for prosocial involvement in the community scale. This was a protective factor, which consisted of three items scored on a four-point scale (α = 0.82). All items were reverse scored for analysis, and the combined mean constituted the scale score. Participants responding less than two items received a missing value for this scale. School Academic failure. The first CTC scale used to assess psychosocial wellbeing in the school domain was the two-item Academic failure scale. As a risk factor, it assessed a student’s risk of academic failure. The first item used a five-point response scale, which was recoded

346

using decimals to transpose scores onto the same four-point scale used by the second item. Calculating the mean of both items created the scale score. Only participants who responded to both items received a score for this scale. Low commitment to school. The second measure used in the School domain was the Low commitment to school scale. This scale was a protective factor containing nine-items, three scored on a seven-item response scale, and six on a five-item response scale (α = 0.76). The first three items were rescored using decimals to transpose scores into a five-point scale, consistent with the remaining six items. The eighth item was reverse scored, and the mean of all nine items created the scale score. Participants responding to less than six items received a missing value for this scale. Opportunities for Prosocial Involvement. The Opportunities for prosocial involvement in school scale was also used for the school domain. Also a protective factor, it has five items, each scored on a fourpoint response scale (α = 0.70), and each of which were reverse scored for analysis. The mean of all five items created the scale score. Participants responding to less than four items received a missing value for this scale. Rewards for prosocial involvement. The Rewards for prosocial involvement scale, a protective factor, was the last scale used for the School domain. It consisted of four items, all scored on a four-point response scale (α = 0.74), and all reverse scored for analysis. The mean of all items created the scale score. Participants responding to less than three items received a missing value for this scale. Peer and individual Social skills. The Peer and individual domain included two protective factors from the CTC survey aimed to measure positive social behaviour, such as respect, honesty, awareness of others, communication and social problem solving (Arthur, Hawkins, Pollard, Catalano, & Baglioni, 2002). These included the Social skills scale, which consisted of four items on a four-point response scale (α = 0.59). Items were recoded individually for analysis so that high scores reflected high social skills. Calculating the mean of all items created the scale score. Participants responding to less than three items received a missing value for this scale. Belief in the moral order. The second protective factor used in the Peer and individual domain was the Belief in the moral order scale (α = 0.58). This used four items scored on a four-point response scale, the last of which was reverse scored. Calculating the mean of the four items created the scale score. Participant responding to less than three items received a missing value for this scale.

347

Emotional control Emotional control. Emotional control was measured using the four-item Emotional control scale, and was reported as a protective factor against things like adolescent substance use (Prior, Sanson, Smart, & Oberklaid, 2000). The scale was adapted from the Swinburne University Emotional Intelligence Test (SUEIT; Palmer & Stough, 2001) for use in large scale adolescent health studies, such as the Australian Temperament Project (ATP; Prior et al., 2000). The Dartington Social Research Unit (SRU) reported two alpha values for this scale; one at 0.74, and another from a feasibility pilot study of 0.83 (SRU, 2013). Each item used a four-point response scale and was reverse scored for analysis. Calculating the mean of all items created the overall scale score. Participant’s responding to less than three items received a missing value for this scale. Social support Multidimensional Scale of Perceived Social Support – significant other (MSPSS-SO). Social support was measured using a four item significant other subscale taken from the full version of the Multidimensional Scale of Perceived Social Support (MSPSS) developed by Zimet, Dahlem, Zimet, and Farley (1988). They claim it differed from pre-existing social support measurement tools in that it is a “subjective assessment of social support adequacy” (Zimet et al., 1988, p. 32). It is also said to be less onerous to administer than other similar scales (Canty-Mitchell & Zimet, 2000; Zimet et al., 1988), and distinctive in that it contains items relating to the social support of a significant other, which is important for adolescent populations (Canty-Mitchell & Zimet, 2000). Zimet et al. (1988) published test-retest reliability statistics for the Significant other subscale as of 0.72, and a Cronbach’s alpha of 0.91, with a population of university students. A later study by Canty-Mitchell and Zimet (2000) also reported an alpha of 0.91 for the Significant other subscale with a population of lower socioeconomic adolescents. Each item was scored on a seven-point scale, the mean of which was used as the scale score. Cases not responding to all items received a missing value for this scale. Trusted adult in life. Social support was also assessed using a one-item measure taken from the HowRU (DEECD, 2010) survey. It has been termed here as Trusted adult in life, and categorised as a protective factor. It used a seven-point scale, and the raw score provided the final scale score. Psychological distress Kessler Psychological Distress Scale (K10). Psychological distress was measured using the Kessler Psychological Distress Scale (K10), a ten-item risk factor scale used to assess the psychological distress of an individual. The K10 has been used widely with many populations (Andrews & Slade,

348

2001), including a sample of Australian adolescents, for which a Cronbach alpha of 0.89 was reported (Hides, Lubman, Elkins, Catania, & Rogers, 2007). It is considered particularly important here as psychological health as a construct has been shown to be a key indicator of psychosocial wellbeing (Keyes, 2007). Scores from all ten items were added to create the total scale score, and therefore participants needed to answer all ten items to receive a score. Results were used to determine an overall psychological distress score (10-50) for both participants and groups. Psychological wellbeing Mental Health Index Psychological Wellbeing (MHI-PWB) Psychological wellbeing was measured using the Mental Health Index Psychological Wellbeing (MHI-PWB) scale. The MHI-PWB is one of two global summary scales taken from the 38 item Mental Health Index (Veit & Ware, 1983; Ware Jr, Manning Jr, Wells, Duan, & Newhouse, 1984), and contains 14 items measuring positive psychological states, with a reported Cronbach alpha range of 0.91 to 0.92 (Florian, Mikulincer, & Taubman, 1995). Following standard practice, (Davies, Sherbourne, Peterson, & Ware, 1988; Department of Health and Ageing, 2003) all items were reverse scored for analysis and summed to give the scale score. Only cases responding to all 14 items received a score. Psychological Needs Basic Psychological Needs – Short Form (BPN-SF). The extent to which a young person perceived their psychological needs were being met was measured using the Basic Psychological Needs – Short Form (BPN-SF), a reduced version of the Basic Needs Satisfaction in General Scale (BNSG-S; Gagné, 2003; Johnston & Finney, 2010). Based on self determination theory (Ryan & Deci, 2000, 2001), this scale aims to measure a respondent’s level of satisfaction with their psychological needs in general, and in different areas of their life. While the original version of this scale contained 21 items and used a seven-point response scale (Gagné, 2003), the BPN-SF, as used by Samman (2007) and Steger and Samman (2012), uses nine items with a four-point response scale to form an overall measurement of psychological needs, as well as three subscales. These subscales use three items each to assess needs in relation to autonomy (“I feel like I am free to decide how I live my life”, α = 0.93), competence (“Most days I feel a sense of accomplishment from what I do”, α = 0.82), and relatedness (“People in my life care about me”, α = 0.94). The version of the BPN-SF used here (see Table 6.1), however, was taken from the Victorian HowRU (DEECD, 2010), and differs in that it uses the original response scoring (Johnston & Finney, 2010). Also, wording for the last item of the competence subscale was changed from “I often feel capable” to “I do not often feel very capable”. The scale was scored by reverse scoring the last competence item so that a high score on each item indicated a high level of psychological needs satisfaction. Summing the score of all nine items created an overall scale score. Summing scores for the three

349

items in each subscale created respective subscale scores. Scores were calculated only when each respondent had responded to all scale, or subscale items.

References Andrews, G., & Slade, T. (2001). Interpreting scores on the Kessler Psychological Distress Scale (K10). Australian and New Zealand Journal of Public Health, 25(6), 494-497. Arthur, M. W., Hawkins, J. D., Pollard, J. A., Catalano, R. F., & Baglioni, A. J. (2002). Measuring risk and protective factors for use, delinquency, and other adolescent problem behaviors: The Communities That Care Youth Survey. Evaluation Review, 26(6), 575-601. Bond, L., Thomas, L., Toumbourou, J., Patton, G., & Catalano, R. (2000). Improving the Lives of Young Victorians in Our Community: a survey of risk and protective factors. Melbourne: Centre for Adolescent Health. Canty-Mitchell, J., & Zimet, G. D. (2000). Psychometric properties of the Multidimensional Scale of Perceived Social Support in urban adolescents. American Journal of Community Psychology, 28(3), 391-400. Davies, A. R., Sherbourne, C. D., Peterson, J., & Ware, J. E. (1988). Scoring manual, adult health status and patient satisfaction measures used in RAND's health insurance experiment. Santa Monica, CA: RAND Corporation. DEECD. (2010). The State of Victoria's Children 2009: Aboriginal children and young people in Victoria. Victoria, Australia: Department of Education and Early Childhood Development. Department of Health and Ageing. (2003). Mental Health National Outcomes and Casemix Collection: Overview of clinician rated and consumer self-report measures, Version 1.50. Canberra: Department of Health and Ageing. Florian, V., Mikulincer, M., & Taubman, O. (1995). Does hardiness contribute to mental health during a stressful real-life situation? The roles of appraisal and coping. Journal of Personality and Social Psychology, 68(4), 687-695. Gagné, M. (2003). The role of autonomy support and autonomy orientation in prosocial behavior engagement. Motivation and Emotion, 27(3), 199-223. Hides, L., Lubman, D. I., Elkins, K., Catania, L. S., & Rogers, N. (2007). Feasibility and acceptability of a mental health screening tool and training programme in the youth alcohol and other drug (AOD) sector. Drug and Alcohol Review, 26(5), 509-515. doi: 10.1080/09595230701499126 Johnston, M. M., & Finney, S. J. (2010). Measuring basic needs satisfaction: Evaluating previous research and conducting new psychometric evaluations of the Basic Needs Satisfaction in General Scale. Contemporary Educational Psychology, 35(4), 280-296. Keyes, C. L. M. (2007). Promoting and protecting mental health as flourishing: A complementary strategy for improving national mental health. American Psychologist, 62(2), 95-108. doi: 10.1037/0003-066x.50.12.965. Palmer, B. R., & Stough, C. (2001). The Swinburne University Emotional Intelligence TEST (SUEIT) technical manual. Melbourne, Victoria: Organisational Research Unit, Swinburne University of Technology. Retrieved from http://hdl.handle.net/1959.3/2922

350

Prior, M. R., Sanson, A., Smart, D., & Oberklaid, F. (2000). Pathways from infancy to adolescence: Australian Temperament Project 1983-2000. (0642394784). Melbourne, Victoria: Australian Institute of Family Studies. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68-78. Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A review of research on hedonic and eudaimonic well-being. Annual Review of Psychology, 52(1), 141-166. Samman, E. (2007). Psychological and Subjective Well-being: A Proposal for Internationally Comparable Indicators. Oxford Development Studies, 35(4), 459486. SRU. (2013). Evidence2Success School Well-being Survey Item Construct Dictionary. Retrieved 13th February, 2014, from Dartington Social Research Unit http://www.pkc.gov.uk/CHttpHandler.ashx?id=21362&p=0 Steger, M. F., & Samman, E. (2012). Assessing meaning in life on an international scale: Psychometric evidence for the Meaning in Life Questionnaire-Short Form among Chilean households. International Journal of Wellbeing, 2(3), 182-195. Veit, C. T., & Ware, J. E. (1983). The structure of psychological distress and well-being in general populations. Journal of Consulting and Clinical Psychology, 51(5), 730-742. Ware Jr, J. E., Manning Jr, W. G., Wells, K. B., Duan, N., & Newhouse, J. P. (1984). Health status and the use of outpatient mental health services. American Psychologist, 39(10), 1090-1100. Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52(1), 30-41.

351

Appendix J: Information Statement or Professionals

352

353

Appendix K: Consent Form or Professionals

354

Appendix L: Bereavement Group Program Protocol

355

Appendix M: Bereavement Group Scale Scoring Information The same scoring procedures reported for the KTS study were used for this study, with the exception of two measures. Only the scoring of these two measures is described here. The Cronbach alpha levels stated for these scales were again reported by Bond et al. (2000). School Low commitment to school. The version of the CTC Low commitment to school scale used in this study contained seven items rather than nine items used in the original version. Of these items, the first was scored on a seven-item response scale, and the remaining six on a five-item response scale (α = 0.76). The first item was rescored using decimals to transpose scores into a five-point scale, consistent with the remaining six items. The sixth item was reverse coded so a high score on all items indicated high risk of “low commitment to school”. The mean of the seven items created the scale score. This was done for each case that had responded to at least six scale items, with missing scores scored as a zero. Participants responding to six or less items received a missing value for the scale. As a risk factor, reduced scores at follow-up indicated improvement. Educational expectation. Along with the measures included in the KTS study, psychosocial wellbeing in the School domain was also measured using a one-item measure taken from the HowRU survey. Termed here as Educational expectation, it is designed to measure the level of education a young person hopes to achieve. The raw score of the item was used as final score, with a higher score indicating higher expectation of education attainment. This has been categorised here as a protective factor, therefore, a higher score at follow-up indicated improvement. References Bond, L., Thomas, L., Toumbourou, J., Patton, G., & Catalano, R. (2000). Improving the Lives of Young Victorians in Our Community: a survey of risk and protective factors. Melbourne: Centre for Adolescent Health.

356

Appendix N: Bereavement Group Survey

357

358

359

360

361

362

363

364

365

366

367

368

Appendix O: Bereavement Group Survey Results Table O.1 Protective factors: Mean Change Scores Mean change scores for protective factors Scale

Domain

n

df

p

M

SD

95% CI

Opportunities for prosocial involvement

School

5

4

0.86

0.04

0.48

-0.55, 0.63

Rewards for prosocial involvement

School

5

4

0.72

-0.10

0.58

-0.81, 0.61

Educational expectation

School

5

4

0.23

-1.60

2.51

-4.72, 1.52

Emotional control

Emotional control

5

4

0.65

-0.15

0.68

-0.99, 0.69

MSPSS - significant other

Social support

5

4

0.24

-1.60

2.62

-4.85, 1.65

Trusted adult in life

Social support

5

4

0.33

-1.60

3.21

-5.58, 2.38

Mental Health Inventory - psychological wellbeing

Psychological wellbeing

4

3

0.22

9.50

12.40

-10.23, 29.23

Basic psychological needs - short form (whole score) Psychological needs

4

3

0.60

-5.25

17.69

-33.40, 22.90

Basic psychological needs - short form (autonomy)

Psychological needs

5

4

0.54

-2.00

6.63

-10.24, 6.24

Basic psychological needs - short form (competence) Psychological needs

4

3

0.51

-2.25

5.97

-11.74, 7.24

Basic psychological needs - short form (relatedness)

4

3

0.62

-1.25

4.57

-8.53, 6.03

Psychological needs

Note. CI = confidence interval. Mean change scores calculated using paired sample t-tests with baseline scores subtracted from follow-up scores. Green mean changes scores indicate positive changes and red mean change scores indicate negative changes.

Table O.2 Risk factors: Mean Change Scores Mean change scores for risk factors Scale

Domain

n

df

p

M

SD

95% CI

Academic Failure

School

5

4

0.57

-0.20

0.73

-1.10, 0.70

Low commitment to school

School

5

4

0.07

0.27

0.25

-0.04, 0.57

K10

Psychological distress

4

3

0.68

-1.50

6.61

-12.01, 9.01

Note. CI = confidence interval. Mean change scores calculated using paired sample t-tests with baseline scores subtracted from follow-up scores. Green mean changes scores indicate positive changes and red mean change scores indicate negative changes.

369