encontrar pontos comuns de integração entre as várias abordagens .... mudança, ou apenas perante um mecanismo com sete, cinco ou três níveis de ...... specific problem, under which set of circumstances'' (Paul, 1967, p. 111). ..... Stage one of this process defines ...... 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20.
UNIVERSIDADE DE LISBOA FACULDADE DE PSICOLOGIA
IF YOU WANT TO GET AHEAD, GET A SEQUENCE OF GENERAL STRATEGIES: INTEGRATIVE DECISION MAKING IN REAL-WORLD PSYCHOTHERAPY.
Nuno Miguel Silva Conceição Dissertação Orientada pelo Professor Doutor António José dos Santos Branco Vasco
DOUTORAMENTO EM PSICOLOGIA Psicologia Clínica
2010
UNIVERSIDADE DE LISBOA FACULDADE DE PSICOLOGIA
IF YOU WANT TO GET AHEAD, GET A SEQUENCE OF GENERAL STRATEGIES: INTEGRATIVE DECISION MAKING IN REAL-WORLD PSYCHOTHERAPY.
Nuno Miguel Silva Conceição Dissertação Orientada pelo Professor Doutor António José dos Santos Branco Vasco
DOUTORAMENTO EM PSICOLOGIA Psicologia Clínica 2010
A Doctoral Grant (Ref. SFRH/BD/25135/2005) of the Science and Technology Foundation (FCT), Portugal, and the Program POCI2010, which is funded by the Portuguese Ministry of Science, Technology and Higher Education, and the European Social Fund (Community Support Framework III), sponsored the present work.
ii
ACKNOWLEDGEMENTS
I am extremely thankful to António Branco Vasco, for the influence and orientation Leonel Garcia Marques, for statistics and high definition music Ken Critchfield for sharing experimental experience Jeffrey Borckardt for time-series analysis Ana Sousa, Andreia Tavares, Joana Ferreira, Joaquim Vitor, Ana Chaganlal e Hugo Sousa for rating the sessions and the first three of them for some paper work too Belina Duarte and Nuno Ramos for their plentiful friendship and generous editing and reviewing support Ana Paula Camacho for professionalism in solving administrative problems Rui Costa and Sara Haga, for sharing important phd-related information in important moments All participants in the study, for giving away their time My patients, for letting me learn and grow with them My parents, sister and all the children in the family, for structure and for play Beatriz’ parents for house of the sun Apple Computers, for seamless flow in bidimensional life
and, last, but not least, Beatriz, for the love sublime
iii
In Poems of the West and East by Goethe (1998):
Jedes Leben sei zu führen, Wenn man sich nicht selbst vermißt; Alles könne man verlieren, Wenn man bliebe, was man ist.
Every life is worth the choosing If oneself one does not miss; Everything is there for losing If one stays just as one is.
iv
ABSTRACT
The purpose of this dissertation was to analyse equivalent therapist's strategyrelated operations and patient's strategy-related change processes as they unfold and relate along the psychotherapeutic process as practiced in real world settings. Ideally, in order to optimize therapeutic responsiveness and effectiveness, therapist´s operations and patient´s processing abilities should proceed in tandem. Processing capacity was conceptualized as the productive use, by the patient, of strategy-related change processes or general psychological processes underlying skilful means to resolve life challenges. There were two research phases, both set within the framework of Paradigmatic Complementarity. The cross-sectional study used self-report data from therapists in real world practice, while the longitudinal study used observed data from 129 videotaped sessions of four patients of a single therapist. The findings from both studies suggest it is relevant to consider the sequencing of general strategies as a phase-by-phase map of the therapeutic process. First, strategyrelated therapist's operations and patient's change processes can be operationalized in a equivalent format and measured with acceptable reliability, for psychotherapy as practiced in the real-world. Second, the time patients need to show high levels of strategy-related change processes suggests general processing capacities take months to build up, rendering it less a moment-by-moment process. Third, strategy-related patient change seems to unfold sequentially according to a sequence of 3, 5 or 7 dimensions, thus providing sequential maps of intermediate outcomes before hitting termination. Fourth, these intermediate outcomes, predict changes in final outcome, making unique contributions over and above the therapeutic alliance, rendering them good enough
v
candidates to stand as mechanisms of change common to any approach. Fifth and finally, decision-making by the therapist can be responsive to this macro developmental sequence influencing its unfolding and being influenced by it. Limitations
and
implications
for
real-world
essentially
integrative
psychotherapy theory, research, practice and training are then discussed.
Keywords:
Decision-Making;
Integrative
Strategies; Change Mechanisms
vi
Psychotherapy;
Sequencing;
General
RESUMO
O presente projecto de investigação procurou analisar estratégias gerais em termos de operações terapêuticas, por parte do terapeuta, e em termos de processos de mudança, por parte do paciente, à medida que vão evoluindo e se vão relacionando entre si, ao longo do processo psicoterapêutico tal como é praticado em contexto naturalístico. Quando se fala de estratégias gerais de uma sessão, ou grupo de sessões específicas, fala-se daquilo que o terapeuta está a tentar promover (na sessão ou grupo delas), como já foi feito nas investigações anteriores, mas também daquilo que está a ser assimilado pelo paciente. Está-se, portanto, perante o processo de tomada de decisão do psicoterapeuta, a tentar estudá-lo e a tentar apoiá-lo. Tal como sugerido por Goldfried (1980) é neste nível intermédio de abstracção de estratégias gerais, entre a teoria e a técnica, que é mais provável que possamos encontrar pontos comuns de integração entre as várias abordagens terapêuticas. E, de facto, um cada vez maior número de terapeutas se define como não sendo de escola única (Norcross, 2005). Para além disso, a psicoterapia, tal como praticada em contexto naturalístico, pouco tem a ver com muita da investigação que se tem feito, mais assente em regras do que em princípios, cenário que só mais recentemente começou a alterar-se no sentido de aproximar investigação e prática clínica. Idealmente, e no sentido de maximizar a responsividade terapêutica, as operações do terapeuta e a capacidade de processamento do paciente deveriam entrelaçar-se mutuamente de forma a promover um desenvolvimento e uma aprendizagem, ou mudança, sustentados. Capacidade de processamento aqui é definida como a utilização produtiva que o paciente faz de processos de mudança relacionados
vii
com estratégias gerais quando estes subjazem aos meios hábeis de resolver desafios na vida quotidiana. Na literatura existem vários modelos de fases do processo psicoterapêutico, como que a indiciar uma responsividade em termos de fases da terapia, aquilo a que poderíamos chamar de responsividade fase-a-fase (e.g. Beitman, 1987; Howard, Lueger, Maling, & Lutz, 1993; Benjamin, 1993, 2003; Vasco, 200; Goldfried, 2006). Mas, de facto, nem sempre é muito explícito que o terapeuta procure tomar decisões no sentido de promover predominantemente objectivos estratégicos característicos de uma determinada fase, em detrimento de outros, no sentido de respeitar uma qualquer sequência de fases do processo. Com isto quero dizer que o processo de tomada de decisão nem sempre toma a grande escala temporal em consideração. Se assumirmos que podem existir grandes dimensões de trabalho terapêutico que devem seguir-se a outras grandes dimensões de trabalho terapêutico, independentemente dos conteúdos e temas específicos de cada caso, em vez de indiciar poderemos incitar à explicitação de um mapa sequencial de responsividade fase-a-fase. A questão é até que ponto estas fases poderão ser traduzidas em estratégias gerais a promover (ou a não promover) e em capacidades (ou vulnerabilidades) do paciente, em termos de processamento de informação, no sentido integrativo do termo. Este projecto enquadra-se no contexto do meta-modelo de Complementaridade Paradigmática. Este contempla, até ao momento, quatro componentes: “(a) “princípios gerais de mudança terapêutica” (o “quê” da terapia); (b) a aliança terapêutica (o “estar” em terapia); (c) “conceptualização do paciente e do problema” (o “quê” mais o “como” da terapia); (d) “sequência temporal de fases relativas a objectivos” (o “quando” da terapia).” (Vasco, 2006, p. 15).
viii
É neste último componente que a presente investigação se desenvolve, apesar de manter estreitas relações com o componente dos princípios gerais de mudança terapêutica (Castonguay & Beutler, 2006). Este componente consiste em sete fases sequenciais de mudança terapêutica: (1) Confiança, motivação e estruturação da relação; (2) Aumento da consciência da experiência e do Self; (3) Construção de novos significados relativos à experiência e ao Self; (4) Regulação da responsabilidade; (5) Implementação de acções reparadoras; (6)
Consolidação da mudança; e (7)
Antecipação do futuro e prevenção da recaída. Partiu-se desta proposta para estudar o fenómeno de desenvolvimento de capacidades de processamento e tomada de decisão clínica responsiva. De acordo com a revisão de literatura, parece existir muita evidência a favor da eficácia equivalente entre as várias abordagens terapêuticas. Existe substancial evidência a favor de princípios de mudança comuns a várias perturbações psicológicas. Da mesma forma, existe considerável evidência a favor de processos psicológicos comuns a várias perturbações. Existe bastante evidência de que aquilo que se mede constrange aquilo que se observa. Existe alguma evidência a favor de mecanismos de mudança comuns a várias abordagens. Mas, existe escassa evidência a favor de mapas sequenciais de mecanismos de mudança. Existiram duas fases de investigação. A fase transversal fez uso de dados provenientes de questionários de auto-relato de terapeutas em contexto naturalístico, enquanto a fase longitudinal fez uso de dados provenientes da observação e cotação, por juízes treinados para o efeito, de 129 sessões, de um único terapeuta, gravadas em vídeo. Os resultados de ambos os estudos sugerem que é suficientemente relevante e rigoroso considerar a sequencialização de estratégias gerais enquanto mapa fase-a-fase do processo terapêutico.
ix
Primeiro, as operações terapêuticas em termos de implementação de estratégias gerais, assim como os processos de mudança que se esperam resultantes da assimilação dessa implementação, podem ser operacionalizados de forma equivalente e medidos com precisão e validade razoáveis. Segundo, o tempo que os pacientes precisam para demonstrar elevados níveis de processos de mudança relacionados com estratégias gerais de processamento é da ordem de meses, tornando difícil que se esteja perante um fenómeno de momento-a-momento. Terceiro, a mudança do paciente relacionada com estas estratégias gerais parece desenrolar-se de forma sequencial, de acordo com uma sequência de 3, 5 ou 7 dimensões, fornecendo assim mapas sequenciais de resultados intermédios (antes de se atingirem os resultados finais) empiricamente apoiados. Quarto, estes resultados intermédios prevêem mudanças nos resultados finais, tendo contribuições únicas para além da aliança terapêutica, os que os torna bons candidatos a mecanismos de mudança comuns a várias abordagens. Quinto, e finalmente, a tomada de decisão dos terapeutas pode claramente ser responsiva a esta macro sequência desenvolvimentista, influenciando o seu desenrolar e sendo influenciada por ela. Em geral, conseguiu-se: a) reforçar a evidência já acumulada em torno do componente sequencial 7 fases do modelo de Complementaridade Paradigmática, bem como a sua articulação com outros componentes do modelo, com dados transversais e longitudinais, de auto-registo e de vídeo; b) relacionar o fenómeno fase-a-fase deste modelo de 7 fases com outro modelos de fases e com a literatura de mecanismos de mudança; c) articular o fenómeno de responsividade fase-a-fase em termos de processo interno de tomada de decisão clínica, em ligação directa com as capacidades e vulnerabilidades dos pacientes, no que ao processamento da informação diz respeito.
x
Em termos de implicações, tal proposta baseada em estratégias gerais e princípios gerais de mudança pode favorecer o treino de terapeutas em formação, pode promover uma maior ligação entre a prática e a investigação, particularmente em contextos naturalísticos, assumindo que a linguagem é suficientemente transteórica e compreensível, não só por terapeutas de qualquer escola, como também pelos seus próprios pacientes. Pode também contribuir para uma especificação de mecanismos de mudança comuns a várias abordagens. E, como não existem na literatura testes empíricos de sequenciação de vários mecanismos de mudança, este projecto pode estar a contribuir para que se venha a abrir essa frente de investigação. Se estamos perante sete, cinco ou três mecanismos de mudança, ou apenas perante um mecanismo com sete, cinco ou três níveis de capacidade de processamento mantem-se em aberto a questão. Ainda assim, pegar num mapa sequencial para tomar decisões molares quando, ao nível molecular, se está em impasse, pode ser uma forma de seguir em frente.
Palavras-chave: Tomada de Decisão; Psicoterapia Integrativa; Sequenciação; Estratégias Gerais; Mecanismos de Mudança
xi
ÍNDICE
ACKNOWLEDGEMENTS ..................................................................................................... iv ABSTRACT ......................................................................................................................... v RESUMO .......................................................................................................................... vii ÍNDICE ............................................................................................................................. xii LIST OF TABLES .............................................................................................................. xiv List of Figures................................................................................................................ xvi INTRODUCTION .................................................................................................................. 1 Literature Review ........................................................................................................... 11 Equivalency Across Approaches ................................................................................ 11 Common Factors Across Approaches ........................................................................ 12 Temporal Patterns of Principles of Change Across Approaches................................ 14 Principles of Change Across Disorders ...................................................................... 20 Where Did All The Specificity Go? ........................................................................... 22 The Renewed Interest in Transdiagnostic Approaches .............................................. 24 Evidence-Based Practice and Practice-Based Evidence............................................. 27 The Black Box of Specificity: Linking Patient Change Process with Outcome ........ 30 Outcomes: What is Measured and What is Left to Measure ...................................... 35 Outcomes Across Approaches.................................................................................... 40 Expansion of Research Designs and Instrumentation ................................................ 42 Qualitative approaches and single subject designs................................................. 43 Qualitative approaches. ...................................................................................... 43 Single subject designs......................................................................................... 48 Measuring interventions and change processes in real-world therapy.. ................. 50 Comparative psychotherapy process scale (CPPS). ........................................... 51 Therapeutic Focus on Action and Insight (TFAI).. ............................................ 52 Comprehensive Psychotherapeutic Interventions Rating Scale (CPIRS)........... 52 Multitheoretical List of Therapeutic Interventions (MULTI). ........................... 53 The Therapeutic Realizations Scale - Revised (TRS-R).. .................................. 54 Achievements of Therapeutic Objectives Scale (ATOS). .................................. 54 Client Task Specific Change Measure Revised (CTSC-R). ............................... 55 The Quest for Mechanisms of Change ....................................................................... 55 Mechanisms of change of EST’s packages - approach and disorder specific? ...... 57 Dialectical behavior therapy (DBT). .................................................................. 57 Transference focused psychotherapy (TFP). ...................................................... 59 Mentalization based therapy (MBT)................................................................... 61 Schema focused therapy (SFT)........................................................................... 62 Cognitive therapy (CT)....................................................................................... 64 Interpersonal psychotherapy (IPT). .................................................................... 65 Empirically found common mechanisms of change across approaches................. 68 Purpose of this Research Project ................................................................................ 71 STUDY 1: CROSS SECTIONAL EXPLORATION AND REPLICATION PHASE .......................... 81 Method............................................................................................................................ 82 Participants ................................................................................................................. 82 Instruments ................................................................................................................. 84 Development of the General Strategies Inventory (GSI). ...................................... 84 xii
Brief Symptom Inventory (BSI).. ........................................................................... 88 Estimate of Improvement Scale (EIS-PC).............................................................. 88 Working Alliance Inventory - Short Form (WAI-S). ............................................. 89 Session Outcome scale (SOS). ............................................................................... 89 Procedure .................................................................................................................... 90 Results ............................................................................................................................ 91 Descritive Statistics & Internal Consistency Analysis ............................................... 91 Multidimensional Scaling......................................................................................... 100 Factor Analysis of the GSI ....................................................................................... 105 Therapist’s strategy-related operations................................................................. 106 Patient’s strategy-related change processes.. ........................................................ 113 Relationships Between General Strategies and With Other Process/Outcome Variables................................................................................................................... 119 General Strategies’ Movements Over Time ............................................................. 122 Intermediate Outcomes (Strategy-Related Change Processes) and Final Outcome. 128 Discussion..................................................................................................................... 131 STUDY 2: LONGITUDINAL EXPLORATION PHASE ........................................................... 149 Method.......................................................................................................................... 150 Participants ............................................................................................................... 150 Instruments ............................................................................................................... 152 General strategies inventory (GSI-Observer Perspective).................................... 152 Procedure .................................................................................................................. 152 Rater training. ....................................................................................................... 152 Data coding procedure.......................................................................................... 153 Data Analysis............................................................................................................ 155 Results & Discussion.................................................................................................... 157 Reliability ................................................................................................................. 157 Results Of The Four Cases Along The 20 Observation Points ................................ 159 Patient's strategy-related change processes. ......................................................... 159 Therapist's strategy-related operations.. ............................................................... 162 Relationship between therapist's strategy-related operations and patient's strategy-related change processes……………………………………………166 Patient's strategy-related change processes as intermediate outcomes................ 175 Results Of Case W Along The 69 Observation Points............................................. 177 Patient's strategy-related change processes. ......................................................... 179 Therapist's strategy-related operations. ................................................................ 180 Relationship between therapist's strategy-related operations and patient's strategy-related change processes.................................................................. 183 Patient's strategy-related change processes as intermediate outcomes................. 187 Sequence of Mechanisms of Change.................................................................... 188 Conclusions .................................................................................................................. 191 GENERAL DISCUSSION ................................................................................................... 195 REFERENCES .................................................................................................................. 211
xiii
LIST OF TABLES
Table 1 - Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 1 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) _________________________ 92 Table 2 - Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 2 Of GSI-Top/T (N=227) And GSI-Pcp/T (N=219) ________________________ 93 Table 3 - Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 3 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) _________________________ 94 Table 4 - Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 4 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) _________________________ 95 Table 5 - Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 5 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219)._________________________ 96 Table 6 - Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 6 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) _________________________ 97 Table 7 - Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 7 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) _________________________ 98 Table 8 - Internal Consistency For The Subscales Of The GSI-Top/t and GSI-Pcp/t 100 Table 9 - Multidimensional Scaling With 28 Items Of The GSI-Top/t ___________ 102 Table 10 - Multidimensional Scaling With 28 Items Of The GSI-Pcp/t___________ 104 Table 11 - Factor Loadings Of The 28 Items Of The GSI-Top/t ________________ 108 Table 12 - Factor Loadings Of The 15 Items Of The GSI-Top/t ________________ 112 Table 13 - Factor Loadings Of The 28 Items Of The GSI-Pcp/t ________________ 114
xiv
Table 14 - Factor Loadings Of The 15 Selected Items Of The GSI-Pcp/T_________ 117 Table 15 - Maximum, Minimum, Means And Standard Deviations Of The Empirically Derivated Scales Of Strategies ___________________________________ 118 Table 16 - Correlations Between Empirically Derived Subscales Of Therapist’s Strategy Related Operations (Upper Right) And Between Subscales Of Patient’s Strategy Related Change Processes (Lower Left) ____________________ 120 Table 17 - Correlations Between Subscales Of Therapist Change Processes (Upper Right) And Between Subscales Of Patient Change Processes (Lower Left) Of The General Strategies Inventory – Short Form/Therapsit Perspective ____ 121 Table 18 - Correlations Between Subscales Of General Strategies And Other Process And Outcome Variables ________________________________________ 122 Table 19 - Means Of The Empirically Derivated Scales Of General Strategies Over Time By Group Of Session Number_______________________________ 123 Table 20 - Means Of The Empirically Derivated Scales Of Patient’s Strategy Related Change Processes By Group Of Predominant Therapist’s Strategy-Related Operations___________________________________________________ 125 Table 21 - Hierarchical Regression of Overall Mean Scores and Final Outcome (n =90) ___________________________________________________________ 130 Table 22 - Analyzed Psychotherapeutic Processes: Patient Information __________ 151 Table 23 - Teams of Raters by Cases by Session Dates _______________________ 156 Table 24 - Case W by Observed Sessions’ Dates ___________________________ 158 Table 25 - Values Of Patient's Strategy-Related Change Processes (Dimensions 4 And 5) Across Time _______________________________________________ 176 Table 26 Case W - Pattern of Change: Cross-correlation functioning showing directional and temporal relationship of change in pCap4 with change in pCap5.______________________________________________________ 190
xv
List of Figures
Figure 1: Phases’ Sequential Model of Paradigmatic Complementarity....................... 19 Figure 2: Scree plot for factor analysis of the 28 Items from GSI-Top/t .................... 107 Figure 3: Scree plot for factor analysis of the 15 selected items from GSI-Top/t....... 111 Figure 4: Scree plot for factor analysis of the 28 items from GSI-Pcp/t. .................... 113 Figure 5: Scree plot for factor analysis of the 15 selected items from GSI-Pcp/t. ...... 115 Figure 6: Profile of dimensions of general strategies as operations and as change processes......................................................................................................... 119 Figure 7: Profiles of patient’s strategy-related change processes by time groups....... 127 Figure 8: Evolution of patients’ strategy-related change processes/capacities (CAP) along the process. ........................................................................................... 161 Figure 9: Distribution of therapist's strategy-related operations in the four cases along the process. ..................................................................................................... 163 Figure 10: Predominant therapist's strategy-related operations in the four cases along the process. ..................................................................................................... 165 Figure 11: Responsiveness of therapist's strategy-related operations (tOp) to patient's strategy-related change processes (pCap) in case Z along the process. ......... 167 Figure 12: Responsiveness of therapist's strategy-related operations (tOp) to patient's strategy-related change processes (pCap) in case Y along the process.......... 169 Figure 13: Responsiveness of therapist's strategy-related operations (tOp) to patient's strategy-related change processes (pCap) in case W along the process......... 171 Figure 14: Responsiveness of therapist's strategy-related operations (tOp) to patient's strategy-related change processes (pCap) in case X along the process.......... 173 Figure 15: Therapist's strategy-related operations (tOp) and patient's strategy-related change processes (pCap) in case W along the 69 sessions............................. 179 Figure 16: Responsiveness of therapist's strategy-related operations (tOp) to patient's strategy-related change processes (pCap) in case w along 69 sessions ......... 182 xvi
Figure 17: Level of therapist's strategy-related operations before and after patient's strategy-related change processes reach structural change status. ................. 185 Figure 18: Level of therapist's strategy-related operations eight sessions before and after patient's strategy-related change processes reach structural change criterion 187 Figure 19: Level of patient's strategy-related change processes over 69 sessions. ...... 189! !
xvii
INTRODUCTION
From the beginning of my psychotherapeutic training, I was confronted with the mysteries of psychological processes. In particular, I have been always interested in understanding decision-making processes, and during these last ten years of my life I have spent a considerable amount of time time dwelling on the decision-making process of psychotherapists, being a patient, a psychotherapist and a researcher myself. I was compelled to make sense out of what I was doing with people and I became curious to experience what my therapist was doing with me, as well as what I was doing with myself. What a trip! I can say with some conviction, that research on decision-making in psychotherapy is in its infancy and, despite a seminal work by Caspar (1997), only more recently are the mysteries of what goes on in a psychotherapist’s mind beginning to see the light. The recent bonus voiceover, in which the psychotherapist comments on the therapy as it progresses, made available in the Psychotherapy Video Series of the American Psychological Association, that present distinguished psychologists demonstrating specific approaches to a wide range of patient problems, is but an example of the increasing transparency and clarity the subject is receiving. Over the years, there have been theoretical and research developments on several models that have potential value to help clinicians make clinical decisions based on patient characteristics and behaviors for achieving a productive level of responsiveness towards patients’ needs (for reviews see Conceição, 2005, Conceição & Vasco 2005; Vasco, 2006).
Responsiveness refers to behavior that is affected by
emerging context, including others’ behavior. Therapists and patients are responsive to each other on time scales that range from months to milliseconds (Stiles, Honos-Webb, & Surko, 1998). 1
This wealth of models and theories describe the complex processes within and between participants as they engage in psychotherapy, guiding psychotherapists in their practice. What might be scaring for some, is the large number of variables described in the literature that could be involved in integrative clinical decision making aiming at maximising effectiveness and responsiveness: it probably allows more than one million possible interactions. For me, this makes research in this field quite challenging and relevant, with regard to a need of simplification. Indeed, ongoing responsive decisionmaking is an issue with which psychotherapists in real world settings deal every day, making research in this field always welcomed as a way to provide for an integration between research and practice. Would practice-based evidence shed light on the same issue, it would naturally also be welcomed (Barkham, Hardy & Mellor-Clark, 2010). Therapists’ competence involves systematically responding to emerging markers or needs of their patients to facilitate positive outcomes, at least in ways consistent with their approach. The seamless integration of (therapeutic) skills, and the ability to be both alone and intimately related in the moment with the person whom she or he is working with, while at times uncomfortable or painful, can produce moments of enlightenment and pleasure bearing resemblance with the ones, one experiences with art (Gross, in press). Used to guide therapeutic action, theory is the psychotherapeutic map or the representation of clinical reality of how therapy unfolds—what to attend to, what, when and how to do, and how to assess progress. Psychotherapy theories are explanations of human functioning, dysfunction and the process of change. Contrary to the situation where one enters a shop to buy a map to explore Iberia, for example, in psychotherapy there is, however, no one "best" road map; rather, there are a plethora of viable maps from which to choose (Wampold, 2010). Consequently, increasingly more clinicians
2
have come to recognize the limitations of their (once) preferred maps and report being influenced by multiple theoretical orientations (Orlinsky & Rønnestad, 2005), especially those in real world practice. Two main directions for research emerging from the psychotherapy integration literature pertaining to decision making, mentioned by Schottenbauer, Glass, and Arnkoff (2007) were a) the need to use methods that are geared to discovering what is unfolding in therapy, rather than studying one-time clinical decisions, and b) the high value of developing an empirically derived common theory of decisional processes for psychotherapists. Entertaining the idea of making an empirical contribution for psychotherapy in research in general and for psychotherapy as practiced in real world (most probably of an integrative nature) in particular, I remembered that just before the onset of the psychotherapy integration movement, Goldfried (1980) offered a new map or stratification of psychotherapy from which different approaches could be compared for the sake of delineating the most meaningful points of convergence.
These
commonalities tend(ed) to be obscured by the widely diverging terminologies employed by therapists of each single orientation. Goldfried's (1980) article was a plea for less parochialism among adherents of different theoretical orientations and more rapprochement, with the goal of enhancing psychotherapy integration. He suggested that the best way one is likely to identify therapeutic similarities of the most clinical relevance is by identifying clinical strategies, which are less abstract than theory but more general than techniques or procedures. These strategies arise from rational and thoughtful intervention plans and can become regarded as “principles” of change if they are supported by empirical evidence (Castonguay & Beutler, 2006).
3
Rather than attempting to valorize certain specific procedures wrapped in a particular package (Westen, Novotny & Thompson-Brenner, 2004), I like to believe that, hopefully, we are about to face a change in the way that psychotherapy and research evidence is defined, shifting us from relying on a narrow range of methods, models and patients to the investigation of one or more research-informed principles of effective therapy (Castonguay & Beutler, 2006; Levitt, Neimeyer, & Williams, 2005). And in thinking about the evidence base for a principle-based approach to psychotherapy, it is worth remembering that despite the differences that guide therapeutic work from different theoretical orientations, there is far more overlap than is commonly assumed, an overlap begetting sound differentiation in terms of the psychological or strategy-related change processes involved. The enormous progress in psychotherapy research has culminated in recognition of several treatments that have strong evidence in their behalf. The Dodo bird points us toward looking at ways, whether in the form of factors, strategies, or principles, that psychotherapists are more similar than different. However, data collected from practiceresearch networks, that is, practice-based evidence (Barkham, Hardy & Mellor-Clark, 2010) point in the direction that some therapists are considerably better than others promoting good outcome and eventually good process in their patients. The question still remains, as to what makes psychotherapy work, what underlies good process. We cannot yet provide sound evidence-based explanations for how or why even the most well studied interventions produce change, that is, the mechanism(s) through which treatments operate. Perhaps the reason for this was that the question of ‘‘how?’ was notably missing from the guiding question for psychotherapy research since the late 1960s, originally introduced by Kiesler (1966),
4
has been ‘‘what treatment, by whom, is most effective for this individual with that specific problem, under which set of circumstances’’ (Paul, 1967, p. 111). Despite this miss, there have been a few rigorous researchers who were aware of the wisdom in this position (e.g. Greenberg & Pinsof, 1986; Safran, Greenberg & Rice, 1988; Garfield, 1990; Greenberg, 1991; Greenberg & Foerster, 1996), urging for the study of the link between patient change process and outcome, and for the effects of particular processes to be demonstrated. By these effects, they were emphasizing the link between therapist interventions and patients change processes under a metaphor of absorption or assimilation by the patient of the therapeutic act. Indeed, considerable rigorous and relevant research on change processes followed these calls, especially at the moment-by-moment level of intervention/development (e.g. Elliot, Watson, Goldman & Greenberg, 2004; Safran, Muran, Samstag & Stevens, 2002; Stiles, 2002). From a comprehensive perspective, the generic model of psychotherapy (Orlinsky & Howard, 1986) since its inception, also provided an empirically valid conceptual representation of psychotherapy process and outcome relationships, dwelling with the question of what is effectively therapeutic about psychotherapy. It aimed at providing a theoretical foundation from which research investigating complex process-outcome relationships common to all psychotherapies could proceed. This model offers a conceptualization of outcome as a process that builds over time from the in-session process events of therapeutic realizations, to shorter-term micro-outcomes through intermediate meso-outcomes to longer-term macro-outcomes. In a zeitgeist of accountability, the time has already come for clinicians to routinely monitor treatment outcome, at least in the United States of America and in some European countries like the United Kingdom (e.g. Lambert, 2010). But what about monitoring the mediators of those outcomes, particularly the ones pertaining to
5
the psychological architecture of the patient, that is, his or her processing capacity with regard to strategy-related change processes to succefully promote those monitored outcomes? Recently, the call to study mediators and mechanisms of therapy received widespread emphasis. In a humorist twist, reporting that discussions and theory about why psychotherapy changes people are plentiful, but supportive evidence quite rare, Kazdin (2007) equated mechanism with the weather of psychotherapy research and cited MarkTwain (1835–1910) noting that, “everybody talks about the weather but nobody does anything about it.” Also concerned with this state of affairs, Doss (2004) stressed that, despite a recent surge of interest in the mechanisms and processes of change during psychotherapy, investigations to date have yielded lamentably few interpretable results. To increase the interpretability of future studies he then presented a multiphase model of programmatic change research, where he argued that investigators should first develop an understanding of change mechanisms, and only subsequently conduct targeted process research to identify important change processes. Indeed, if we go back to the insight Goldfried (1980) had in San Francisco, he not only allowed clinicians to increase their clinical repertoire as well as their creative potential but, I believe, also already provided researchers with a clinically relevant and conceptually meaningful focus for the investigation of mechanisms of change in psychotherapy. Unfortunately, however, studying mechanisms of change is no easy task, and most empirical studies focus on one or two mechanisms not organized according to any map or sequence. Detailed exploration of change mechanisms may provide data relevant to another question of interest in current psychotherapy research: the psychotherapy doseeffect relationship. What kind of changes is achievable in which amount of time and
6
how to guide that unfolding has always interested me. Given evidence that long-term therapies may be more effective than short-term therapies (Orlinsky, Rønnestad, Willutzki, 2004), it seems worthwhile to develop ways to study the relationship between therapeutic operations and patient change mechanisms leading to (final) outcomes in short and long therapies to analyse the similarities and difference in level or type of change. Despite the influence of time-limited psychotherapy and a move, in our society towards fast food as well as shorter therapy services, data is still inconclusive with regard to the amount of therapy necessary to attain desired outcomes (Hansen & Lambert, 2003). For that aim, a more comprehensive conceptualization of outcomes, final and intermediate (mechanisms) is needed.
A map or representation of their
organization, sequential or other, could also be useful. Are there outcomes that have to come out first before others can see the light? Taken together, what is achievable and how processes unfold in a psychotherapeutic process, yielding good process, interests me from an integrative perspective as applied in real world settings. When I sit in front of a patient, clarifying the map, on a phase-to-phase level, in which development occurs, is a challenge to me, not only as a psychotherapist but also a researcher, both as a research-oriented psychotherapist, or as a practice-oriented researcher. My work in psychotherapy, as with many other colleagues, involves both overcoming the pain of the past, facilitating the development of skills to enjoy an adult cheerful life, while entertaining the dream of the future. How do patients become more skilfull in the art of conquering life led me to come back to my early (then abandoned) training in cognitive neuropsychology and make a reference to the term processing and mental mechanisms (Bechtel, 2009). This is not so much an answer as it is a label for a research problem.
7
My challenge in this dissertation is to contribute to the quest of defining and explaining larger chunks of change, described in terms of mental mechanisms responsible for information processing of certain psychological processes or strategies, letting go of any clinging to content. Inspired by some expert colleagues on momentby-moment level of change (e.g. Greenberg & Safran; 1987; Pascual Leone, 2009), and after a previous first shot (Conceicao, 2005), I now more systematically attempt the so called phase-to-phase level approach to the general process of change in psychotherapeutic processes, borrowing from Vasco (1999; 2001) the idea of the temporal sequencing of strategic objectives of the therapist. Effectively dealing with daily life’s tasks or struggles implies capacities to process increasingly complex (psychological) information. The path from vulnerability into capacity or skillful means can probably be accomplished by meeting certain ongoing performance criteria, like in skills improvement, i.e. mindfulness practice, learning a new language, learning to be a psychotherapist, learning to play an instrument, learning to cook - the learning processes never end. Will there be invariance in terms of our pattern of performance and in terms of the development of our processing capacity or skillfull means? The phenomenon of how patients develop increasing capacities, when undergoing psychotherapy, is at the heart of the current research project. In this process, the patient has to produce performance that meets certain standards, like a given level of awareness, regulation of agency, or regulation of needs. Indeed, an adaptation of the logic underlying the research paradigm of process analysis applies here. Process analysis borrows from task analysis (Greenberg, 2007; Pascual-Leone, Greenberg, & Pascual-Leone, 2009) but is the study of how a participant develops and makes use of new sets of skills that start to be applied across situations (Pascual-Leone & Greenberg, 2007).
8
To promote optimal development the therapist has probably to be careful to balance the intensity or complexity of the processing required of the patient. I believe, processing strategies or general principles vary in demand characteristics, and a therapist’s concern can be to reduce the likelihood of overload constraining or even hindering the flow of processing and development. If one is able to identify specific mental mechanisms of change, that stem from or constrain targeted developmental learning experiences or therapist interventions (Basseches & Mascolo, 2009), one might best facilitate change if one decides to work in the patient’s zone of proximal development (Vygotsky, 1924/1978; Leiman & Stiles, 2001). This choice may help therapists to be maximally effective in facilitating processing in general. Note that, despite borrowing much from my expert colleagues on moment-bymoment therapeutic action, I am emphasizing the phase-by-phase level of intervention. And in doing so, my memory takes me to the context where I met one of the greatest loves of my life. We were doing some coursework on developmental psychology, namely applying some Piagetian tasks to test the notion of object permanence. And there was that untroubled and tranquil child who, no matter what we would do, on a moment-by-moment level (yes, we were creative!), she would never sustain curiosity in the object(s) once removed from her visual field. Needless to say, to a psychologically informed audience, she was perfectly healthy, simply younger. Much has run in the river of Piagetian and Neo-Piagetian perspectives (McClelland & Robert Siegler, 2001; Morra, Gobbo, Marini & Sheese, 2008) since then, but the implications of this story for my psychotherapy practice, and my interest in the phase-to-phase level of responsiveness, in particular, remain alive. In a similar fashion as to what Dodo Bird says, it seems to me that there is evidence in the literature that point us toward looking at ways, whether in the form of
9
change (psychological) processes, or change mental mechanisms, in which patients are more similar than different. So I am interested in differentiating these similarities within and between participants of the psychotherapeutic encounter, as well as to make bridges between them. To accomplish these goals I made use of both cross-sectional and longitudinal methods hoping to provide more comprehensive, robust findings; qualitative research will follow some day after this research project is finished. The literature on similarities in therapist's strategy-related operations and patient's strategy-related change processes and mechanisms will be reviewed next, paradoxically to sustain the project of a better differentiation between them.
10
Literature Review
Equivalency Across Approaches A relatively young profession, psychotherapy has been characterised by a rich array of theories, quite different procedurally, scientifically and philosophically. Psychotherapy research has shown that differing therapies often demonstrate effectiveness. Nearly two-thirds of all patients experience a positive outcome from psychotherapeutic treatment (Lambert & Barley, 2002; Asay & Lambert, 1999; Wampold, 2001; Luborsky & Barrett, 2006). What makes psychotherapy work is another question.
No procedural or intervention factors have been identified that
consistently capture a large portion of the variance in improvement, and despite the push for determining effective treatments, about half of the variance (Norcross & Lambert, 2006) remains somewhat of a mystery. In some ways, what makes psychotherapy work depends on whom you ask: (a) those who believe that particular treatments for various disorders are effective because of the specific ingredients of the treatments, or (b) those who believe that therapeutic ingredients common to all (or most) treatments are responsible for the benefits of psychotherapy (Wampold, 2010). Luckily, as in many branches of psychology, there is evidence for both sides. Reviews of the psychotherapy research literature have, for decades, concluded that all treatments are generally equivalent (Asay & Lambert, 1999; Lambert & Barley, 2002; Lambert & Ogles, 2004; Wampold, Ollendick, & King, 2006; Wampold, 2010b). Several alternative explanations have been suggested for this equivalence and all warrant further investigation before any commitment to a particular direction: (a) different therapies might operate through different processes but still achieve similar 11
outcomes; (b) different outcomes may have indeed occurred, but they were not detected given methodological limitations, and (c) common factors shared by all therapies are the healing element. Common Factors Across Approaches Taking into account the fact that psychotherapies are shown to be fundamentally equivalent in general effectiveness, Wampold (2001) goes on to defend that more research resources should be directed away from efforts to validate particular therapies and toward research on common factors, those aspects of treatment that are not specific to any school of treatment, such as empathy, working alliance, the extent to which patients are able to engage in treatment, the facilitation of hope, the opportunity for emotional release, support, advice, and encouragement to try out new explanations or perspectives and behaviors. The idea of common factors has a history of more than 60 years, beginning with Rosenzweig (1936), continuing with the influential work of Frank and Frank (1991): (a) an emotionally charged and confiding relationship between the healer and the patient, (b) a healer who is given special status and is perceived to have the powers to heal, (c) a powerful and cogent rationale for the healer's actions, and (d) a set of treatment actions that are consistent with the rationale. As reviewed by Conceicao (2005) there have been several proposals of common factors (e.g. Greencavage & Norcross, 1990; Tracey, Lichtenberg, Goodyear, Claiborn & Wampold, 2003). While much has been written about the subject since then, until recently, relatively few scholars have provided useful ways of identifying the important common factors, studying them, and utilizing them more effectively in psychotherapy. One of the reasons for this lack of progress probably was that: “what usually happens is that a proponent of a common factor or a set of common factors reviews the history of the factor, points to 12
its ubiquitous presence in all forms of therapy, cites the relevant research (if there is any) and moves on, confident that his or her case has been proved.” (Weinberger, 1995, p. 46). Hill (1995), commenting on Weinberger's paper, suggested the then too abstract, global, and vague common factors needed to be defined more concretely, so that they could be measured separately from other aspects of therapy process. She stressed the need to study mechanisms of change in the common factors, reminding us that the concept of equifinality suggests that even though different factors have different mechanisms of change, they may all produce the same outcome. In the last years some authors have been particularly active on the defense of common factors (Hubble, Duncan, & Miller, 1999; Weinberger, & Rasco, 2007; Duncan, Miller, Wampold, & Hubble, 2010; Wampold, 2010), but Hill’s (1995) call for the study of their mechanisms of change remains somewhat unheard and therapists run the risk of taking or implementing them in undifferentiated manners, because their use is rarely specifically illustrated (an approach more common to the specific factors proponents). One of the useful ways of utilising common factors more effectively in psychotherapy, I believe, is the sequencing of them into a map of how therapy unfolds, used to guide therapeutic action. This heuristic, in my view, has never received enough emphasis. A notable exception, for example, is Lampropoulos (2001) who assembled a comprehensive list of eight factors by (a) selecting the most frequently reported factors, (b) grouping together the overlapping, similar, and subsumed factors, and (c) arranging them in a prototypical, heuristic time sequence in psychotherapy and change. As a result, according to Lampropoulos (2001) the therapeutic process would be summarised as follows: (1) formation of a therapeutic relationship (bond, positive 13
personal skills and qualities), and establishment of a working alliance (contract, goals, tasks); (2) accomplishment of catharsis and relief from distress (i.e. ‘emotional regulation’ via empathy, and support); (3) instillation of hope and raising of expectations (to actively engage patient in therapy); (4) self-exploration, awareness, and insight into problems (feedback, reality testing); (5) provision of a theoretical explanation (rationale) for patients’ behaviour and change (via interpretation, restructuring, reframing); (6) problem confrontation (exposure, working through, use of techniques); (7) acquisition and testing of new learning in and outside psychotherapy (behavioural–cognitive–experiential–interpersonal learning, via suggestion, persuasion, identification, modelling, etc.); and (8) control over the problem and mastery of the new knowledge (self-attributions of change and self-efficacy enhancement; generalised use of the solutions; change maintenance and relapse prevention). Temporal Patterns of Principles of Change Across Approaches Common factors are sometimes used interchangeably with the construct of principles of change.
Despite a great deal of common variance between both
constructs, a closer look at the literature yields somewhat different scenarios. Goldfried (1980) suggested that the place to look for the explanation of the equivalence of effectiveness among different psychotherapies, the Dodo Bird effect, was not simply common factors, but to common principles that were present in most, if not all, psychotherapies. These principles, or strategies, were not as specific as techniques or as abstract as theories, but occupied a middle ground between them. The most important component of Goldfried's analysis, in my opinion, was his conceptualization of different levels of abstraction. He argued that the possibility of finding meaningful consensus exists at a mid-level level of abstraction somewhere between observable techniques and more theoretical interpretations of why these 14
techniques may work.
This conceptualization provided an accessible pathway for
integration in clinical practice and training, and it opened up a window for future research through the operationalization of core clinical strategies, which could be investigated within and between treatment approaches and problem areas. Any commonalities found at this level would represent strategies or principles of change that cut across these different orientations. One point of consensus can be found in the similarities across orientations with regard to phases or stages of therapeutic change (Vasco, 2006). A good point of departure may be the four-stage learning model. It is not clear who originated the very first 'conscious competence’- learning model. As well as various modern authors, like Abraham Maslow, one of the predominant minds in the field of psychology during the 20th century, sources as old as Confucius and Socrates are cited as possible earliest originators. Anyway, in the very broadest sense, it is possible to apply the Four Learning or Competence Stages to psychotherapy and view the change process as entailing four stages, from unconscious incompetence, to conscious incompetence, then to conscious competence and only finally to unconscious competence. Patients arrive with little understanding of the hindering reasons behind their problems (unconscious incompetence). They gradually increase awareness of those same processes responsible for their incompetence (conscious incompetence) but are not yet clear as to what they need to do in order to improve. Soon after they attempt to change such problematic processes (conscious competence) attempting new ways of thinking, feeling and acting. Finally, repeated attempts at new more effective patterns of functioning, over time, become more tacit and effortless (unconscious competence). Golfried (1980; 2006) also had his own proposal.
After translating the
theoretical jargon associated with different therapeutic schools into ordinary English, he
15
discerned a few common principles at that intermediate level of abstraction including (a) the presence of patients' expectations that change is possible, (b) the existence of an optimal therapeutic relationship, (c) providing feedback to help patients become more aware of aspects of themselves and others, (d) the encouragement of corrective experiences, and (e) facilitation of ongoing reality testing. In his classic contribution (Goldfried, 1980), the sequencing of theses principles or strategies was not that much emphasised. My sense is that, Goldfried (2006) now more explicitly stresses that an initial positive expectation that therapy and the therapist can be helpful, along with a certain amount of motivation to engage in the therapy process, together with the existence of an optimal therapy relationship, providing a safe context within which change (i.e. new ways of functioning, both within and outside of the therapy session) may be explored. According to him, patients’ problems rely frequently on old or inaccurate representations of themselves and others, and most approaches to therapy attempt to increase patients’ awareness, providing them with a clearer view of their perceptions, emotions, actions, and needs. When patients increase their awareness of their own functioning and their relationships with others, they are more willing to experience corrective experiences, to take risks and learn to function more effectively (e.g., through the therapy relationship, through specific tasks in or between sessions). A cycle of ongoing reality testing required for lasting therapeutic change is provided when patients continually become aware, have corrective experiences to support this new awareness, and use these experiences to attribute new meanings to their awareness (Goldfried, 2006). In another integrative model, that sequencing of general strategies had been emphasised from the very beginning.
The now called, future-oriented process of
therapeutic change (Beitman, 1987; Beitman, Soth & Good, 2006), entails a four-step
16
model interlaced with a conceptualization of an integrated psychotherapeutic interpersonal relationship that moves through the stages of engagement, pattern search, change, and termination.
This model stresses this sequential four-step model that
underlies but is less visible in different schools of psychotherapy. These steps include (1) identifying problematic responses, (2) activating the observing self, (3) identifying mismatches between expectation and experience, and (4) changing either expectations or external reality to realign them with one another. Stage one of this process defines the emotions, behaviours, and symptoms that are in need of change. During the second stage, a therapist guides the patient toward an objective view (the observing self) of his or her interior states, thoughts, desires, images and emotions. Consciously observing the self leads to step three: recognising mismatches between one's experience and one's expectations. Stage four concentrates upon resolving this disconnection in such a ways that desired future expectations are realised rather than not.
The future-oriented
supplement to this template involves, sequentially, activation of the patient's observing self, identification of problematic expectations, modification of problematic expectations, and co-construction of new expectations.
The patient will consider
termination of psychotherapy when he or she feels capable to act independently, is willing to let go of normal intensity and frequency of therapeutic collaboration, and can both maintain what has been learned and generalize it in daily living (Beitman et. al., 2006). These are but two of the most representative examples of integrative models, that explicitly, and from a theoretical point of view, propose the sequencing of general strategies or general principles independent of patient characteristics (see Vasco, 2006; Conceição, 2005 reviews of other proposals). As far as I am aware, there is no empirical evidence directly supporting these proposals. Such common stages or phases that may
17
have been observed clinically are thus in need of empirical confirmation, one of the research fronts in Paradigmatic Complementarity, from which, this research Project emanates. Paradigmatic Complementarity is a metamodel for psychotherapy integration that invests on the adequate articulation of existing models, mainly integrative ones, to somewhat constrain or tame the proliferation of integrative models. It is a model of sufficient breadth to orient the decision-making process of the psychotherapist as he or she practices in the real world. At the same time, it is of sufficient utility for guiding exploratory psychotherapy research that aims to identify emerging patterns of relationships between process and outcome, in such a way as to collect practice-based evidence, that may have relevance for the practicing clinician in naturalistic settings. As I was mentioning before, one of the research fronts in Paradigmatic Complementarity, the temporal sequence of phases structuring strategic therapeutic objectives (see figure 1), is believed to constitute a somewhat unrecognized similarity (actually, a general principle of change) among different theoretical orientations, both of an integrative and single-school (Vasco, 2001; 2006). According to this proposal, the therapeutic process unfolds as both patient and therapist progress along the following phases of general strategies: (a) trust, motivation, hope building and relationship structuring; (b) increasing awareness of self and experience; (c) new meaning making regarding self and experience; (d) regulation of responsibility; (e) implementation of repairing actions; (f) consolidation of change, and (g) relapse prevention and projection of self in the future.
18
g5. realize that he/she can cope autonomously with life’s future challenges g4. integrate experiences from the past, present and anticipated future into a coherent self-narrative. g3. strengthen the feeling of self-coherence and purpose in his/her life g2. anticipate resources required to cope with future scenarios beyond the termination of the therapeutic process g1. project himself/herself in the future, effective and affectively relating to self, others and the world f5. develop relationships/situations that support his/her choices f4. generalize the expression of his/her identity to the different domains of his life f3. accept the inevitability of a certain degree of vulnerability or conflict as a result of experiencing and expressing his/her identity f2. emotionally nurture and nourish himself/herself regarding the expression of his/her identity and growth f1. deal with internal or external obstacles to the consolidation of his/her identity and growth e5. effectively deal with situations and respect his/her needs at the same time e4. ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs. e3. integrate different parts or needs within himself/herself into a congruent and satisfied enough whole e2. stand up for her/himself by dealing with internal or external blocks to the expression of his/her identity e1. choose life-styles that allow him/herself to live the present as well as to promote his/her personal development d5. recognize himself/herself as the only agent of his/her own choices d4. understand when and how he/she contributes to his/her usual difficulties d3. acknowledge that he/she can promote or hinder the satisfaction of his/her needs d2. take responsibility for taking care of her/himself by tentatively mobilizing internal or external resources toward that goal d1. commit to respecting and validating his/her needs, regardless of their correspondance to others’ expectations c5. form new plausible links or explanations for his/her problematic attitudes and behaviour c4. understand what he/she is trying to attain when behaving in hindering ways c3. overcome the processes that hinder awareness of experience or helpful meaning making c2. identify existing patterns in his/her usual ways of functioning. c1. make sense of his/her problematic experiences in terms of past or present circumstances b5. translate problematic experience into its elements (e.g., cognition, emotion, behaviour) b4. feel curious and interested in observing how he/she treats himself/herself b3 become aware of parts or needs (of)within himself/herself which are in conflict b2 increase awareness of how he/she relates to others and others relate to him/her b1. explore or experience the impact of relevant situations on himself/herself a5. feel motivated to approach his/her problems from a psychological standpoint a4. experience the setting as safe and trust (in) therapist’s ability to help him/her a3. negotiate the structure and the rules of the therapeutic process in order to make it viable a2. feel a collaborative relationship with the therapist a1. feel hopeful that he/she might improve with the help of psychotherapy
© Paradigmatic Complementarity Phases Sequential Component; Vasco & Conceicao (1998:2008): A) Trust, motivation, hope building and structuring; B) Increasing awareness of self and experience; C) Meaning making regarding self and experience; D) Regulation of responsibility; E) Implementation of repairing actions; F) Consolidation of change, and G)Anticipation of the future and relapse prevention.
Figure 1: Phases’ Sequential Model of Paradigmatic Complementarity. 19
Principles of Change Across Disorders Such quest for empirical confirmation has recently been pursued not regarding stages or phases but general principles or strategies that had been observed clinically. Perhaps the most significant and comprehensive work on common and specific therapeutic factors, and the one most clearly influenced by Goldfried's article, is the volume, Principles of Therapeutic Change that Work, edited by Castonguay and Beutler (2006). In the preface, Castonguay states very explicitly (p. 4) that the impetus for this work came from Goldfried's (1980) article whose title, in fact, is similar to that of the book. This volume, reviewing the research literature on transtheoretical principles that have been found to contribute to successful change, resulted from the joint Task Force between APA’s Division 12 and the North American Society for Psychotherapy Research that commissioned an investigation of the relevant principles of change with regard to participant factors, relationship factors, and treatment factors. They took account of the findings of Divisions 12 (Clinical Psychology) and 29 (Psychotherapy) task forces on empirically supported treatments and relationships, respectively. For dysphoric disorders, anxiety disorders, personality disorders, and substance abuse disorders, respectively, a fourth chapter in each section is a summary and integration of the principles identified in the other three chapters. In their conclusion, the editors report on the common principles of change (those spanning two or more of the four disorders) vs. those that are unique (those that predicted outcomes within only one of the patient disorder groups). 19% of the change principles for participants were common as were 82% of the relationship variables and 55% of the techniques. In a consensus meeting, a list of both common principles and unique principles was consolidated, yielding a total of 61 principles was identified, most of which were common to a variety of disorders (Castonguay & Beutler, 2006b). 20
Whether or not one regards this work as weak, medium or strong support for common factors, it certainly provides a wealth of empirically supported information for the clinician and researcher. Interestingly enough these principles did not address the relative success or failure of brand names of treatment (Chambless, Crits-Christoph, Wampold, Norcross, Lambert, Bohart, 2006), that is, this project did not address the common factors as percentages holding across different theoretical orientations, which was Goldfried's (1980) original interest. This slight nuance is, in my opinion, worth emphasising, in that it highlights the latent idea that common principles may exist across different theoretical orientations (as common factors) but common principles also exist across different disorders, a point to which I will return later. It should be noted that these principles are consistent with the recent focus on evidence-based practice in psychology, which is a broader concept than simply empirically supported treatments or empirically supported therapy relationships. Evidence-based practice in Psychology (EBPP) is defined as the integration of the best available research with clinical expertise in the context of patient or patient characteristics, culture, and preferences (APA Presidential Task Force on EvidenceBased Practice, 2006; see also Fisher & O'Donohue, 2006; Goodheart, Kazdin, & Sternberg, 2006; Norcross, Beutler, & Levant, 2006).
EBPP also focuses on the
application of empirically supported principles of psychological assessment, case formulation, therapeutic relationship, and intervention in order to promote effective psychological practice and enhance public health. In this paradigm, treatment decisions should be made, whenever possible, on the basis of the results of empirical research that tests what treatment works for what problems experienced by patients with what important characteristics that might moderate the treatment outcome (Chambless, CritsChristoph, Wampold, Norcross, Lambert, Bohart, Beutler & Johannsen, 2006).
21
Where Did All The Specificity Go? An evidence-based practitioner bases his or her activities on the available evidence, in whatever form it exists, and actively pursues evidence related to innovative activities in development. The problem is that the field of psychotherapy research has ironically been dominated by an assumption that therapy techniques are the most powerful determinants of outcome. Much accepted evidence in the dominant circles has been, broadly speaking, technique-based. Ideally one has not to disregard the critical role of evidence, but to be more careful in the definition of evidence. In my opinion, one can always look for findings where the variance lies, without clinging to the ease of research design, as it happens in randomized controlled trials (RCTs). Goldfried had sagely stated “In our attempt to study the effectiveness of our therapeutic procedures, we have expended far too much energy investigating techniques that may not be all that powerful clinically” (Goldfried, 1980, p. 995). But, without regard to the small contribution made by technique to outcome, in contrast to the much larger effect of patient and relationship factors (Wampold, 2001), and with the increasing influence that managed health care has over treatment indications for the consumers of mental health services, efforts have focused on attempting to identify what treatments are most effective (Chambless & Ollendick, 2001). RCTs dominated the therapy effectiveness literature via a paradigm called Empirically Supported Treatments (ESTs) that are characterised by a manualized (i.e., reliably teachable) treatment approach for a particular sample or symptom, tested by an RCTs, preferably validated at more than one site. Insurance companies pay most reliably for ESTs, governmental agencies require the ESTs paradigm for research funding, and even APA nears mandating that training programs focus primarily on ESTs.
22
RCTs methods rely on concrete outcome measures to gauge effectiveness. The ESTs movement requires that therapeutic practice is itself rule-governed or constrained (if disorder x then therapies y and/or z), and therefore consistent and replicable, whose efficacy can be demonstrated by direct analogy to medical treatments for clearly defined maladies. Reed, Kihlstrom and Messer (2006) point out that an over-reliance on RCTs methods for determining effectiveness is likely to support those treatments that are best suited for this methodology, namely, medication. Not surprisingly, the majority of treatments
that
have
been
deemed
evidence-based
are
cognitive-behavioral
interventions, as its ingredients have been more accessible to empirical validation than those of other. RCTs methods exclude those treatments that may be perfectly effective, but do not allow for concrete measures or a clearly delineated treatment manual that specifies the conceptual and structural boundaries of the treatment to enable therapists to monitor their adherence to the treatment techniques (Addis, Cardemil, Duncan, Miller, 2006; Wachtel, 2010). Many patients receiving these therapies improve only a little or not at all and have a relatively high likelihood of relapse; and results of this kind of research may not apply that well to real world clinical practice since up to 90% of applicants for specific research projects are rejected, for the sake of forming homogeneous groups.
For
example, complex cases, not that rare, who have been difficult to treat for one reason or another, have been unresponsive to prior treatments (many medications and different psychotherapies), have a history of multiple suicide attempts and hospitalization, are highly comorbid (often including severely personality disordered, bipolar disorder, drug and alcohol abuse), and difficult to keep on protocol are routinely screened out of RCTs.
23
Not surprisingly, EST has been criticised, among other things, for: (a) its heavy reliance on pathologizing criteria for the designation of patient groups, deriving from its subscription to the medical model; (b) the marginalization of other psychotherapy approaches with different values; (c) its discrimination against legitimate alternative research paradigms; (d) the reliance upon outcome measures that do not assess other types of change; (e) little attention to patient variables in shaping outcome; (f) the emphasis on technical factors in treatment to the detriment of common factores or general principles; (g) the manualization of therapist behavior; (h) ignoring the dodo bird verdict; (i) recognition of pervasive methodological errors in the analysis of groupadministered treatments on the EST list; (j) the lack of explanatory value in terms of mechanisms of change. Wachtel (2010), for example, believes this research is of questionable scientific value and argues that much of this EST literature represents a problematic caricature of science that impedes the development of a discipline of psychotherapy genuinely grounded in the scientific method. The Renewed Interest in Transdiagnostic Approaches Traditionally, Cognitive-Behavioral Therapy (CBT) has progressed within a disorder-specific approach: distinct cognitive-behavioral models are developed to account for the onset and maintenance of symptoms within each specific disorder. The disorder-specific approach thus hypothesizes that the key onset and/or maintenance processes are significantly different across disorders and that different treatment interventions may be necessary for different disorders. Manualized treatment in research settings is, however, certainly not representative of treatment in clinical practice. One practical consideration in the use of treatment manuals to deliver treatment is the sheer number of manualized treatments. Over 150 different manualized treatments currently exist and have been deemed 24
“effective;” however, these manualized treatments only provide treatment for approximately one-third of the diagnostic disorders (Chambless & Ollendick, 2001). With this overwhelming number of treatment manuals, it would be impossible for real world clinicians to learn every empirically supported treatment. It is also important to consider that only about 6% of therapists report using manuals in all cases, whereas 47% of therapists report never using treatment manuals (Westen, Stirman & DeRubeis (2006). Most patients do not present for therapy with only one disorder – comorbidity rules, and patients do not generally compartmentalize their disorders in neat packages so that clinicians may treat one disorder at a time (Franklin, DeRubeis & Westen, 2006). Comorbidity also implies that many disorders may have etiologic factors in common. Besides, patients with the same diagnosis are not necessarily homogenous, not only with regard to comorbid conditions, but also in symptom presentation. Two patients may receive the same diagnosis, but be completely different, even in their symptoms. For example, with nine total criteria for a disorder and a threshold for diagnosis at five criteria, there are 256 possible ways to present with that diagnosis. Diagnosis cannot fully capture the intricacies of the person’s presentation or the context in which their problems exist (Reed, Kihlstrom & Messer, 2006). In response to these problems, transdiagnostic approaches to treatment have been attracting greater attention. A unified treatment approach emphasizing commonalities across disorders is not new to clinical psychology. However, the current context of this renewed interest in a unified treatment perspective, stemming directly from the ESTs/EVTs circles, may offer fresh insight into change processes and principles, given that it follows many years of cognitive research and treatment on various specific disorders.
25
A range of cognitive and/or behavioral maintenance processes are hypothesised to be shared across psychological disorders, that is, processes that are elevated in a wide range of psychological disorders relative to nonclinical controls and that causally contribute to the development and/or maintenance of symptoms (Harvey, Watkins, Mansell & Shafran, 2004). A transdiagnostic approach thus proposes that there is added value for understanding and treating psychological disorders by focusing on common and unified factors across disorders.
The treatment protocol can be applied, with
minimal adaptation or modification, to a range of different disorders. Two main approaches are evident in the development of transdiagnostic interventions. One is pragmatic, as illustrated by the work of Erickson, Janeck, and Tallman (2009), where the protocols are developed largely on the basis of clinical experience with interventions applicable to a range of disorders (e.g., relaxation training can be applied to many different disorders).
Barlow's recent shift from specific
treatments for specific anxiety disorders to what he calls a “unified protocol” based on principles of change (e.g. Allen, McHugh & Barlow, 2008) could be another example. The other approach is largely theory-driven, with protocols being developed to target cognitive and behavioral processes thought to be involved in a wide range of psychological disorders (Harvey, Watkins, Mansell, & Shafran, 2004; Mansell, Harvey, Watkins & Shafran, 2009). Mansell et al. (2004) attempted to systematically identify processes that could be identified in all disorders studied and therefore inform the development of a form of therapy that does not require knowledge of the patient’s specific diagnosis to operate effectively. Their review covered key processes that have been researched across adult Axis I disorders. For ease of explanation, they categorized the processes within the broad domains of (a) Attention (selective attention - external or internal, attentional avoidance); (b) Memory (explicit selective memory, recurrent
26
memories, overgeneral memory); (c) Reasoning (interpretational bias, expectancy bias, emotional reasoning); (d) Thought (recurrent negative thinking, positive and negative metacognitive beliefs, thought suppression), and (e) Behavior (behavior avoidance, safety behaviors) Cognitive-behavioral transdiagnostic model of emotional disorders will always leave space for common and specific processes, in terms of disorders. But in my perspective, it is worth noting that what is happening with CBT also happened in other integrative approaches, like Dialectical Behavior Therapy, Emotion Focused Therapy, EMDR, to mention just a few. Originally they were developed with a specific disorder in mind, and end up being generalized for other disorders.
The disorder-specific
theories appear to differ primarily in content rather than in process. Disorder specific in-session processes reflective of good process can easily lose its specificity. Evidence-Based Practice and Practice-Based Evidence Summing up, Goldfried (1980) suggested the explanation of the Dodo Bird effect could come not only from looking at common factors (Hubble, Duncan, & Miller, 1999), but even more so in looking for common principles that were present in most, if not all psychotherapies.
Castonguay and Beutler (2006) in their identification of
strategies or principles of change, found that a large part of them were common across disorders.
And currently from the EST/EVT circles one is witnessing a renewed
interesst in transdiagnostic approaches acrosss disorders. In any case the Dodo Bird points us toward looking for ways, whether in the form of factors, strategies or principles, or psychological processes, that we are more nearly alike rather than different. And by “we” I am referring to both therapists and patients. I am not suggesting that specificities should be abandoned. They have fueled much research on what ingredients make therapy successful: “What aspects of therapy 27
and what kinds of therapy, provided how and by what kind of therapist, under what circumstances, for what kinds of patients with what kinds of problems, are likely to lead to what kind of results?” (Orlinsky, Rønnestad & Willutzki, 2004, p. 362). I am nonetheless emphasizing the need for the specification of psychotherapy change processes, based in psychological processes. Chwalisz (2001) contends that evidence-based practice is merely a more explicit application of the scientist-practitioner model, and it should not be wed to a particular paradigm or research design. She suggests a neutral enough definition of evidencebased practice that includes (a) being aware of all possible treatment options that have appropriate (broadly defined) support, (b) formally informing patients about treatment options and making evidence-based suggestions, (c) engaging in empirically supported treatment activities within an evidence-based treatment environment, and (d) continually conducting research that supports practice activities. Such a definition would bring this paradigm of EBP to closer dialogue with its complementary paradigm, PBE that utilizes data from treatments as routinely delivered in practice settings (Barkham, Hardy & Mellor-Clark, 2010). Both EBP and PBE ideally should keep close attention to the most important common factor, the therapeutic relationship. The APA Division of Psychotherapy Task Force had been commissioned to (1) identify elements of effective therapy relationships; and (2) determine efficacious methods of customizing or tailoring therapy to the individual patient on the basis of his/her (nondiagnostic) characteristics (Norcross, 2002). Now, the conclusions of a second iteration of Empirically supported Relationships, Task Force II, jointly sponsored by APA Division of Clinical Psychology and Division of Psychotherapy (2009 – 2011) are that: (a) the therapy relationship makes substantial and consistent contributions to outcome independent of the type of
28
treatment; (b) practice and treatment guidelines should address therapist behaviors and qualities that promote the therapy relationship; (c) efforts to promulgate best practices or EBPs without the relationship are seriously incomplete and potentially misleading; (d) the relationship acts in concert with treatment methods, patient characteristics, and clinician qualities in determining effectiveness, and (e) adapting or tailoring the relationship to patient characteristics (in addition to diagnosis) enhances effectiveness (Norcross, 2010). Norcross (2010) in a SEPI keynote address integrating two of the most crucial developments in our field (evidence-based practices and the therapy relationship), acknowledged that what’s missing from EBPs is still the person of the therapist, the therapy relationship and the patient’s (nondiagnostic) characteristics. He reviewed the meta-analytic research compiled by a recent task force on (1) effective elements of the therapy relationship, (2) effective means of tailoring that relationship to individual patients, and (3) discredited relationship behaviors. Concluding that clinical experience and controlled research consistently demonstrate that the therapy relationship accounts for as much outcome as the particular treatment method, he presented a refreshed distribution of explanation factors of psychotherapy outcome, namely 5% of psychotherapy outcome variance is attributable to Interaction; 7% to the Individual Therapist; 8% to Treatment Method; 10% to Therapy Relationship; 25% to Patient Contribution, while 45% remains Unexplained Variance. Addis, Cardemil, Duncan, & Miller (2006) argue that “rather than attempting to fit patients into manualized treatments via EBP, we recommend that therapists and systems of care tailor their work to individual patients through practice-based evidence” (p. 147). Practice-based evidence utilizes information about patient gains throughout the therapy process to advance the quality of care (Barkham, Hardy & Mellor-Clark, 2010).
29
In this way, therapists can determine what seems to be working for this particular patient at this particular time, and can subsequently customize interventions as the need arises. Indeed, feedback to therapists about particular patient’s progress has been shown to increase treatment effectiveness (Lambert, 2010). So, rather than fueling the debate between what treatments or what relationship factors are most effective, Beutler (2009) suggests the redefinition of psychotherapy, both in clinical and research applications, as: the therapeutic management, control, and adaptation of patient factors, therapists’ factors, relationship factors, and techniques factors that are associated with benefit and helpful change. This redefinition has led to a marriage among treatment methods (Nathan & Gorman, 2002); participant predictors, and Empirically Supported Relationships (Norcross, 2002), and research-informed principles of change (Castonguay & Beutler, 2006; Chambless, Crits-Christoph, Wampold, Norcross, Lambert, Bohart, 2006). Since it is impossible to tease apart the ingredients that go into the recipe for successful therapy, it is most advantageous to consider all the factors together, in concert with each other. Hill (2005) proposes the following with regard to the interactions between therapist techniques, patient involvement and the therapeutic relationship: “The therapist cannot use techniques effectively if the client is not involved and there is not a good relationship; the client is not likely to be involved if the therapist does not skillfully use the techniques and there is not a good relationship; and a good relationship cannot exist without a competent therapist and involved client” (p. 438). The Black Box of Specificity: Linking Patient Change Process with Outcome To date, much of psychotherapy research has focused on the packages of interventions offered or the amount of therapy the patient has received. The therapeutic alliance is among the few factors which have consistently been shown to impact on 30
outcome, but how different interventions work or what effect they have on patients and patient in-session response to treatment has received less attention. According to Hill (2005), therapist’s techniques may very well account for much more of the variance than has been previously found, if their effects are not studied all together in a packet as is often done. It is also difficult to isolate the impact of therapist techniques since they are inextricably intertwined with patient involvement and the therapeutic relationship (issues like timing, appropriateness, quality, responsiveness to patient needs, or context). It can be case that these techniques operate in the same manner across orientations, even though they are not used in the same amount and intensity. In other words, the impact of therapist techniques might be similar across different types of therapy, with different patients, although some therapists would be more likely to use some techniques than others (Hill, 2005), as some patients would be more likely to benefit from some techniques than others. In two related studies (Kasper, Hill & Kivlighan, 2008; Hill, Sim, Spangler, Stahl, Sullivan, Teyber, 2008) Hill and collaborators studied the effects of a technical intervention, therapist immediacy (talking in the here-and-now about the therapeutic relationship), intensely in two cases of brief therapy, using a case study format. They looked at the different ways in which immediacy was used across cases and examined the immediate and delayed effects of immediacy, accounting for context and therapist and patient variables. Some effects were consistent across the two cases, whereas others were different. To determine the effects of other techniques, similar case studies could be done, but my impression is that it very much depends on the definition of technique. Traditionally technique has been associated with brand or packet of therapy, and this example of studying the effects of an intervention, like immediacy, has nothing to do with such an association. It becomes more principled like. The work of Safran and
31
collaborators (e.g. Safran & Muran, 2001; Muran, Safran, Gorman, Samstag, EubanksCarter, & Winston, 2009), provides an even more clear example of technical interventions more defined as principles of intervention as they describe how to recognize, address and repair ruptures in the therapy relationship. The same could be said of the work of Greenberg and collaborators, specially the later work of Pascual Leone (Pascual Leone & Greenberg, 2007), that more clearly transcends the specific tasks of Emotion Focused Therapy, as he describes the process of how patients resolve the generic problem of emotional distress. Finally the work of Stiles (2002) would also count as another example of principled-like techniques tha facilitate how patients assimilate problematic experiences. All these are examples of research that yield clinically useful information that supports desired therapy change.
Somehow they are technical interventions, or
instances of moment-by-moment responsivity, but they transcend theoretical technique, in that they do not need to be tied to the theories they emerged from. Principles allow for greater flexibility than a simple rule prescribing behavioral responses. Depending on their theoretical orientation, therapists may engage a variety of (technical) interventions to pursue the same intention and in doing so they may pursue the same overarching goal, allowing greater sensitivity to changing needs of both participants in the psychotherapy session.
These principles may allow therapists to cross the
boundaries of therapy orientation by focusing on underlying common intentions that may be enacted in different forms. Interestingly, London (1964) had already envisioned that in the future one would speak of principles of change rather than theoretical orientations. Levitt, Neimeyer, & Williams (2005) believe that a principled approach to the regulation of clinical decision-making is more promising than the manualized-protocol-
32
-and-treatment-assignment orientation of the ESTs movement, and that the attempt to invoke a consensual application of broad standards actually pays further dividends by promoting the goals of therapy integration and by engendering educative discussion of therapist intentions across practitioners of different orientations.
As opposed to
generating principles for change only from the analysis of experimental methods, they advocate the creation of principles for change by examining intentional processes as they unfold within a change process. Identifying patterns that existed within therapists’ subjective understanding can lead to models of psychotherapy that are sensitive to contextual important factors. These understandings, in the form of principles, can direct the application of interventions. For example, the principles presented by Williams and Levitt (2007) regarding the facilitiation of agency were derived from expert therapists across different therapy orientations and read as follows:
(a) therapists strived to
promote patient agency by encouraging patients to introspect, but they lacked this skill or ability, when therapists would use an educational mode; (b) therapists sought to guide patients to decide what was in their power to change and what was not and (c) when therapists perceived that patients were not progressing, they desired to increase their patients’ awareness of how and why they were inhibiting change. Principles like these can be used by a variety of orientations to better facilitate patients’ agency. In addition, the contextually sensitive descriptions of issues related to agency can be helpful for considering items on measures to assess the development of agency within therapy. According to Williams and Levitt (2007) by moving agency into the foreground of therapists’ heads, therapists and researchers may be better attuned to facilitating patients’ agency and have concrete principles on which to draw while studying or facilitating moment-to-moment interactions in therapy.
33
The truth, however, is that the predominant randomized controlled trial research frequently utilized process measures to demonstrate therapists’ adherence to a particular brand or packet of treatment (Waltz, Addis, Koerner, Jacobson, 1993). Measuring adherence to a specific form of psychotherapy tends to be limited in focus, assessing only those core features associated with the specific form of treatment being utilized. Such measures are like a black box hiding more than what they reveal, which may be inappropriate or ineffective for studying the processes of other forms of psychotherapy, especially those practiced in naturalistic treatments, or for comparing process across different types of cases. Greenberg and Foerster (1996) suggest that one of the major problems with clinical trials is their failure to account for the absorption of treatment by the patient or activation of change processes when evaluating the effects of different treatments on outcome. “A therapist may deliver the treatment, but does it take? Does it set the anticipated change processes in motion? It's all very well, for example, to suggest that a patient experiment with expressing feelings to a significant other in an empty chair, but if a treatment group contains some patients who become engaged in the process, some who do so only intellectually, and others who refuse, a true test of the active ingredients of the treatment will not be obtained. If, however, in a specified treatment, those patients who go through the process of change in the required manner are studied to see if they change more than those who don't, the effects of the change processes are then truly being tested. The link between patient change process and outcome needs to be studied and the specific effects of particular processes need to be demonstrated. To do this, rationally, empirically derived patient change processes close to the level of performance need to be specified and measures of these processes constructed” (Greenberg & Foerster, 1996; pp. 445-446).
34
Greenberg and Foerster (1996) are but emphasising the need for the study of the link between patient change process and outcome, and for the effects of particular processes to be demonstrated. In order to evaluate the effects of psychotherapy, not only do we need to know whether the therapist made a particular intervention, but we also need to know whether the patient processed the intervention, and in what ways, and then behaved in some way to indicate that the intervention had an impact. Detailed exploration of change processes and change mechanisms may provide data relevant to another question of interest in current psychotherapy research: the psychotherapy dose-effect relationship (Howard, Kopta, Krause & Orlinsky, 1986; Kopta, 2003). I will later devote substantial time writing about mechanisms, but let’s first have a look at this three-phase model conceptualization of outcome.
Why I
consider important? Because of the usually present void regarding change processes and mechanisms when studying and defining outcome. Outcomes: What is Measured and What is Left to Measure Although therapy has generally been found to be effective, data is inconclusive with regards to the amount of therapy necessary to attain desired outcomes. How many sessions a patient needs has become an important factor in efficacy research. With the advent of managed health care, insurance companies often arbitrarily limit the number of sessions a patient may receive. The influence of time-limited psychotherapy and a move towards shorter therapy services makes this question of pivotal importance to service providers, specially to clinicians who, in this managed-care era, are under pressure to provide effective relief in the shortest time possible.
More generally,
longer-term treatment is associated with better outcome (Orlinsky et al., 2004). Patients who present with multiple diagnoses will, on average, require longer-term treatment in order for it to result in positive outcome (Westen, Stirman, & DeRubeis, 2006). While 35
data support the use of brief therapies for some problems, their use with more severe problems is doubtful (Lambert & Ogles, 2004), as their use for some goals, I believe, may also be. The mean number of therapy sessions a patient receives differs by site and by provider and while the modal number of sessions patient receive is one, most receive less than five sessions, though their research suggests that between 13 and 18 sessions are typically required for 50% of psychotherapy consumers to achieve improvement in their symptoms (Hansen, Lambert, & Forman, 2003). An additional 25% will see significant change after 50 sessions. These results also indicate that 50% of patients are not well served by a limited number of sessions. There are many factors that impact the relationship between number of treatment sessions and outcome, including chronicity and reactivity of the patient’s symptoms, the type of problem presented, the type of treatment and method of assessing outcome, as well as the goals or outcomes desired by the patient, I would again add.
Nonetheless, several studies have concluded that the
more psychotherapy a patient receives, the more change the patient is likely to achieve (Hansen, Lambert, & Forman, 2003). In therapy outcome research this positive relationship between the dose of therapy (the amount of psychotherapy a patient is exposed to) and the amount of change that may be expected is called dose-response effect (Howard, Kopta, Krause, & Orlinsky, 1986). For example, Howard, Lueger, Maling & Martinovitch (1993) and Martinovitch (1998) reported data demonstrating that most change occurs quickly and early in therapy and the rates of change level off after a particular point, after which more therapy does not necessarily mean more change, or at least not the accelerated rate of change seen early in therapy. This is called the "diminishing returns effect" of psychotherapeutic change. Possible explanations of the diminishing returns effect are:
36
(1) patients experience change at different rates, with some faster and other slower; (2) the rate of improvement may be dependent on the symptoms and diagnosis for which they seek help (rapid improvers, once out of therapy, do not contribute to aggregate data concerning change for later sessions, which then is limited to data collected from patients with more complex conditions and slower change rates); (3) patients' idiosyncratic interpretations of self report outcome instruments (initially the same amount of change is experienced and reposted as larger than subsequent change), and (4) change across various domains may occur at different rates. Improvement in some domains may be facilitated by or even dependent on successful change in other domains; the observed diminishing returns effect may be the result of psychotherapy progressing in phases in which certain domains of patient functioning are impacted at different times and at different rates, perhaps even in a dependent, progressive fashion. As mentioned before and as reviewed by Vasco (2006) and Conceição (2005) numerous theorists have proposed phase models of psychotherapy. Probably the only one explicitly dealing with domains of psychotherapy outcomes, as traditionally defined, however, is the three-phase model. Howard et al. (1986) proposed this model of therapy as as specific explanation of the dose-response and diminishing returns effects referred to above.
They posited that psychotherapy occurs in stepwise,
dependent phases, each phase representing a different domain of change impacted by psychotherapy.
They theorized that these domains are differentially impacted by
psychotherapy, with regards to rate and extent of change. The phases are stepwise and dependent in that the major tasks of each phase must be completed in order for the subsequent phase to be sucessful. The first phase, remoralization, consists of therapists enhancing their patients' subjective well being, trough the alleviation of hopelessness and pessimism and the
37
promotion of optimistic expectations about the benefits of therapy and reframing problems as at least partially internal rather than completely external. Arousing hope that their problems can be solved results in an early occurring enhancement of patients’ subjective well-being. Remoralization is completed relatively quickly in therapy (by session four or before), and it is possible that some patients come to therapy already having completed this phase (Howard et al., 1993). In the second phase, remediation, the focus is on resolving patients' symptoms, life problems, or both, since the interventions are concerned with facilitating mobilization of patients' coping skill, encouragement of more effective coping skills, or both. When a positive therapeutic alliance has been established and the patient has mobilized existing or learned new coping skills, symptom remission tends to be achieved. Many patients terminate treatment after this second phase is accomplished, i.e., when their symptoms have abated (Howard et al., 1993). However, some patients will continue in therapy in order to address issues of a more enduring nature, perhaps maladaptive behavioral, cognitive or emotional longstanding patterns that have contributed to the development of the immediate symptoms for which they sought treatment and block the access to new ways of dealing with various aspects of self and life. This third phase, rehabilitation, incorporates issues of interpersonal functioning, social role functioning, and general life functioning, focusing on more enduring changes that are impacted much more slowly than other domains of change associated with psychotherapy.
Once the longstanding and
maladaptive patterns of living are modified and new ways of dealing with various aspects of life and self are established, it results in an enhancement of life functioning. The length of time required for phase three to be completed depends on a number of different factors, particularly the accessibility and malleability of these patterns in each
38
patient. Efforts to maintain gains from therapy are also part of phase three (Howard et al., 1993). Howard et al. predicted that "improvement in subjective well-being would occur quickly, improvement in symptomatology would occur more slowly, and improvement in life functioning would occur even more slowly," (Howard et al., 1993, p. 680). They also hypothesized that improvement in therapy would demonstrate a causal, stepwise pattern from phase to phase. Howard, Lueger, Maling, and Martinovich (1993), and Kopta, Howard, Lowry, & Beutler (1994) are just two of several examples of studies that found supporting evidence. While measuring certain outcome domains is certainly relevant for psychotherapy as practiced in real world my question is what is left to measure. As mentioned before, evidence starts to accumulate with regard to the importance for therapists and mostly for patients of routinely monitoring outcomes (Barkham, Hardy & Mellor-Clark, 2010; Lambert, 2010).
Indeed several widespread used measures
appeared recently in the field of practice-based evidence, namely, the Outcome Questionnaires System (OQ; Lambert, Hansen, & Harmon, 2010), Treatment Outcome Package (TOP; Kraus & Castonguay, 2010) and Clinical Outcomes in Routine Evaluation (CORE; Barkham, Mellor-Clark, Connell, Evans, Evans, & Margison, 2010). This is good news because the number of outcome measure was so enormous that having these three may spare a lot of resources to clinicians and researchers. But my concern is if mainstream outcome research may not be asking the right questions, in terms of the criteria that are used to evaluate outcome, and as a result ends up not doing justice to what happens in therapy? As Sales (2010) put it, “it is not because information is gathered in real settings, that it reflects (real) psychological change” (parenthesis added). The field may be
39
moving towards the monitoring of outcomes but there is still the risk that practioners do not learn much about doing better therapy. I agree with the Duncan’s (2010) idea that awareness of being bad can make us better therapists, but I doubt systematically monitoring our clients’ treatment response and the therapeutic alliance (as rated by the client) and to discuss these phenomena with the client do most of the work of the therapeutic act. Aren’t we, psychologists, specialized in psychological processes, and we psychotherapists, specialized in change strategies, principles and processes, after all? Or should we not? Perhaps the field could consider monitoring mechanisms of change as long as they would be in plain English and based on general strategies or principles of change or plain psychological processes. Outcomes Across Approaches Although changes in these symptomatic qualities may be connected with personal growth, generally they are not considered to be the central foci of humanistic or integrative change. What do symptom changes or their absence mean in patients’ voice (McLeod, 2011)? Or what is the patient actually capable of? So, on the one hand adherence process measures are not able to reveal what therapists are actually doing in the session and, on the other hand, most outcome measures are not able to reveal the change processes patients undergo as a result of the absorption of therapists’ interventions and therapeutic setting in general. In other words, what are patients doing in session and outside the session with regard to change processes, or psychological processes. Hypothesis about intermediate results on the way to final outcome (which is, more often than not, limited in scope) remain difficult to investigate. Change processes and change mechanisms need adequate instrumentation to be captured. For example, asking about emotional pathology is not the same as asking about comfort with emotion, the resolution of negative emotion, or emotional expression. So if one were 40
about to entertain the hypothesis of emotional regulation as a mechanism of change, one would better choose an appropriate measure. Actually, there are critiques that these measures might be befitting a medical approach to psychotherapy and do not assess types of change that are prized across therapy orientations and, as such, generate poor assessments of humanistic, psychodynamic or even integrative approaches.
In fact, outcome measures can
privilege psychotherapy orientations most associated with their theory of origin and rarely assess the different goals of the therapy orientations being evaluated (Levitt, Stanley, Frankel, & Raina, 2005). Measures that are especially attuned to the function of a certain therapy may be especially apt to reflect the types of changes that are valued within that approach, compromising the ability to track or assess whether these therapies have helped patients develop in other relevant areas. Two therapies can be shown to produce equivalent results in several measures, but different in others. Levitt, Stanley, Frankel, and Raina (2005) indeed examined the nine most commonly used psychotherapy outcome measures in 85 studies comparing humanistic psychotherapy with control groups or other approaches (some conducted by humanistic therapists but many by therapists with other allegiances as well). Of the nine measures assessed, most did not adequately inquire about important goals of humanistic psychotherapy - the symbolization of, access to, or expression of new emotions; the development of new meaning connected to internal experience; increased awareness of new choices; and interpersonal awareness or awareness of personal growth or insight. Instead, as Levitt et al. (2005) concluded, there was a focus on physical symptoms and psychopathology, as would befit investigations that are based on a medical approach to psychotherapy.
41
As I have mentioned before, easily manualized therapies predominate in outcome research, receiving the empirical support clinicians need to justify practice. Therapeutic orientations that are less easily operationalized have received less empirical support.
Reviews of efficacy research have provided conclusions that privilege
manualized, outcome-oriented, pure-form therapies, while leaving more integrative therapies behind (Luborsky, Barrett, Antonuccio, Shoenberger, Stricker, 2006). Most practicing clinicians ascribe to an integrative or eclectic orientation whereby they extract principles from the interventions they feel will best help their patients with their problems. This poses a serious problem for translating evidence-based treatments to the real world of practice. In order for a therapy to be deemed evidence-based, there must be clear and specific techniques, as well as a specific behavioral outcome (Wachtel, 2010). Integrative researchers may be at a disadvantage if their treatments are being evaluated by their ability to achieve goals that are not their own. The study from Levitt et al. (2005) confirms the desirability of developing and selecting theory-congruent measures. Like with humanistic psychologists if integrative or real world practitioners are to show that therapy works the way they say it will work, their instruments need to reflect their own theories of development and models of psychotherapy. And ideally the client’s voice would not be forgotten in such an enterprise. Expansion of Research Designs and Instrumentation If one becomes interested in bridging the link between patient change process and outcome, demonstrating the specific effects of particular patient change processes promoted by the therapist (operations), as stressed by Greenberg and Foerster (1996), and if one is not interested in a specific brand of therapy one has probably to expand the most frequent research designs and instrumentations. A task-analytic research program 42
outlined by Greenberg and collaborators (Greenberg & Foerster, 1996, Greenberg, 2007, Pascual Leone, Greenberg, & Pascual Leone, 2009) is an ideal method for constructing empirically grounded models of how people change in psychotherapy. This method belongs to a more general category of qualitative approaches that may vary in the level of quantitative rigour and relevance they include. Taken together, this line of inquiry stands as a viable path through which researchers can develop understandings of in-session processes of change, since these methods allow a focus on subjectivity that is appropriate for understanding therapy and allow participants to articulate and contextualize elements of change that appear to be important in their own experience. I would include case studies and single-subject designs in this general category. According to McLeod (2011) “When clients are interviewed about their experience of what ‘came out’ of their therapy, on the whole they do not talk in terms of symptom reduction. What they describe, instead, is something that is more of a dynamic process, where they learn new skills, ideas and strategies that they can then apply in the future in order to live more satisfying lives'' (end of chapter 13). Alternatively, one could also develop measures more specifically designed to assess the kinds of processes and mechanisms I am referring to. Next I will briefly review these two approaches before reviewing part of the emerging literature on mechanisms of change. Qualitative approaches and single subject designs. Qualitative approaches. Of particular importance in this category is the study of ‘‘significant psychotherapy change events or episodes or moments’’ (Elliott, 1984; Rice & Greenberg, 1984, Fitzpatrick & Chamodraka, 2007). Qualitative research methods
43
are used extensively to identify and define these change events, especially in efforts to discover these episodes or their components (e.g. Hill, Knox, Thompson, Williams, Hess, & Ladany, 2005). The goal of the methodological approach to the episode is the identification and exhaustive description of ‘‘moments that, in accordance with specific criteria, stand out in the therapeutic process as ‘significant,’ ‘noteworthy’ or ‘relevant for change’’. Change moments are windows on the inner workings of the process of change in psychotherapy. These episodes have variable temporal limits and may last anywhere from a couple of therapeutic interactions to 20 to 40 min or even more than one session (Rice & Greenberg, 1984). The study of in-session episodes allows for a generic or common factors approach to unveiling the active ingredients (independent variables) associated to change, able to be applied to different theoretical approaches, and in that way contribute to a cumulative knowledge base on these healing factors. Using qualitative analyses can provide a foundation for later verificational assessments of how psychotherapy works. An understanding of specific processes of moment-by-moment change can generate suggestions for practices that are sensitive to the context within which that change happens. Empirical data on the nature of change may aid us in approaching a conceptual definition of change. Across the significant moment and other qualitative research studies, the factors identified as significant have differed.
A broad spectrum of themes have been
examined in interview studies, namely patients’ descriptions of moments of misunderstanding (Hill, Nutt-Williams, Heaton, Thompson, & Rhodes, 1996) insight events (Elliott, Shapiro, Firth-Cozens, Stiles, Hardy, Llewelyn, & Margison, 1994), and helpful events (Paulson, Truscott, & Stuart, 1999), to name just a few. Many of the studies on assimilation of problematic experiences, resolution of alliance ruptures and
44
facilitation of emotional change (Safran & Muran, 200; Stiles, 2002; Elliott, Watson, Goldman, & Greenberg, 2003) were also conducted within such a significant moments’ template. Relevant in terms of generic therapeutic change, Mahrer and Nadler (1986) report 12 categories of ‘‘good moments’’ in therapy, including the provision of significant material, description and exploration of feelings, emergence of previously warded-off
material,
expression
of
insight
and
understanding,
expressive
communication, expression of good working relationship with the therapist, expression of strong feelings toward the therapist, expression of strong feelings in personal life situations, manifest presence of a substantively new personality state, undertaking new ways of being and behaving in extratherapy life situations, reporting changes in target behaviors, and expression of a welcomed state of well-being. A meta-analytic review of qualitative studies by Timulak (2007) synthesizes patient-identified
impacts
of
helpful
events
in
nine
core
categories:
(a)
awareness/insight/self-understanding; (b) behavioral change/problem solution; (c) empowerment; (d) relief; (e) exploring feelings/emotional experiencing; (f) feeling understood; (g) patient involvement; (h) reassurance/support/safety; and (i) personal contact. These episodes are intensely productive instances of the therapeutic process that are identified as being helpful for the patient’s healing process, and therefore, allow a deeper understanding of what works in therapy and in that way inform clinicians on how to contribute to patients’ change process (Timulak, 2007). Summing up, many of these studies are particularly helpful to foster responsiveness at a moment-to-moment level (for reviews see Vasco, 2006; Conceição, 2005). What about qualitative inquiries into the temporal sequential progression of the therapeutic process, possibly highlighting change mechanisms or intermediate results?
45
In an interesting qualitative grounded theory study, Levitt and Williams (2010) propose the process of expanding and integrating awareness can be understood as a common factor across psychotherapy orientation. Although the scope and focus of awareness may differ, they stress, all therapies guide patients to expand their awareness. They further state that although it has received a great deal of attention within psychotherapy approaches, it is a topic that has not received much study across approaches, as they did. The findings of Levit and Williams (2010) yielded five main dimensions. First, therapists helped patients to develop curiosity about themselves so they could become motivated to engage in the tasks of therapy. Second, this curiosity allowed them to sustain a reflexive exploration into vulnerable and emotionally charged topics, which could be anxiety provoking. Third, patients required increased structure (e.g., through guided exploration or interventions) and/or support (e.g., a sense of safety and being valued) as needed in relation to their level of risk. Fourth, the therapists guided patients toward encounters with threatening experiences and the development of not only new intellectualized ideas but actual experiences of new awareness, in particular an experience of difference from the way a problem was experienced at the beginning of therapy.
These new experiences were based on the exploration of a previously
inaccessible aspect of their selves (e.g., a cognitive, emotional, social, or interpersonal aspect) that became available within a sustained reflexive analysis. Finally, fifth, by symbolizing this new experience in metaphors and language, patients gained greater access to that experience and could integrate it through further discourse and come to rely on it in understanding themselves and forming expectations of the future (Levitt & Williams 2010).
46
Another interesting attempt to study change episodes within an integrated generic model of change, is that of Krause, de la Parra, Arístegui, Dagnino, Tomicic, Valdés, Echávarri, Strasser, Reyes, Altimir, Ramírez, Vilches, and Ben-Dov (2007) who adhere to the notion, that the essence of change relates to the transformation of the patient’s subjective perspective. According to these authors, these representational changes can be conceptualized as changes in the subjective constructs and theories of patients, which are defined as a set of personal cognitions about oneself and the world that serve to guide individuals’ behavior and optimize self-value. This process of change in the subjective patterns of interpretation and explanation shows an evolution in successive stages and involves an increasing, although not necessarily linear, process of construction of psychological patterns of explanation and interpretation as well as a progressively increasing level of complexity and elaboration of these patterns that build on previous and less complex ones. This model of evolution of psychotherapeutic change has been supported empirically and operationalized through a hierarchy of generic change indicators grouped in three categories. The first grouped category (initial consolidation of the structure of the therapeutic relationship) is composed by the following ordered items: (1) Acceptance of the existence of a problem; (2) Acceptance of his/her limits and of the need for help; (3) Acceptance of the therapist as a competent professional; (4) Expression of hope (“moral boost” or “remoralization”); (5) Questioning of habitual understanding, behavior, and emotions (“opening up”); (6) Expression of the need for change; and (7) Recognition of his/her own participation in the problems (Krause et al., 2007).
The second grouped category (Increase in permeability toward new
understandings) is composed by the following ordered items: (8) Discovery of new aspects of self; (9) Manifestation of new behavior or emotions; (10) Appearance of
47
feelings of competence; (11) Establishment of new connections among: aspects of self, aspects of self and the enviornment, or aspects of self and biographical elements; (12) Reconceptualization
of
problems
and/or
symptoms;
(13)
Transformation
of
valorizations and emotions in relation to self or others (Krause et al., 2007). Finally, The third grouped category (Construction and consolidation of a new understanding) is composed by the following ordered items: (14) Creation of subjective constructs of self through the interconnection of personal aspects and aspects of the surroundings, including problems and symptoms; (15) Founding of the subjective constructs in own biography; (16) Autonomous comprehension and use of the context of psychological meaning; (17) Acknowledgment of help received; (18) Decreased asymmetry between patient and therapist; and (19) Constrution of a biographically grounded subjective theory of self of his/her relationship with surroundings (Krause et al., 2007). Krause et al. (2007) showed that these indicators were indeed generic in the sense that they can be observed in different forms of therapy and that change indicators indeed occur in the theoretically established sequence (i.e., from those of less hierarchy in the early therapeutic process to those more highly ranked toward the end). Single subject designs. Many scholars argue for the importance of using singlesubject designs in psychotherapy research (Greenberg, 1986; Jones, 1993; Hilliard, 1993; Elliott, 2002) and are becoming increasingly valuable (Heppner, Kivlighan, & Wampold, 1999), since they can be crucial to understanding the relationship between psychotherapy processes, in-session patient changes, and outcomes.
The primary
criticism of single-subject research is the difficulty in generalizing findings (external validity) and in drawing conclusions about causality (internal validity). But to see the value of single-subject research, one must view it as intrasubject research (Greenberg, 1986). Through this lens, internal validity issues can be addressed 48
by collecting systematic, objective, qualitative data through a variety of methods, administering measures over a period of time, comparing post-session change to stable pre-treatment levels, and assessing the magnitude and onset of the change that occurred. External validity in the single-case design is addressed through the replication of results across individuals that are expected to demonstrate both similarities and differences. We can define quantitative single case research briefly as the repeated collection of quantifiable (numerical) data on a single clinical case, typically involving repeated measurement over a baseline period (Morley, 2007). The main strategy for doing this is to focus on within-subject variability and to seek to understand and explain the causes of that variation (Barlow & Hersen, 2008; Kazdin, 1981; Morley, 1996), ruling out alternative explanations for a finding. Single-subject designs have the potential to generate rigorous and relevant data meeting the twin demands of practice-based evidence discussed before, opening a productive discourse between practice and laboratory. Statistical sophistication can also be a reality in such designs. Time-series designs have explicitly been recognized as important methodological approaches that can fairly test treatment efficacy and/or effectiveness (Chambless & Ollendick, 2001). With single-case time-series one can also address mechanisms of change as well as sequencing (Borckardt, Nash, Murphy, Moore, Shaw, & O'Neil, 2008). Either efficacy or the proposed mechanisms of change have tentatively been tested through single-subject designs, in several approaches, like Schema Therapy (Nordahl & Nysæter, 2005), Cognitive Analytic Therapy (Kellett, 2005; Kellett, 2007) Functional Analytic Psychotherapy (Kanter, Landes, Busch, Rusch, Brown, Baruch, Holman, 2006; Busch, Kanter, Callaghan, Baruch, Weeks, & Berlin, 2009), to mention just a few.
49
Measuring interventions and change processes in real-world therapy. Many psychotherapy researchers, practitioners, and theorists presume that the actions of the therapist are what bring about change in therapy. Studies of therapy process and outcome concentrate heavily on therapists’ delivery of certain treatment interventions. As I mentioned before, a number of instruments have been developed to assess treatment adherence, technique, and process for a variety of therapies.
However,
measures of adherence are, by definition, wedded to a specific treatment manual (Waltz, Addis, Koerner, & Jacobson, 1993), which limits the applicability and utility of these instruments to more real-world therapies that (a) do not use a treatment manual, (b) do not use the specific manual for which the instrument was developed, or (c) use techniques from different manualized treatments (both within or across theoretical orientations). Many studies examine whether treatments using various theoretical orientations differ in the interventions therapists use and whether specific therapist actions relate to change in the therapy process and outcome (Orlinsky, Rønnestad, & Willutzki, 2004). Due to the, still somewhat unknown, importance of therapeutic interventions, many survey instruments have been developed to assess the interventions delivered in a therapy session, but more often than not they only measure interventions of one or a small number of orientations, leaving interventions not included in those particular orientations overlooked. Besides, scales assessing interventions from only a limited number of orientations may not describe fully what is happening in these real-world therapies, that tend to be of an integrative nature. Apart from being different in some characteristics, we expect psychotherapies to be the same in other characteristics. To what extent do specific psychotherapies contain interventions from different
50
psychotherapy orientations?
Instruments developed to measure interventions from
multiple therapeutic orientations could present a welcome advantage. Comparative psychotherapy process scale (CPPS). A few of such measures start to emerge in the literature. The Comparative Psychotherapy Process Scale (CPPS) (Hilsenroth, Blagys, Ackerman, Bonge, & Blais, 2005; Hilsenroth, Defife, Blagys & Ackerman, 2006) is intended to be a brief measure assessing the degree to which any delivered therapy used global techniques and adhered (more or less) to the distinctive features of Psychodynamic/Interpersonal (PI) and Cognitive Behavioral (CB) approaches.
The seven general techniques that consistently and significantly
distinguished PI from CB treatments are: (a) focusing on patients’ affect and the expression of emotion; (b) exploring patients’ attempts to avoid topics or engage in activities that hinder the progress of treatment; (c) focusing session on patterns in patients’ actions, thoughts, feelings, and relationships; (d) exploring patients’ past experiences; (e) focusing on patients’ interpersonal experiences; (f) focusing discussion on the therapeutic relationship; and (g) exploring patients’ wishes, dreams, or fantasies. On the other hand, the six general techniques that consistently and significantly distinguished CB from PI treatments are: (a) assigning homework and outside of session activities; (b) actively directing session activity; (c) teaching specific coping skills; (d) focusing on patients’ future experiences; (e) providing patients with information about their treatment, disorder, or symptoms; and (f) focusing on patients’ cognitive experiences (e.g., dysfunctional or irrational beliefs). The CPPS developed to be a more general instrument is intended to offer a reliable alternative that has general real-world (i.e., not manualized) applicability to a variety of treatments (both within and across theoretical orientations).
51
Therapeutic Focus on Action and Insight (TFAI). Samoilov, Goldfried, & Shapiro 2000 recently developed a measure of therapeutic process, Therapeutic Focus on Action and Insight (TFAI). The TFAI contains 12 coding categories tapping the two factors: construction of meaning and facilitating action. It is designed to be used with video or audiotapes and written in a language allowing less-experienced raters (e.g., undergraduate students) to reliably code the measure. Comprehensive Psychotherapeutic Interventions Rating Scale (CPIRS). Trijsburg et al (2002; 2004) composed another instrument that may have also some potential value for studying psychotherapies as they are conducted in clinical practice, and that may also be helpful in determining the degree in which a treatment is eclectic/integrative or pure form. The Comprehensive Psychotherapeutic Interventions Rating Scale (CPIRS) was designed to cover several orientations. The CPIRS consists of interventions derived from the main psychotherapy orientations, yielding twelve factors (with 67 interventions in total), representing specific and common factors. The specific factors are behavioral, cognitive, experiential, psychoanalytic, psychodynamic, and strategic interventions and experiential procedures [chair work]). The other factors - facilitating, authoritative support, coaching, directive process, and structuring interventions - may be viewed to a certain extent as common factors. Particularly interesting and revealing for me is that this structure is not necessarily stable, depending from the therapists in each sample on which the factor analysis is conducted, as was evident from the first to the second study using this measure. The volatility of the dimensions probably reveals integrative decision-making by the therapists in real-world clinical settings.
52
Multitheoretical List of Therapeutic Interventions (MULTI). More recently, McCarthy & Barber (2009) introduced a new measure of psychotherapy interventions, the Multitheoretical List of Therapeutic Interventions (MULTI), that assesses the three perspectives on the therapy process using a similar format and content, while including interventions from the most widely practiced and researched therapy systems. The current MULTI consists of 60 items and eight subscales: Behavioral Therapy (BT), Cognitive Therapy (CT), Dialectical Behavioral Therapy (DBT), Interpersonal Psychotherapy (IPT), Person Centered (PC), Psychodynamic (PD), Process-Experiential (PE), and Common Factors (CF). Most of the items were specified uniquely to one subscale; however, 14 items appear in two subscales and five items appear in three subscales. Again, interesting and revealing for me is that, although this theory-driven model of MULTI sub-scales adequately explained the variation in their data (i.e., low RMSEA values), it was not among the simplest ways possible to reduce the data (i.e., low CFI values). This suggests that there might be more overlap in intervention use among therapists belonging to different orientations. As McCarthy & Barber (2009) acknowledge, competing factor models that organize interventions into fewer subscales might explain the relations among interventions more parsimoniously than the model the authors suggested. These are four exemplars of the few measures that exist in the literature that could be used to measure therapy as it may happen in naturalistic settings. On the other hand, with regard to existing measures to assess change processes or intermediate results as they may happen in real-world, I believe there are even less, if any. Nonetheless let me review three interesting measures with some potential in that regard.
53
The Therapeutic Realizations Scale - Revised (TRS-R).
The construct of
therapeutic realizations is comparable to what others have referred to as session impacts, in that it refers to session-level effects of therapy for the patient (Kolden, Strauman, Gittleman, Halverson, Heerey & Schneider, 2001). Therapeutic realizations are one of five universal, session-level change processes explicated in the Generic Model of Psychotherapy. This scale has four factor-analytically derived components: Remoralization, Understanding.
Unburdening,
Past-Focused
Insight,
and
Present-Focused
The Remoralization subscale is comprised of items that reflect a
renewed sense of optimism and positive affectivity, as exemplified by the specific therapeutic impacts of confidence, hope, enhanced self-control, reassurance, and encouragement. The Unburdening subscale captures the emotional-cognitive process of reflective self-expression, the experience of relief realized in interpersonal opportunities to verbalize troubling thoughts and feelings with a trusted listener.
Past-Focused
Insight refers to a type of learning that occurs in psychotherapy characterized by the realization of connections between temporally remote experiences (intrapersonal as well as interpersonal) and present feelings, thoughts, actions, and ways of relating. Finally, Present-Focused Understanding involves the acquisition of new knowledge, skills, attitudes, and ways of coping. Achievements of Therapeutic Objectives Scale (ATOS). Consonant with the need Greenberg and Foerster (1996) stressed to account for the absorption of treatment when evaluating the effects of different treatments on outcome, as I previously referred to, McCullough, Larsen, Schache, Andrews, & Kuhn (2003) developed a research tool, the Achievement of Therapeutic Objective Scale (ATOS) designed to evaluate the extent of therapeutic effects of therapy that the patient is absorbing or assimilating. The focus is on the achievement of treatment-specific objectives as opposed to a 54
theoretically defined idea of optimal outcome. Unlike the previous measures discussed, it is a measure of the absorption or ‘receipt’ by the patient of the therapy/interventions given by the therapist. It attempts to identify the adaptive shifts in patients’ behaviour that occur as a result of treatment, as seen from an observer perspective. The ATOS scale contains seven subscales that represent the main objectives of Short Term Dynamic Psychotherapy as well as the common factors in psychotherapy; Insight, motivation, exposure, new learning, inhibition, self perception, and alliance & relations (McCullough, Kuhn, Andrews, Valen, Hatch, & Osimo, 2004). Client Task Specific Change Measure Revised (CTSC-R).
Also in in an
attempt to demonstrate a link among in-session processes, postsession change, and posttherapy outcome, the CTSC-R is used to determine whether patients can differentially identify changes associated with specific interventions, providing a measure of postsession change, as opposed to a global evaluation of the session. CTSCR, as the name indicates is a revised version (Watson, Greenberg, Rice, & Gordon, 1998). It consists of sixteen 7-point items to identify changes that would be anticipated to occur in patient-centered (CC) psychotherapy (three items), PE psychotherapy (five items), and cognitive-behavioural psychotherapy (four items); the remaining four items indicated more general changes that might be engendered by any of the three approaches.
The Quest for Mechanisms of Change I started this review with the fact that psychotherapy is a remarkably effective healing practice. I then noted that common factors and principles may exist across
55
different theoretical orientations but common principles and change processes also exist across different disorders, i.e. across different patients. In an era of evidence-based practice, a large proportion of outcome variance remains to be explained, highlighting that factors remain hidden in a mysterious void of the psychotherapy process. Are there powerful healing factors that have not yet been identified? Do we lack the proper methods to accurately capture different dimensions of change as well as to accurately capture the mechanisms that promote those changes? Are there mechanisms that remain to be discovered and are they context-specific or common among approaches and patients? In fact, the mechanisms (or mediators) by which psychotherapy creates change are not well understood. Preferred efforts to address mediators of change come from studies of data generated as part of an RCT where there is already documentation that there has been meaningful beneficial change and that the change is not due to other variables (those controlled by the design used for the RCT). In my perspective, however, hypotheses about mechanisms of change can be derived from other sources including the common factors and principles among different therapies, for different disorders; from the common processes across disorders; from observations of therapists or researchers that have been consolidated into a possible mechanism as demonstrated in single-cases revealing links between theory, observed change processes, and final outcome; or even from other disciplines in the field of psychology. That is, in part, what I have been trying to illustrate with this review of the literature. Another this that motivates me in this review is to stress the need to address psychotherapy as it may be practiced in the real-world, increasingly of an integrative nature. Let me now review some efforts to address mediators of change coming from studies of data generated as part of RCT’s.
56
Major models of empirically supported psychotherapy do propose specific mechanisms of change responsible for the effects of the treatment packages. Despite these well-articulated models of change, there are few empirical studies demonstrating clearly that the proposed mechanisms account for the effects of each treatment. Further, there is very little evidence supporting that effective mechanisms of change are unique to specific treatment modalities rather than common to diverse psychotherapeutic approaches and techniques. Mechanisms of change of EST’s packages - approach and disorder specific? I will first examine several studies with proposals of several theoretical approaches, some integrative some single theory oriented, regarding mechanisms of change in borderline patients. I have chosen this group of papers, because there are yet not that many to choose from, due to the novelty of this research front, but also because these patients tend to be quite heterogeneous, a characteristic bearing more resemblance with patients from the real-world therapies as they are practised in naturalistic settings. Dialectical behavior therapy (DBT). Lynch, Chapman, Rosenthal, Kuo and Linehan (2006) outlined potential mechanisms underlying changes in the patient that may mediate the effects of certain DBT-specific interventions.
Considering the
intervention of mindfulness, the hypothesised mechanisms of change were behavioural exposure and learning new responses, emotion regulation, reducing literal belief in rules, and attentional control. Regarding validation as an intervention, they proposed as mechanisms of change, increasing the stability of self-views, reducing emotional arousal and enhancing learning, increasing motivation, modelling and contingency management.
In behavioural targeting and chain analysis as interventions, the
hypothesised mechanisms of change were, aversive contingencies, exposure and
57
response prevention, enhancing episodic memory, in vivo learning of skillful behaviour, generalization of behaviour change. Considering opposite action, the mechanisms were exposure and response prevention; broadening the patient’s repertoire and learning of new responses and cognitive modification.
Dialectics as an intervention had as
mechanisms, enhanced orienting responses and in vivo learning and modelling. Finally regarding the patient-therapist system as an intervention, the proposed mechanisms where reducing polarization and maintaining therapist efficacy. As the reader can notice, in this paper, Lynch, Chapman, Rosenthal, Kuo and Linehan (2006) are not always clear as to what constitute therapist operations and what are patient change processes or mechanisms of change. According to Chapman and Linehan, (2005), the strategies used in DBT to change borderline behavioral patterns can be further reduced to the following process: the reduction of ineffective action tendencies linked with (intense) dysregulated emotions, that is, encouraging the patient to act in a manner inconsistent with the action urges associated with dysregulated emotional states.
The primary mechanism
responsible for change in the indicators of treatment (final) outcome (i.e., reduced parasuicidal behaviour; improved social functioning; reduced depression, and hopelessness) involves changes in the behavioural components of the patient’s emotional responses (intermediate outcomes).
Improvement in emotion regulation
constitutes the mediator or mechanism of action in DBT, with emotion dysregulation including a set of specific behaviours that are directly targeted in treatment. Note, however, that DBT does not view emotion dysregulation as a construct per se, but rather as a set of specific behaviours (that comprise the domain of the emotion system and that can be observed and directly modified) to be targeted in treatment, like: (a) enhanced control over attentional focus or problem-solving when under distress, (b)
58
improved inhibition of urges to act in impulsive or self-destructive ways, and (c) increased regulatory control over emotion-related physiological responses. A recent measure was developed to assess skills training outcome. DBT theory states that BPD individuals suffer from a skills deficit, which leads to most of their problems. Therefore, skills use is hypothesized to be a mechanism of change. It will of course also be an outcome; skills training may lead to increase in frequency of skills use (outcome), and frequency of use may explain treatment changes (mediator) (Neacsiu, Rizvi, Vitaliano, Lynch, & Linehan, 2010). What comes out as the mediator or mechanism of change seems to be emotion regulation assessed or operationalized by optimal skills use, since, from a behavioral perspective, DBT cannot postulate skills use as deriving from a latent psychological variable. It is more like, does the person have the behavior in his or her repertoire and is the frequency of the behavior high enough and appropriate for the context? If the person does not have it, DBT teaches it. If they have it, DBT teaches when and how to use it. This behavioral expression is what is hypothesized to be the mechanism of change.
That is, if the person does emit skillful behavior (thoughts, emotions,
cognitions, sensing, and all overt behaviors) in crisis situations, will that make the crisis better? Neacsiu, Rizvi, Vitaliano, Lynch, and Linehan (2010) are just in the way of establishing the validity of the concept of skilful behaviour, providing empirical support for the DBT notion of behavioural change as a central mechanism of change. The new measure just developed is expected to be a useful research tool for answering important questions about mechanisms of change in treatment. Transference focused psychotherapy (TFP). In this model, the hypothesized mechanisms of change derive from a developmentally based theory of BPD, which conceptualizes the disorder in terms of unintegrated and undifferentiated affects and 59
representations of self and other, whereby BPD patients are said to have a difficulty integrating them.
The lack of integration corresponds to a “split” psychological
structure in which totally negative representations are split off from idealized positive representations of self and other. The putative global mechanism of change is patient’s increase in reflection resulting from/in the exploration and integration of these polarized affect states and representations of self and other into a more coherent whole (Levy, Clarkin, Yeomans, Scott, Wasserman & Kernberg, 2006). Therapist’s interventions or operations begin with the structured treatment approach (e.g., the use of a treatment manual, treatment contract, hierarchy of problems addressed, and group supervision for therapists) and the use of the triad of clarifications, confrontations and interpretations. It is expected that the therapist through an objective and nonjudgmental attitude (a) manages to convey a message that he or she can tolerate the patient’s negative affective states as they are reactivated in session and, (b) expects the patient to increase his or her ability to have a thoughtful and disciplined approach to emotional states, as they are regulated and explored for new understandings in session (Levy et al., 2006). The increase in reflectiveness involves two levels. The first level entails an articulation and reflection of what one feels in the moment. The patient increases in his or her ability to experience, articulate, and contain an affect and to contextualize it in the moment. A second, more advanced level of reflection is the ability to place the comparison and understanding of momentary affect states of self and others into a general context of a relationship between self and others across time, in perspective. These different contradictory states are highlighted and attributions are sought as to why they have remained split off (Levy et al., 2006). In order to assess the above described integration and reflectiveness of representation a recent self-report measure
60
(Multidimensional Measure of Reflective Function) is currently being developed (Hill, Levy, & Meehan, 2005), whereby the therapist is asked about the reflective capacities of their patients. This work is, however, very preliminary and Levi and colaborators were not able to share the final measure before the completion of this review. Mentalization based therapy (MBT). Fonagy and Bateman (2006) assume that BPD patients have a limited capacity to mentalize, that is, to comprehend and use their knowledge of their own and others’ states of mind (thoughts, feelings, wishes, and desires). They propose that a constitutionally vulnerable individual, who experiences developmental trauma in an attachment context, becomes psychologically vulnerable in later attachment contexts as a result of instability of the self. In an attempt to cope, the individual, with a fragile capacity to represent affect and effortfully control their attentional capacities, dissociates the mind from others’ minds and relies on earlier psychological mechanisms to organize the experience and in doing so reveals fragments of the self. In MBT the patient’s mentalizing (the experience of another human being’s having the patient’s mind in mind) is seen as a critical facet of the therapeutic relationship and the essence of the mechanism of change. The patient is encouraged to (a) counteract the normal pattern of attachment related deactivation of mentalizing of negative emotions and social and moral judgments, and (b) not to relinquish mentalization at the slightest suggestion of attachment-related brain activation (Fonagy & Bateman, 2006). In terms or therapist’s operations in MBT, as stressed by Fonagy and Bateman (2006), (1) the therapist focuses on patients’ current mental state with the aim of building up representations of them; (2) avoids situations in which the patient talks of mental states that he or she cannot link to subjectively felt reality; (3) the therapy creates a transitional area of relatedness in which thoughts and emotions can be “played 61
with.”; (4) the enactments over the course of the treatment are not interpreted or understood in terms of their unconscious meaning but in terms of the situation and affects that preceded them. These attempts to enhance mentalization are made in the context of an attachment relationship. The attachment system is supposedly activated through (1) the discussion of current/past attachment relationships, (2) the therapist’s encouragement and regulation of the patient’s attachment bond to him or her through the creation of an environment that assists with the patient’s regulation of affect, (3) the therapist’s attempt to engender attachment bonds between the patient and others. The therapist must balance the intensity of attachment relationships and the complexity of mentalization required of the patient. According to Fonagy and Bateman (2006) the therapist focuses on the patient’s mentalization on relationships that have relatively low levels of involvement and only gradually to focus the patient’s thinking on relationships closer to the patient’s core self. Similarly the tasks of mentalization vary in demand characteristics, with clarification at the most superficial end and exploring of repudiated intense emotions in relation to the attachment figure at the more complex end. Schema focused therapy (SFT). In SFT patients with BPD are conceptualized as being under the sway of five aspects of the self (schema modes) and the goal of the therapy is to reorganize this inner structure. Kellogg and Young (2006) state that although most patients who have personality disorders are trapped in a rigid style that is amenable to schema work, BPD patients are trapped in a state of flux, with a rapidity of emotional change, often from adoration to hatred.
Since many of these patients
frequently endorsed the top range of almost every item on the schema inventories, the solution was to understand that these patients were “flipping” through clusters of schemas and coping styles and that it would be more useful to envision the patient’s personality as consisting of various modes or different aspects of the self. 62
There are five central modes in the borderline constellation: (1) the abandoned and abused child, (2) the angry and impulsive child, (3) the detached protector, (4) the punitive parent, and (5) the healthy adult modes. The development of the healthy adult is one of the goals of the therapy, and it is typically first embodied in the therapist and then, through the therapy process, internalized by the patient (Kellogg & Young, 2006). From my perspective, the conceptual division Kellogg and Young (2006) make between mechanisms of change and phases of treatment is not at all clear because they hypothesise four core mechanisms of change that are at the core of schema treatment: (1) limited reparenting, (2) experiential imagery and dialogue work, (3) cognitive restructuring and education, and (4) behavioral pattern breaking. These interventions are used during the three phases of treatment: (1) bonding and emotional regulation, (2) schema mode change, and (3) development of autonomy. Their proposal for mechanisms of change derives most exclusively from the technical or strategic level from the perspective of therapists operations. However, my personal clinical reading of this proposal, with my integrative lens, reveals some hints at mechanisms like self-validation regarding needs and emotions, emotion regulation, mentalization or reflective functioning, meaning attribution regarding difficult past and present situations, increase in problem solving skills, agency regulation and strengthening the development of a sense of identity in daily living.
Even the
conceptualization in terms of modes suggest them as core higher level schemas, self wounds or parts of the self whose harmony or dialogue is to be sought, first by becoming aware of them, knowing them in action, linking and explaining them, making them more flexible (activating, deactivating, softening and/or strengthening them) and promoting negotiation between them through agency regulation, and putting them
63
somehow at the service of emotion, need and self-regulation towards the development of a sense of self and identity. Cognitive therapy (CT). According to CT, BPD patients are characterized by dysfunctional beliefs that are relatively enduring and inflexible and that lead to cognitive distortions such as dichotomous thinking. When these beliefs are activated, they lead to extreme emotional and behavioral reactions, which provide additional confirmation for the beliefs. BPD is conceptualized as vicious circles of distorted cognition, negative affect, strong impulses, and maladaptive behavioral responses that confirm or perpetuate the original distorted beliefs. These circles create a sense of hopelessness and helplessness (Wenzel, Chapman, Newman, Beck, & Brown, 2006). In terms of therapists operations, the systematic evaluation of cognitive distortions is supposed to reduce negative affect, thereby allowing patients to select appropriate behavioral responses in a more thoughtful way and to practice the skills learned in therapy. CT also helps patients to develop and implement new behaviors to negotiate interpersonal conflict and manage personal distress. Behavioral experiments involve the formulation of specific, testable hypotheses about the patient’s behavior with the aim of gathering evidence for the evaluation and refinement of underlying beliefs. These experiments are used as methods to test currently held beliefs and predictions and translate newly developed beliefs into behavioral change. The need for the development of a conceptual road map that provides a sense of context, comprehensiveness, and direction for the development of a treatment plan and treatment goals, in other words, structure, is also of paramount importance to CT (Wenzel et al., 2006). Like in cognitive therapy for other disorders, Wenzel et al. (2006) hypothesize that a change in dysfunctional beliefs is the primary mechanism of change associated 64
with CT, and suggest additional mechanisms including enhancement of skills, reduction in hopelessness, and improvement in attitude toward treatment. The development of problem solving skills is crucial to treatment success and should also be studied systematically as a mechanism of change. The process of problem and goal definition, generation of alternative responses, evaluation of the alternatives, and implementation and evaluation of the selected strategy equips patients with an adaptive behavioral repertoire for coping with stress.
Relationship enhancement skills, such as anger
management, effective communication, and assertiveness, are also associated with improvements in interpersonal relationships and decreases in mood fluctuation related to interpersonal conflict. They also speculate that reduction in hopelessness and greater positive expectancies for treatment as mechanisms of change associated with CT for BPD. Interpersonal psychotherapy (IPT). Markowitz, Skodol and Bleiberg (2006) acknowledge the way Interpersonal Psychotherapy (IPT) works is unknown, in general and specifically with BPD patients.
They depart from usual IPT to speculate on
potential mechanisms with BPD, however, in my opinion they also interchangeably refer to therapist’s operations or patient’s change processes or mechanisms, without differentiating them. In terms of therapists operations, IPT is said to employ common factors starting with a structured approach that prevent the dyad from becoming discouraged and disorganized. Therapists also help patients to feel understood, and provide support, expertise, a rationale for improvement, therapeutic optimism and encourage success experiences.
BPD is reconceptualized within an illness model, as a chronic and
treatable illness, that is, a lifelong personality as an alien, removable syndrome, as this is taken to presumably comfort most patients and provide hope that remolding lifelong 65
character might be possible.
Keeping the focus of treatment out of the room
(extraoffice focus), the focus on those relationships outside the office, may deemphasize the therapeutic relationship and increase the importance of solving “real life” interpersonal relationship problems and may decrease the tendency for therapeutic ruptures, increasing the likelihood of the therapy’s continuing and the patient’s improving (Markowitz et al., 2006). From my readings what might stand as the predominant mechanism of change in IPT, is the learning of new skills.
In the common factor "Success Experiences"
mentioned by Markowitz et al. (2006) it is hypothesized that through conquering a life crisis (complicated bereavement, role dispute, etc.), patients learn new skills in the process, and improve not only their life but also their mood, bringing control to their lives and improved self-image. The focus on rehearsal for the patient’s gaining such successes and the emphasis on what the patient does between sessions ensures that the patient owns his or her successes.
Although IPT makes no claims to change
personality, interpersonal skills (assertiveness, effective expression emotions) frequently open up new possibilities for interpersonal functioning and consequent personality change. Markowitz and collaborators have developed an Interpersonal Psychotherapy Outcome Scale to measure how much interpersonal change the patient and therapist feel the patient has accomplished in work on the focal problem area. The IPOS, as its title suggests, is, however, really more of an outcome than a process instrument. It basically asks the patient which focal interpersonal psychotherapy (IPT) area(s) their therapy focused on. It's a rather simple instrument assessing patient-perceived change in the focal problem area that then can be correlated with final improvement.
66
From my reading of these papers, it is not obvious that the mechanisms of change are specific of the therapeutic approach, nor of the disorder analysed. Independent from the differences in terms of therapists operations, all seem to promote a more hopeful, resilient, agentic and connected self, with more structure, more selfacceptance, increasing awareness and curiosity, more capacity to experience and process diverse psychological self-states, more able to reflect on that experience (tolerate and mentalize) and make new meanings, and more able to implement the skills needed to satisfy his or her needs in life. Probably all approaches arouse the attachment system at the same time it applies intra and interpersonal demands, although some do it more explicitly as in the following continuum from left to right MBT-TFT-DBT-SFTCT-IPT. The first thing that came to my mind when reviewing these papers on mechanisms of change with BPD, was that there seemed to be little order among what could be different levels of analysis regarding mechanisms of change. Actually, I had to spend some resources teasing apart what appeared to be clear confusions between therapeutic technique and/or strategy (therapeutic operations) and patient change processes or mechanisms. More often than not, the levels of analysis were mixed by the authors. Etiological mechanisms, disorder maintenance mechanisms, therapeutic action or operation strategies/processes/interventions/mechanisms, patient change processes resulting from the absorption of that therapeutic acts, mechanisms of change resulting from the accumulation of those change processes, and the translation of those mechanisms in behavioral terms by the patient, are all but different parts that need to be differentiated and operationalized if one wants to advance in this field of psychotherapeutic science. In the near future, it could be very interesting to conduct
67
detailed process studies using the materials from these empirically supported treatments by investigators without theoretical blinders. Just before finishing this literature review, and stating the purpose of my research project, let be leave BPD as, nonetheless, a specific disorder, and illustrate my point with another interesting study. Empirically found common mechanisms of change across approaches. Connolly Gibbons, Crits-Christoph, Barber, Stirman, Gallop, Goldstein, Temes and Ring-Kurtz (2009) examined a set of mechanisms proposed as important mechanisms of change within a pooled data set comprising a range of therapeutic modalities, including cognitive
therapies,
dynamic
psychotherapies,
and
other
supportive
control
psychotherapies. They were interested in validating the role of each mechanism within specific treatment modalities and, furthermore, in examining whether these mechanisms were unique or common to diverse psychotherapies. They selected three mechanisms of change, including self-understanding of interpersonal patterns to represent the mechanism of change espoused by dynamic models of psychotherapy, compensatory skills (or coping skills) to represent one of the major mechanisms of change described in cognitive models of psychotherapy, and views of the self to represent a mechanism of change common across diverse psychotherapeutic models. By changes in self-understanding Connolly Gibbons et al. (2009) were refering to changes in self-understanding of interpersonal patterns, in other words, acquiring understanding of the repetitive maladaptive impairing relationship conflicts. A closer look of mine at the measure used to assess this mechanism, reveals the items try to capture increase in awareness, new meaning making and regulation of responsibility and needs, interpersonal needs. The fact that mostly distal causal attributions, in terms of meaning making, are measured by this scale might, increase the probability of it 68
being "psychodynamic-specific". The clear interpersonal taste of the items may also help the measure differentiate psychodynamic from cognitive approaches that tend to be more intrapersonal.
However increase in awareness, new meaning making and
regulation of responsibility and needs are certainly not specific of a psychodynamic approach. Compensatory skills implies the acquisition of compensatory or cognitive coping skills to deal with distressing events and thoughts, including the generation of initial explanations and then alternative explanations for negative internal and external events as well as the creation of concrete problem-solving plans and selection of appropriate behavioral responses to resolve difficult situations. A closer look at the measure used by Connolly Gibbons et al. (2009) to assess this mechanism, reveals that the items mostly capture meaning making constructs, which might render it a measure of a common mechanism. Despite the language of the measure is too cognitively flavoured, cognitive flexibility and problem solving might me common mechanisms among different approaches. By change in views of the self Connolly Gibbons et al. (2009) were refering to self-discrepancys since it had been found that three psychotherapy treatments were all associated with decreased self-discrepancy. A closer look at the measure used to assess this mechanism leads me to agree with the hypothesis of if being a common mechanism. Connolly Gibbons et al. (2009) explored these three theory-driven mechanisms of change concurrently across the cognitive and dynamic psychotherapies by first examining whether any of these mechanisms changed significantly across treatment and whether the changes were specific to the models of psychotherapy from which each mechanism was derived. Next, they examined whether change in each mechanism
69
predicted change in symptoms across cognitive, dynamic, and other psychotherapies. To unravel the temporal course of change in mechanism variables and outcome variables, they further examined whether change in the mechanism variables predicted follow-up symptom course, controlling for change in symptoms across the treatment. They hypothesized that self-understanding would change significantly more in the dynamic psychotherapies compared with the cognitive therapy condition, whereas compensatory skills would change significantly more in the cognitive treatments compared with the dynamic psychotherapies.
Additionally, they hypothesized that
significant changes in self-concept would occur across both cognitive and dynamic treatments.
As hypothesized, self-understanding changed significantly more in the
dynamic psychotherapies, than in cognitive therapies. Changes in compensatory skills and self-concept, however, were common across the dynamic and cognitive psychotherapies examined. Finally, Connolly Gibbons et al. (2009) hypothesized that each of these mechanisms would remain a significant predictor of symptom reduction when controlling for the other mechanisms examined. When all three mechanism variables were entered simultaneously in regression analyses, the “action” was carried by changes in compensatory skills and views of the self; changes in self-understanding were no longer significant, once changes in the other mechanism variables were controlled. These results indicate that changes in compensatory skills and views of the self are the important
mechanisms
of
change
driving
symptom
course
across
diverse
psychotherapies. Decreases in the use of negative coping skills and improvements in how individuals view themselves appear to be important changes that drive symptom course across diverse psychotherapeutic approaches.
70
The results indicate that self-understanding of interpersonal patterns is not a significant mechanism of therapeutic action, once the effects of changes in compensatory skills and views of the self are controlled, however, as Connolly Gibbons et al. (2009) suggest, this might be due to the conceptual overlap between the constructs of self-understanding and compensatory skills, an explanation and a result that paves the way for integrative theoretically relevant mechanism variables and measures to measure them. I will finally announce the purpose of this research project. Purpose of this Research Project As reviewed previously and earlier on (Vasco, 2001; 2006; Conceição, 2005) the notion that psychotherapy proceeds in stages is a common heuristic tool that transcends theoretical orientation. In fact, the construal of change within the psychotherapeutic process as a process involving progress through a series of stages, where stage is the temporal dimension that represents when particular changes within the patient occur or within the therapeutic process occur (therapist, patient, interaction), is certainly not new. Somehow, these conceptualizations try to impose some structure onto the complexity of change and developmental processes. All can serve as guides that help therapists to select interventions by defining the developmental tasks or change processes that need to be accomplished at each stage. As I stressed in the introduction, therapists and patients can be responsive to each other on time scales that range from months to milliseconds (Stiles, Honos-Webb, & Surko, 1998).
The construct of stage or even phase ends up being used
interchangeably by each author with regard to the time scale, be it months or milliseconds. This state of affairs calls for a better differentiation between levels of intervention or change as initially stressed by Vasco (2006) and further developed in Conceicao (2005). This is not to say that decisions taken during larger periods of time, 71
do not have their millisecond, second or minutes counterparts. Indeed the idea is that moment-by-moment decisions can, or perhaps even should, be influenced, prescriptively or proscriptively by phase-by-phase decisions. On the one hand, I believe that these two layers of decision making deserve a better differentiation, and on the other hand, phase-by-phase responsiveness deserves empirical research demonstrating its potential usefulness. If we go back to the less clear state of affairs with regard to the usage of the construct stage or phase, another problem emerges for me, namely that more frequently than not, it must be tied to a content or theme. In his formulation of Transtheoretical Model (TTM), Prochaska (1979), being perhaps the first and most systematic to focus on the level of strategies, had already sought to find compatible matches among several cross-theoretical constructs: (1) levels of change, (2) readiness for change, and (3) strategies for change. The TTM postulated that ten processes of change are different along a temporal dimension as if they were defining the sequence of the stages. I always saw great utility in the conceptualization of a stage such as precontemplation as belonging to a process of change, in that it helped many therapists to reformulate their notions of resistance. Besides, Prochaska’s important contribution was the idea that one's readiness for change signals the potential the effectiveness of strategies used to promote change.
Interestingly, this use of strategies could be
implemented by a change/developmental agent (like a psychotherapist, psychologist, trainer, educator) or self-implemented.
This dual possibility, for me, was already
pointing in the direction that there might be something in the patient that needs to change first, before the next changes are allowed. My greatest dissatisfaction with TTM, however, was that in order to apply it to in my psychotherapeutic practice, that
72
rarely is short-term, it would have to be anchored in a specific problem or goal or theme, like “give up smoking” or “increase jogging to reduce my belly”. I then noticed that the same would happen with other stage models like Assimilation of Problematic Experiences or the the Seven-Phase Model (Meier & Boivin, 2000) but the nature of the content to which the stage was tied was richer in psychological density. Both are developmental frameworks that describe how themes change, diverge or converge.
The respective methodological tools, Assimilation
Analysis and Theme Analysis, both track themes intensively, so that themes’ qualitative change processes and interrelation with each other can be studied (Meier & Boivin, 2000; Knobloch, Endres, Stiles & Silberschatz, 2001). The same issue would be present even in interesting adaptation of these original models, where the content to which the stage was tied was not only richer in psychological density but could also become more comprehensive in scope. Livesley (2003), for example, after minor modifications and keeping a four-stage structure applied the TTM to any levels or problems of the therapeutic process, the time spent (the pacing) on each stage being the variable that changes. Rudolf, Grande, Dilg, Jakobsen, Keller & Oberbracht (2002), slightly changed the Assimilation Stages to study problems at higher-level namely, structural changes. They described therapeutic progress as a function of the structural changes in patients’ modes of coming to terms with so called ‘foci’ – the most important areas causing and sustaining patients’ disorders, like internal conflicts, structural vulnerabilities, maladaptive relationship patterns (Rudolf, Grande, & Oberbracht, 2000). Finally, Benjamin (2003) also adopting a four stage structure based on the TTM described a possible way of working on one of her five proposed steps or stages of the therapeutic process, namely, enabling the will to change (the gift of love), a higher level
73
of change. The thesis of Benjamin (1993) is that every psychopathology is a gift of love - destruction of self or others is done out of love and accompanied by the wish to receive love and are motivated by a wish to please important attachment figures. By behaving in ways consistent with the rules and values of a beloved early figure (or figures), patients seek psychic proximity to the internalized representations of those figures, much as a toddler returns to its caregivers when frightened (Benjamin 2003). Without wanting to enter in great detail about Benjamin’s great thesis, the point I would like to stress is the nature of change sought under her conceptualization, namely, structural personality change and the language that change is described, in terms of processes, not that much content. Rudolf et al (2000; 2002) are certainly also concerned with structural personality change, but their conceptualization is still considerably anchored in (psychodynamic) theory driven content. Structural change is usually not revealed until there have been many therapy sessions, i.e. it may take a while to emerge - therefore the need for a map, and preferably, a map of strategy related patient change processes and mechanisms and strategy related therapists operations to foster that development in the patient. These strategy related change processes or mechanisms of the patient ideally, from my perspective, are not to be anchored in any content or theme, but to simply define a general capacity to process experience in different levels of complexity and with a great generalisation potential to the different areas of his or her life, as he or she will progress in therapy. A better understanding of the important mechanisms of therapeutic change as theoretically proposed or empirically demonstrated provides a good opportunity for further improving the effects of integrative approaches and real-wold therapy in general. According to my literature review, research efforts on mechanisms of change seem to
74
beget sound integrative theory grounded on plain psychological processes written in ordinary English. And research on such general strategies that could be good canditates for mechanisms of change is lacking, as well as are empirical demonstrations of possible sequential organizations of them.
Kazdin (2007) explicitly stressed the
importance of demonstrating mechanisms to precede final outcome, but what what precedes or anticipates final outcome, i.e. intermediate outcomes, may need order. Some mechanisms may have to be there first, before others can follow. Connolly Gibbons et al. (2009) stated that additional research is needed to further unravel the role of specific interventions that produce changes in mechanism variables. But my guess is that if, at the technical level, diverse therapeutic techniques are capable of producing changes in common mechanisms, (like compensatory skills and conceptions of the self, in their study), at the strategic level, it is the promotion of general strategies or strategic objectives common among diverse psychotherapeutic approaches that is responsible for changes in these important mechanisms. This is to say that, besides the need for conceptualizations of common mechanisms of change, research is needed to unravel the role of therapist’s general common strategy related operations that produce changes in patient’s general common strategy related mechanism variables. If these common mechanism variables influence or guide clinical decision making at the strategic level of therapist’s operation, or vice versa, also remains a research question (Conceição, 2005). Aiming at contributing for such research, I therefore chose the theoretical proposal from Paradigmatic Complementarity (Vasco, 2001) that bears some resemblance with others reviewed before (e.g. the 4 Stage Learning Model; Goldfried, 2006; Beitman, 2006, Benjamin, 2003). According to Vasco, (2001; 2006) therapy is understood as a sequence of (seven) clusters of general strategies (see figure 1).
75
I will try to investigate the usefulness of such a proposal for the recent generic call for research on mechanisms of change, and in doing so, I will adopt an integrative perspective. By integrative perspective, I mean two things that mirror the two goals underlying the formation of the Society for the Exploration of Psychotherapy Integration (SEPI) when first formed in 1983: dialogue across the different theoretical orientations and dialogue between research and practice. As others in the field, I believe psychotherapy researchers can use clinical practice as a natural laboratory for identifying promising elements relevant to good process (Westen & Bradley, 2005). The fact that research community seems to be recognizing that examining aggregate effect is not the only empirical window researchers have on the nature of change and that systematic observation of one or a few patients can be scientifically sound and instructive, led me to combine cross-sectional and longitudinal single-case methods, all within naturalistic settings. Naturalistic studies by themselves may not permit strong causal conclusions, yet such evidence can be enormously helpful. The general purpose of the present study is to address the paucity of research on the temporal sequencing of strategy related patient change processes and mechanisms and strategy related therapists operations, in terms of psychological processes not anchored in specific contents. Evidence will be sought for the arguments that: a) the dimensions of strategy related patient change processes may count as mechanisms of change leading to final outcome and, b) responsiveness in terms of strategy related therapists operations may count as a general principle of change. The first objective of this project is the verification of a phase model of change that bridges the start and finish with specified intermediate phases. As previously noted, this model is described with 7 phases, but any alternative model that respects the sequence of the strategies involved in the original model would also be tested against it.
76
We had already collected evidence with regard to the temporal sequence of strategic objectives as promoted by therapists (Vasco, 2006), and evidence with regard to a sequential development of capacities or intermediate outcomes by the patient, that is strategy related patient change processes or mechanisms of change (Conceicao, 2005). In the first part of this research project I will further the previous studies and try to replicate the data stemming from them, to further validate them, and in doing so, I chose a cross sectional design with a new, larger sample, using improved instrumentation. In the second part I will look at 4 long-term longitudinal cases, with one of them particularly intensively, and examine what experiences patients and therapist are actually having in the session with regard to processing general strategies. I will seek to evaluate and refine the proposal made in Paradigmatic Complementarity of sequencing of strategic objectives, to allow for a richer understanding of how patients develop in psychotherapy and how therapists can best facilitate that development. The problem under investigation is how patients develop in terms of processing capacity in the event of undergoing individual psychotherapy and how therapists work during that development, both at the level of general strategies. How patients are able to process the information (strategy related change processes or mechanisms) and what kind of processing is the therapist stimulating (strategy related operations) will be under close scrutiny. Specifically, the following questions will be investigated: - Cross-Sectional Study Research Questions (Hypotheses) i)
Descriptive statistics and internal consistency analysis will reveal that strategyrelated therapist's operations and patient's change processes can be operationalized in a mirrored format and measured at the item level and at the scale level, according to the proposal within Paradigmatic Complementarity, of 7-phase model
77
of the therapeutic process. ii)
The list of strategy related therapist’s operations as well as the list of strategy related patient change processes are sequentially ordered at the item level.
iii) An abbreviated version of the General Strategies Inventory will be derived empirically, demonstrating acceptable reliability and validity as a postsession measure of strategy related change and operations for psychotherapy of any approach. Its structure will respect the theoretically proposed sequence, even if with less dimensions than the original 7. iv) The empirically derived dimensions of strategy related patient change processes and strategy related therapist operations will be used to replicate results from previous studies with the original 7 dimensions: (a) they will vary at different points in therapy in agreement with the postulated sequence, (b) they will relate to other process and outcome variables, (c) they will relate to each other as to suggest some kind of responsiveness at this level of intervention v)
The predominance of the different types of therapeutic work varies along the time of the process and can be used as a measure of time, that goes beyond the number of sessions, that is not as informative as compared to this construct “phase of therapy” or phase of predominant strategy-related operations being pursued by the therapist.
vi)
Intermediate Outcomes (strategy-related change processes or even mechanisms) that are expected to result from strategy-related operations, will predict changes in final outcome (estimate of improvement), making a unique contribution over and above the therapeutic alliance, and will do so according to the postulated sequence.
- Longitudinal Case Studies Research Questions (Hypotheses)
78
vii) When tested in two groups of raters, the General Strategies Inventory-Observer Form using the seven original dimensions will demonstrate good interrater reliability (within-group) viii) Strategy-related change processes of the patients will unfold according to sequence theoretically postulated in Paradigmatic Complementarity replicating the pattern found in the cross-sectional study ix) High levels of patient's strategy-related change processes (representing structural change or structural capacities) will be achieved sequentially, with the next following the previous structural processing capacity, by order. x)
Therapist's strategy-related operations will be more scattered along time but a predominancy of implementation will be verifiable and the unfolding of that predominance will follow the postulated sequence
xi) Decision making of the therapist will be responsive to patient's strategy-related change processes and to patient's structural change or structural capacities of each dimension. Responsiveness will be noted when: (a) the therapist predominantly promotes operations from a certain dimension until the patient reaches structural change's criterion on that same dimension, (b) the therapist will let go of predominantly pursuing strategy-related operations from a certain dimension, when patient's strategy-related change processes from that same dimension have reached structural change's criterion, and (b) the therapist will move on to predominantly promote strategy related operations of the next dimension in the sequence xii) All 4 cases will demonstrate significant increase in the patient's strategy-related change processes of regulating responsibility and implementing repairing actions, both known to correlate with final outcome. In one case, a conventional outcome measure will complement this effectiveness demonstration.
79
xiii) The time a patient needs to reach structural change criterion in a certain capacity will demonstrate that the phenomenon being studied takes months to build up, rendering it more a phase-by-phase instead of a moment-by-moment phenomenon. xiv) A sequential test of mechanisms of change will demonstrate that patient change in a previous dimension will be followed by change in the adjacent dimension, weeks before change in the other second next dimension occurs.
80
STUDY 1: CROSS SECTIONAL EXPLORATION AND REPLICATION PHASE
This study consisted of an attempt to replicate some results from previous studies that also used previous versions of a transtheoretical measure, the General Strategies Inventory, to assess the promotion of those general strategies by the therapist (therapist's strategy-related operations) and the achievement of capacities related to the absorption, by the patient, of those strategies (patient's strategy-related change processes). We needed a measure that could be used across perspectives, therapist, patient, and observer, with the same items: (a) first, a version to assess the perspective of the therapist regarding the promotion and the capacities; (b) second, a version to assess the perspective of the patient regarding his or her capacities and (c) a a version to assess the promotion and the capacities from the perspective of observers. The third perspective was particularly important, since it would used in the second phase of research, the longitudinal study. This study is mostly psychometric, correlational and also of the type “sampling from natural variation” (Grawe, 1999). It is cross-sectional, descriptive and quantitative (Hill & Lambert, 2004). As such it has limitations regarding internal validity, since it does not imply real experimental manipulation of variables. Regarding external validity it is a naturalistic study. However, the sample is not necessarily representative of the Portuguese population of participants, so generalization of results remains conditioned.
81
Method
Participants Participants were 227 therapist-patient dyads recruited either through personal contact with the author or through listservs associated with clinical psychology and psychotherapy. Therapists filled in questionnaires either in paper or online. In the paper version, therapists would invite their patients to complete their versions of the questionnaires.
Uncomplete questionnaires or any refusal to participate was not
controlled, but simply not included. Data was collected with participants in individual therapy in private and institutional settings, from several places of Portugal, mostly from Lisbon, Porto and Coimbra, but also from Braga, Évora and Faro. Half of the data was collected in 2008 and the other half in 2010. After cheking several parameters, both samples were combined into a single, larger one. Patients were seen in the context of adult individual psychotherapy. In order to avoid biases from literacy level and altered states of consciousness, exclusion criteria were: patients with psychotic features or with substance abuse and less then the 11th grade of schooling. Some therapists submitted more than one patient of their case load to the sample (but never more than 6; and only 5 therapists submitted more than 4), but they were not supposed to hand in two protocols of the same patient, even if at different times. Therapist effects were, therefore, not controlled. Incomplete or invalid protocols were also dismissed without being subject to any kind of analysis. Since there could be a tendency for the respondents to be those motivated to participate both in research and in therapy, therapists were asked to to submit also some protocol regarding patients with
82
whom they were having difficulties, a variable that was controlled with a specifically designed measure. Mean session number is 32 (SD = 52). From the 227 cases, 15 did not enter data regarding session number, while 42 (18,5%) were between the first and fifth session, 49 (21,6%) between the sixth and 12th session, 46 (20,3%) between the 13th and 24th session, 44 (19,4%) between the 25th and 50th and 31 (13,7%) had more than 50 sessions. In the global sample of the 227 patients, 16 did not enter data relative to sex and 72 (31,7 %) were male and 139 (61,2%) female, comprising ages from 18 to 66 (M = 33,5 years ; SD = 11,17; 24 did not enter data relative to age). Regarding Literacy level, out of the 227 only 91 were asked about it (the digital sample was spared). From these, 33 (36.2%) had bellow BA 31 (34%) had BA and 4 (4,4%) above BA. Regarding patient pathological features, 53 (23,3%) were assessed by their therapists as having only a depressive disorder, 52 (22,9%) only an anxious disorder, 28 (12,3%) only a personality disorder, 41 (18,1%) comorbidity among Axis I disorders; and 38 (16,7 %) comorbidity among Axis I and II disorders. In 15 cases there was no data regarding this variable. Independently from comorbidity, depression was present in 127 cases, anxiety in 108 and personality disorder in 67. Out of the 227 cases, 131 (57,7%) had female therapists and 43 (18,9%) male therapists. Mean age of the therapists was 35,4 years, between 22 and 67 years (SD = 7.7). Thirthy three (14,4%) patients had therapists with less than 4 years of clinical experience, 47 (20,7%) had therapists with an experience level ranging from 4 to 6 years, 48 (21,1%) had therapists with an experience level ranging from 7 to 9 years, 50 (22%) had therapists with an experience level ranging from 10 to 15 years, and 43
83
(18,9%) had therapists with more than 15 years of clinical experience. In 6 cases we had no information regarding experience level. The way data was collected regarding theoretical orientation prevented me to specify what specific approach therapists followed. Indeed, all cases of therapists reported guiding their practice according to more than one orientation, from the available options: cognitive, behavioural, humanistic/experiential, psychodynamic/ analytic, systemic, and integrative or other. This does not necessarily mean that the sample is a truly integrative one, yet one can say that no pure-form theoretical schools exist in this real world naturalistic sample. This finding is consistent with the literature (Orlinsky & Ronnestad, 2005). Instruments Development of the General Strategies Inventory (GSI). The version of the GSI used in this study was preceded by two previous versions, which development I summarize here. The initial items have been written in theory neutral terminology to capture common factors or processes that underlie many forms of therapy. It was sought to avoid items dealing explicitly with therapeutic technique. The items do not focus on change in specific content areas, but are phrased as generally related to feelings, thoughts, behaviours, and attitudes. Items were supposed to reflect strategies at a midlevel of abstraction (Goldfried, 1980) as reviewed in the introduction. Item Generation. The instrument was generated through a series of sequential ratings and evaluations of the prospective items. In the year 2000, an initial pool of 160 items was generated by the author of this dissertation on the basis of clinical and theoretical discussions with António Branco
84
Vasco, and a content analysis of integrative psychotherapy literature (Journal of Psychotherapy Integration and Handbook of Psychotherapy Integration). Vasco and Conceicao (2001) had chosen to develop the patient form first aiming at maximizing their efforts to minimize clinical jargon and anchor the measure in plain language, planning to later develop the therapist form on the basis of this emerging patient form. Rating by Professionals. Each item on the pool was designed to capture a behaviour (internal) in the patients field of awareness that may be present or absent during a session. Four psychotherapists with an integrative theoretical perspective, working within the framework of the model under study, evaluated these items to reduce conceptual and linguistic bias. Two questions were asked of the raters with respect to each item: (a) “Is this item relevant to the therapeutic process?”; and (b) “Which of the seven dimensions of general strategies of the therapeutic process does this item refer to?”. The relevance of each potential item to the concept of the stage of general strategies, according to conceptualization that we supplied, was rated on a 5-point Likert Scale. A rating of 1 indicated that the item was not related to the process and a rating of 5 indicated that the item was very relevant to the process. Items with a mean relevance rating less than 4.0 were eliminated from the pool.
In addition, raters
classified each item as referencing one of the seven dimensions proposed by Vasco (2001). A percentage of agreement (PA) index was calculated for each item. Items with a PA index of less than 70 % were rejected. 80 items were eliminated from the pool. The first version of GSI ended up being composed by by 64 items distributed in 7 subscales with 11 or 12 items each and the items, from the perspective of the therapist and the patient. For example, the item “In this session I have been increasing my awareness of how I relate to others” is considered as parallel to the item “In this session 85
I tried to help my patient increase his or her awareness of how he or she relates to others”. In a first study with this first version (published in Vasco, 2006) internal consistency ranged from acceptable to good (.89, .62, .74, .61, .78, .86, and .87 from dimension 1 to dimension 7 respectively). By 2005, on a second version of the inventory, several items were removed and others were edited as a result several meetings within the research team taking into consideration previous statistical analysis. The team also reconsidered what was being asked from the patients, because of their difficulty diferentiating the work being done at the session, probably because the items were still not simple enough, or because they have more globalistic impressions of the therapeutic work being done. As a result of this reconsideration, it was proposed to ask the patients about how capable they felt regarding this or that kind of capacities. It was thus possible to have a mirrored measure to assess therapeutic operations (what the therapist promotes) and intermediate results (what the patient is capable of). The second version of GSI ended up being composed by 64 items distributed in 7 subscales also with 8 or 11 items each, in the propomtioon version and 4 items each in the capacities version. Cronbach alphas for the therapist version of promotion subscales were .88, .85, .79, .83, .81, .85 e .79, respectively, while for the capacity subscales were .85, .85, .84, .86, .92, .93, .82, respectively. Relative to other process variables, the capacities subscales behaved possibly like (intermediate) outcome scales in that they correlated significantly and considerably with other process and outcome measures (for details, see Conceição, 2005). By 2008, on a third version of the inventory, several items were again removed and others were edited as a result several meetings within the research team taking into
86
consideration previous statistical analysis on the second version of the inventory. The aims were: a) to shorten the measure as much as possible; b) to try to better differentiate what might be moment specific work from phase specific work and keeping only the more phase specific items; and c) to contruct mirrorred versions of each version, therapist operation and patient capacities, from three perspectives, patient, therapist and observer. One reversed item per subscale was introduced. The third version of GSI ended up being composed by 35 items distributed in 7 subscales all with 5 items each. It is the version used in this research project, which aimed at a final version of the instrument, before it could be tested with other measures from other sequential models or with measures assessing interventions from several therapeutic approaches, like the ones discussed in the introduction. Patients and therapists rate the scales on parallel forms using an ordinal scale of 7 points (1- “not all descriptive” a 7-“totally descriptive”).
I am aware of the
importance of assessing both perspectives, yet for the purpose of this dissertation only the therapist parallel forms will be analysed. Instructions on the form read, ‘Thinking about today’s meeting, please indicate how descriptive each item is regarding what you have been predominantely trying to achieve with your patient in this last session’. The focus on a single, recent session allows for the general strategies to change during therapy and avoids assuming that respondents can accurately remember and summarize details of the strategic goals across weeks or months. The therapist answers the extent to what the items describe the processing she tried to promote in his patient (e.g.: “I tryed to help the patient understand how he contributes to his own dificulties”) and the extent to which she feels his patient is, at the time, capable of that kind of processing (e.g.: “The patient is able to understand how he contributes to his own dificulties”). Inverted items were phrased negatively when presented to participants (“I DID NOT try
87
to help understand how he contributes to his own difficulties” and “The patient IS NOT able to understand how he contributes to his own difficulties”).
Since psychometric evaluation of this inventory is one of the main goals of this first study of the current project, data will be presented on the results section.
Brief Symptom Inventory (BSI). The BSI (Derogatis, 1993) is currently one of the most widely used measures of general psychological symptoms. Each of the 53 items asks patients to rate perceived psychological distress for the past week on a 5point Likert scale (0/not at all , 4/extremely). The Global Severity Index (BSI-GSI) is calculated based on the 53 items. Test/retest reliability (2 week interval) of .90 for the total Global Severity Index has been reported, as well as considerable evidence of convergent, divergent, and predictive validity. Estimate of Improvement Scale (EIS-PC). EIS provides an estimate patients and therapist do of the improvements in patient’s life since the beginning of psychotherapy. Patients and therapists report “to what extent do you feel (you/your patient) improved since the beginning of psychotherapy?”. Five itens capture estimates of improvement relative to (i) initial complaints and symptoms, (ii) personal relationships, (iii) social life, (iv) work/study and (v) the way the patient feels about himself. One last item taps on an estimate of global change since the beginning of psychotherapy. This scale represents intermediate outcomes and involves improvements inside and outside sessions. Internal consitency of this scale is .85 and .89 for patient and therapists perspectives respectively. An exploratory factor analysis of this scale in the patient version results in a single factor solution (eigenvalue = 3,4), with all item
88
loadings ranging from .62 to .79, explaining 57% of variance (Kaiser-Meyer-Olkin Index of .84). The same analysis in the therapist version yield even better results with all item loadings on the single factor (eigenvalue = 3,9), ranging from .68 to .77, explaining 65% of variance (Kaiser-Meyer-Olkin Index of .87). These psychometric characteristics are similar to the ones found in a previous study (Conceição, 2005). In that same previous study, for the therapist perspective, this scale correlated significantly with Outcome Questionnaire (Lambert, Burlingame, Umphress, Hansen, Vermeersch, Clouse, & Yanchar, 1996) global scale (r = -.41; p< 0.05, one-tailed) and Outcome Questionnaire Symtom Distress scale (r = -.36; p< 0.05, one-tailed), Interpersonal Relationships scale (r = -.44; p< 0.05, one-tailed) and Social Role Functioning scale (r = -.29; p=.08, one-tailed). The patient version of this scale had the same trend of negative correlations but did not reach significance, probably due to the small sample size of patients.
In this study the patient version of this scale correlates negatively and
significantly with BSI global scale (r = -.21; p< 0.05; N=121, one-tailed) a pattern that does not reach significance in the therapist version (r = -.10; p= .20; N=74, one-tailed). Working Alliance Inventory - Short Form (WAI-S). This inventory assesses the working alliance and is composed by three dimensions, bond, agreement between therapist and patient on goals and on tasks. Participants report the frequency of feeling and thoughts in relation to the other element of the therapeutic dyad, on a Lykert-type scale (from 1 “never” to 7 “always”). The short version has 12 items, 4 for each dimension (Tracey & Kokotovic, 1989). In this study only the global scale will be used. Internal consistency is .89 for the patient version and .85 for the therapist version. Session Outcome scale (SOS).
This measure is constructed based on the
Session Evaluation Scale (SES) developed by Hill & Kellems (2002) on which two
89
items were added and the response format changed (see Conceicao 2005) resulting in a new measure. Each participant is asked to rate 6 items assessing session outcomes like "I feel this session was valuable", “I am glad I came to this session" or "I am satisfied with this session". In a previous study, Conceicao (2005) found internal good internal consistency for this measure in both therapist and patient perspectives (Cronbach’s alpha .83 and .85 respectively), which was supported by a single factor solution, in a Principal Component Factorial Analysis.
In the current research project these psychometric
characteristics are replicated. Cronbach’s alpha of .84 and .85 for therapist and patient’s perspective, respectively, and an Exploratory Principal Component Factorial Analysis yields a single factor solution accounting for 57% of variance in both perspectives (KMO = .77 for patient and .80 for therapist perspective). Procedure Therapists and patients were asked to complete the battery of self-report measures and a short demographic questionnaire for the most recent session they conducted. Therapists could fill out the questionnaires for more than one patient. All participation was voluntary. These paper instruments were returned either by post or in person. Therapists and patients had no access to each other data and were informed about that.
For the data collected online recruiting only therapist participacion,
potential participants were informed of the study via email. They were provided with information about the study and then chose whether to take part by completing the relevant measures. This collection was made using SURVS, a web-based software developd for surveys. Sessions were sampled cross-sectionally (i.e., from any point during the treatment at the time of data collection). All participation was voluntary.
90
Results
The first aim was to determine the psychometric qualities of the items and scales of the General Strategies Inventory to check: a) its direct usefulness for testing the sequential component of Paradigmatic Complementarity (see Figure 1); and b) the possibility of empirically deriving an abbreviated measure that respected the conceptual coherence of that sequence, even if it would have less than seven dimensions. In the following tables, all items are written in such a way as to contain information regarding it. Item i7:1, for example, means that the item occupies position 7 in the measure, in the sequence of 35 items, and that it is one of the items developed to measure the first dimension of the 7 dimensions of Paradigmatic Complementarity general strategies sequencing model (see Table 1). If the item contains an r at the end of their description it means it was an inverted item, by the time it was presented to the participants, that was reversed. Descritive Statistics & Internal Consistency Analysis Item means, standard deviations, corrected item-total correlation, Cronbach's Alpha if item deleted, and Inter-Item Correlation Matrix for each of the seven dimensions is presented in tables 1 to 7. Each table contains data regarding both forms of the General Strategies Inventory, the Therapist’s Operations Form (GSI-Top/t) and the Patient’s Change Processes Form (GSI-Pcp/t). Only the therapist perspective of both forms will be used in this research project that concerns the decision-making process of the therapist. .
91
Table 1 Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 1 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) Scale One Trust, Motivation and Hope Building and Relationship Structuring Item Description i7:1
feel motivated to approach his/her problems from a psychological standpoint
i14:1
feel hopeful that he/she might improve with the help of psychotherapy
i16:1
experience the setting as safe and trust (in) therapist’s ability to help him/her
i29:1
feel a collaborative relationship with the therapist
i31:1r
negotiate the structure and the rules of the therapeutic process in order to make it viable
Mean
SD
Corrected
Cronbach's
Item-Total
Alpha if Item
Correlation
Deleted
4.04
1.86
.54
.76
5.40
1.31
.64
.75
3.82
1.93
.68
.72
5.34
1.27
.58
.77
3.95
1.96
.69
.71
5.60
1.13
.73
.72
4.27
1.95
.73
.70
5.78
1.07
.74
.72
3.63
2.35
.85
6.01
1.24
.30 .31
92
.85
Inter-Item Correlation Matrix i7:1
i14:1
i16:1
i29:1
i31:1r
.48
.51
.48
.22
.65
.67
.25
.72
.22
.48 .57
.58
.59
.56
.79
.31
.20
.25
.29 .27
Table 2 Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 2 Of GSI-Top/T (N=227) And GSI-Pcp/T (N=219) Scale Two Increasing Awareness of Self/Experience Item Description
Mean
SD
Corrected
Cronbach's
Item-Total
Alphaif Item
Correlation
Deleted
4.55
1.75
.35
.50
4.61
1.42
.64
.80
i9_2
translate problematic experience into its elements (e.g., cognition, emotion, 4.02 behaviour)
1.69
.25
.56
4.32
1.37
.65
.79
i17_2
increase awareness of how he/she relates to others and others relate to him/her
4.76
1.59
.37
.49
4.76
1.38
.75
.77
i20_2
feel curious and interested in observing how he/she treats himself/herself
4.07
1.68
.34
.51
5.08
1.28
.57
.81
5.26
1.83
.34
.51
5.40
1.41
.55
.82
i3_2
become aware of parts or needs (of)within himself/herself which are in conflict
i18_2r
explore or experience the impact of relevant situations on himself/herself
93
Inter-Item Correlation Matrix i3:2
i9:2
i17:2
i20:2
i18:2r
.19
.28
.23
.19
.15
.09
.23
.31
.20
.56 .60
.59
.41
.50
.45
.38
.54 .56
.23 .39
Table 3 Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 3 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) Scale Three New Meaning Making Regarding Self/Experience Item Description
Mean
SD
Corrected Item-Total Correlation
Cronbach's Alpha if Item Deleted
Inter-Item Correlation Matrix i2:3
i11:3
i21:3
i2_3
make sense of his/her problematic experiences in terms of past or present 4.46 circumstances 4.51
1.72 1.51
.40 .69
.68 .84
i11_3
understand what he/she is trying to attain when behaving in hindering ways
4.08 4.12
1.74 1.48
.49 .66
.64 .85
.53
i21_3
overcome the processes that hinder awareness of experience or helpful meaning 3.60 making 3.87
1.62 1.41
.39 .75
.68 .83
.61
.58
i22_3
identify existing patterns in his/her usual ways of functioning
4.51 4.86
1.72 1.40
.57 .73
.61 .83
.59
.62
.65
form new plausible links or explanations for his/her problematic attitudes and 5.24
1.83 1.40
.47 .62
.66 .86
.54
.46
.57
i19_3r behaviour
5.30
94
.29
i22:3
i19:3r
.20
.32
.33
.27
.52
.28
.33
.32 .36
.52
Table 4 Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 4 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) Scale Four Regulation of Responsibility Item Description
Mean
SD
Corrected Item-Total Correlation
Cronbach's Alpha if Item Deleted
Inter-Item Correlation Matrix i13:4
i23:4
i30:4
i13_4
understand when and how he/she contributes to his/her usual difficulties
4.15 4.46
1.69 1.44
.40 .64
.63 83
i23_4
recognize himself/herself as the only agent of his/her own choices
4.12 4.35
1.84 1.53
.55 .76
.56 .80
.61
i30_4
commit to respecting and validating his/her needs, regardless of their correspondence to others’ expectations
4.24 3.96
1.78 1.45
.32 .65
.67 .83
.49
.59
i33_4
take responsibility for taking care of her/himself by tentatively mobilizing internal or external resources toward that goal
4.59 4.17
1.65 1.41
.47 .70
.60 .82
.46
.62
.69
i4_4r
acknowledge that he/she can promote or hinder the satisfaction of his/her needs
5.35 4.87
1.70 1.72
.39 .61
.63 .84
.53
.60
.38
95
.37
i33:4
i4:4r
.11
.30
.33
.31
.42
.34
.30
.18 .24
.51
Table 5 Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 5 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219). Scale Five Implementation of Repairing Actions Item Description
Mean
SD
Corrected Item-Total Correlation
Cronbach's Alpha if Item Deleted
Inter-Item Correlation Matrix i12:5
i25:5
i27:5
i12_5
ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs
3.99 3.59
1.68 1.42
.60 .77
.67 .85
i25_5
effectively deal with situations and respect his/her needs at the same time
4.13 3.48
1.73 1.38
.58 .77
.67 .85
.71
i27_5
stand up for her/himself by dealing with internal or external blocks to the expression of his/her identity
3.94 3.48
1.75 1.44
.61 .81
.66 .84
.74
.72
i28_5
choose life-styles that allow him/herself to live the present as well as to promote his/her personal development
3.74 3.71
1.98 1.56
.52 .78
.69 .85
.65
.71
.77
i10_5 r
integrate different parts or needs within himself/herself into a congruent and satisfied enough whole
4.92 4.03
1.81 1.69
.26 .55
.79 .91
.49
.47
.49
96
.52
i28:5
i10:5r
.55
.44
.19
.49
.45
.21
.46
.24 .18
.49
Table 6 Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 6 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) Scale Six Consolidation of Change Item Description
Mean
SD
Corrected Item-Total Correlation
Cronbach's Alpha if Item Deleted
Inter-Item Correlation Matrix i6:6
i15:6
i32:6
i6_6
generalize the expression of his/her identity to the different domains of his life
3.26 3.39
1.68 1.51
.51 .80
.64 .83
i15_6
emotionally nurture and nourish himself/herself regarding the expression of his/her identity and growth
4.15 3.66
1.62 1.45
.45 .78
.66 .84
.74
i32_6
develop relationships/situations that support his/her choices
3.79 3.97
1.87 1.37
.50 .74
.64 .85
.70
.67
i34_6
deal with internal or external obstacles to the consolidation of his/her identity and growth
4.10 3.61
1.80 1.35
.56 .82
.62 .83
.77
.75
.71
i24_6r
accept the inevitability of a certain degree of vulnerability or conflict as a result of experiencing and expressing his/her identity
4.57 4.63
1.92 1.60
.32 .48
.72 .91
.44
.42
.39
97
.42
i34:6
i24:6r
.40
.42
.19
.35
.32
.20
.45
.22 .33
.47
Table 7 Item Means, Standard Deviations, Corrected Item-Total Correlation, Cronbach's Alpha, And Inter-Item Correlation Matrix For Dimension 7 Of GSI-Top/t (N=227) And GSI-Pcp/t (N=219) Scale Seven Anticipation of The Future and Relapse Prevention Item Description
Mean
SD
Corrected Item- Cronbach's Total Alpha if Item Correlation Deleted
Inter-Item Correlation Matrix i1:7
i5:7
i8:7
i1_7
realize that he/she can cope autonomously with life’s future challenges
3.69 3.50
1.85 1.62
.55 .72
.61 .84
i5_7
anticipate resources required to cope with future scenarios beyond termination
2.67 3.20
1.92 1.52
.41 .78
.66 .83
.71
i8_7
project himself/herself in the future, effective and affectively relating to self, others and the world
3.60 3.59
1.91 1.51
.65 .78
.56 .83
.68
.72
i35_7
strengthen the feeling of self-coherence and purpose in his/her life
3.48 3.77
1.84 1.54
.47 .78
.64 .82
.69
.69
.74
i26_7r
integrate experiences from the past, present and anticipated future into a coherent self-narrative
4.83 4.71
2.08 1.81
.24 .48
.74 .91
.35
.45
.42
98
.43
i35:7
i26:7r
.54
.36
.17
.46
.25
.04
.45
.26 .24
.46
Adjusted item-to-scale correlations provide an estimate of the convergence between the item being evaluated and the rest of the items in its subscale. In calculating the adjusted item-to-scale correlations, the item being evaluated was excluded from the total subscale score so as not to inflate the correlation. An item is considered to possess adequate convergence if its adjusted item-to-scale correlation is equal to or greater than .30 (Nunnally & Bernstein, 1994).
Adjusted item-to-scale correlations were then
computed based on the sample of 227 rated sessions for the GSI. Only in the GSITop/t, three items fell below that criteria: i9_2 (translate problematic experience into its elements (e.g., cognition, emotion, behaviour)); i10_5 r (integrate different parts or needs within himself/herself into a congruent and satisfied enough whole), and i26_7r (integrate experiences from the past, present and anticipated future into a coherent selfnarrative). This provides empirical evidence that they are not measuring the same construct measured by the other scale items. In both forms, a closer inspection on all inverted items, after being reversed, reveals poor psychometric results casting doubts on the construction quality of the inversion. In the GSI-Top/t, correlations for each five items with at least one other from the same scale ranged from .04 to .72, with a substantial number above .40, in subscales one, five, six and seven. These results suggest reasonable factorability but dimensions two three and four may have items that aggregate differently with regard to the theoretical proposal. In the GSI-Pcp/t overall, correlations for each five items with at least one other from the same scale ranged from .20 to .79, with a substantial number above .40 in all subscales.
These results suggest reasonable factorability but the
considerably higher correlations may indicate the presence of a large global factor (like overall patient strategy related change) or, at least, less dimensions with regard to the theoretical proposal.
99
Coefficient alpha is an internal reliability statistic that provides an evaluation of the internal consistency of the items defining a subscale. A coefficient alpha of .70 or greater is generally considered to represent adequate internal consistency (Nunnally & Bernstein, 1994). In addition to displaying the adjusted item-to-scale correlations noted previously, coefficient alphas were computed for the GSI subscales. As shown in Table 8, the original subscales with 5 items each, standardized item alphas based on the sample of 227 rated sessions for the GSI dimension 1, 2, 3, 4, 5, 6 and 7 were .80, .57, .70, .67, .75, .70, and .70 respectively. Considering the subscales only with four items, excluding the inverted ones, the coefficient alpha were .85, .51, .66, .63, .79, .72, and .74, respectively, indicating accetable to good internal consistency with the exception of dimension 2. Table 8 Internal Consistency For The Subscales Of The GSI-Top/t and GSI-Pcp/t
Scale 1
Trust, motivation, hope building and structuring
Scale 2
Increasing awareness of self and experience
Scale 3
Meaning making regarding self and experience
Scale 4
Regulation of responsibility
Scale 5
Implementation of repairing actions
Scale 6
Consolidation of change
Scale 7
Anticipation of the future and relapse prevention
Therapeutic Operations
Patient Change Processes
.80 .85 .57 .51 .70 .66 .67 .63 .75 .79 .70 .72 .70 .74
.81 .85 .83 .82 .87 .86 .86 .84 .89 .91 .89 .91 .88 .90
Note: Values above for 5 item scale and below for 4 item scale; GSI-Top/t, N=227; GSI-Pcp/t, N=219;
Multidimensional Scaling In a previous study (Vasco, 2006) the sequential progression of strategic objectives as promoted by the therapist (in this research project called “therapist’s general
strategy
related
operations”)
was 100
tested
using
such
an
analysis.
Multidimensional Scaling (MDS) is an exploratory approach for identifying structure within several dimensions. This is a statistical technique that assesses the similarity or dissimilarity of a set of objects (strategies in this case) in multidimensional space (Borg & Groenen, 1997). For a similar purpose, now for both forms of GSI (patient’s general strategy related change processes had never been subjected for such an analysis), correlations at the item level were transformed into a matrix for exploratory MDS procedures. The correlation matrices were then submitted to the MDS analysis using STATISTICA, a software package for statistical analysis. MDS is usually beneficial in assessing the parsimony of a multidimensional model while allowing one to view the structural relationship between data. Given the use of proximities in MDS, error is quantified through the error of representation, or the extent to which the data yield a “badness-of-fit” to the proposed model. This function, called “Stress” assesses the mismatch of the proximities and provides an assessment of the fit of the proposed dimensions. In the present study, the Stress function will be used as the initial criterion to evaluate whether a singledimensional model is acceptable. A second criterion that will also be taken into account lies in the interpretability of the coordinates. If the m-dimensional solution provides a satisfying interpretation, but the (m + 1) - dimensional solution reveals no further structure, it may be well to use only the m-dimensional solution (Borg & Groenen, 1997). MDS with the 28 items of the GSI-Top/t, replicates previous results (Vasco, 2006) corroborating the principle of sequential ordering of the strategies used by therapists in general (see Table 9).
101
Table 9 Multidimensional Scaling With 28 Items Of The GSI-Top/t Item
Values
Description
i:16_1
1,87
experience the setting as safe and trust (in) therapist’s ability to help him/her
i:29_1
1,77
feel a collaborative relationship with thetherapist
i:14_1
1,53
feel hopeful that he/she might improve with the help of psychotherapy
i:07_1
1,49
feel motivated to approach his/her problems from a psychological standpoint
i:09_2
1,30
translate problematic experience into its elements (e.g. cognition, emotion, behaviour)
i:20_2
1,06
feel curious and interested in observing how he/she treats himself/herself
i:02_3
0,84
make sense of his/her problematic experiences in terms of past or present circumstances
i:03_2
0,77
become aware of parts or needs (of)within himself/herself which are in conflict
i:17_2
0,53
increase awareness of how he/she relates to others and others relate to him/her
i:11_3
0,45
understand what he/she is trying to attain when behaving in hindering ways
i:21_3
0,42
overcome the processes that hinder awareness of experience or helpful meaning making
i:22_3
0,34
identify existing patterns in his/her usual ways of functioning
i:13_4
0,13
understand when and how he/she contributes to his/her usual difficulties
i:34_6
-0,41
deal with internal or external obstacles to the consolidation of his/her identity and growth
i:23_4
-0,46
recognize himself/herself as the only agent of his/her own choices
i:25_5
-0,51
effectively deal with situations and respect his/her needs at the same time
i:15_6
-0,55
emotionally nurture and nourish himself/herself regarding the expression of his/her identity and growth
i:30_4
-0,65
commit to respecting and validating his/her needs, regardless of their correspondance to others’ expectations
i:27_5
-0,70
stand up for her/himself by dealing with internal or external blocks to the expression of his/her identity
i:12_5
-0,72
ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs
i:33_4
-0,74
take responsability for taking care of her/himself by tentatively mobilizing internal or external resources toward that goal
i:28_5
-0,85
choose life-styles that allow him/herself to live the present as well as to promote his/her personal development
i:32_6
-0,89
develop relationships/situations that support his/her choices
i:08_7
-0,98
project himself/herself in the future, effective and affectively relating to self, others and the world
i:06_6
-1,04
generalize the expression of his/her identity to the different domains of his life
i:35_7
-1,09
strengthen the feeling of self-coherence and purpose in his/her life
i:01_7
-1,33
realize that he/she can cope autonomously with life’s future challenges
i:05_7
-1,59
anticipate resources required to cope with future scenarios beyond termination
102
A stress value = .22 (D-star: Raw stress = 48,27; Alienation = ,24; D-hat: Raw stress = 40,03) for a single dimension solution is at the limit of acceptability, and for the purpose of this research project I believe this constitutes a creative use of Multidimensional Scaling in revealing evidence for the phenomenon of sequencing of general strategies, be it explicit or implicit in the therapists’ head. When interpreting MDS results, it is important to note that additional dimensions will always reduce the stress coefficient and additional variables will always increase the stress coefficient (Cox & Cox, 1994). On the one hand, had I included the 7 inverted items and stress would increase (specifically to .28, in this case, beyond the limits of acceptability). For example, on the other hand, considering the analysis exluding those items, a two dimension solution (in the case of therapist’s strategy related operations) reduces Stress level from, .22 to .17 and is still easy to interpret, in that the second dimension besides the temporal dimension separates strategies of the extremes of the psychotherapeutic process (beginnings and endings) from strategies of the working through of the process. Solutions with more dimensions become difficult to interpret and do not better serve the goals of this study. A new finding is that MDS with the 28 items of the GSI-Pcp/t form also result in a corroboration of the principle of sequential ordering of the patients’ strategy related change processes in general (see Table 10). The resulting stress value = .22 (D-star: Raw stress = 46.15; Alienation = .24; D-hat: Raw stress = 37.84) for a single dimension solution is also at the limit of acceptability.
103
Table 10 Multidimensional Scaling With 28 Items Of The GSI-Pcp/t Item
Values
Description
i:14_1
2,92
feel hopeful that he/she might improve with the help of psychotherapy
i:07_1
1,98
feel motivated to approach his/her problems from a psychological standpoint
i:16_1
1,52
experience the setting as safe and trust (in) therapist’s ability to help him/her
i:20_2
1,39
feel curious and interested in observing how he/she treats himself/herself
i:29_1
1,29
feel a collaborative relationship with thetherapist
i:09_2
0,52
translate problematic experience into its elements (e.g. cognition, emotion, behaviour)
i:13_4
0,40
understand when and how he/she contributes to his/her usual difficulties
i:03_2
0,37
become aware of parts or needs (of)within himself/herself which are in conflict
i:02_3
0,33
make sense of his/her problematic experiences in terms of past or present circumstances
i:17_2
0,24
increase awareness of how he/she relates to others and others relate to him/her
i:11_3
0,13
understand what he/she is trying to attain when behaving in hindering ways
i:22_3
0,11
identify existing patterns in his/her usual ways of functioning
i:23_4
-0,16
recognize himself/herself as the only agent of his/her own choices
i:21_3
-0,30
overcome the processes that hinder awareness of experience or helpful meaning making
i:33_4
-0,44
take responsability for taking care of her/himself by tentatively mobilizing internal or external resources toward that goal
i:15_6
-0,55
emotionally nurture and nourish himself/herself regarding the expression of his/her identity and growth
i:30_4
-0,60
commit to respecting and validating his/her needs, regardless of their correspondance to others’ expectations
i:34_6
-0,62
deal with internal or external obstacles to the consolidation of his/her identity and growth
i:27_5
-0,65
stand up for her/himself by dealing with internal or external blocks to the expression of his/her identity
i:32_6
-0,71
develop relationships/situations that support his/her choices
i:35_7
-0,74
strengthen the feeling of self-coherence and purpose in his/her life
i:06_6
-0,77
generalize the expression of his/her identity to the different domains of his life
i:12_5
-0,79
ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs
i:25_5
-0,86
effectively deal with situations and respect his/her needs at the same time
i:28_5
-0,90
choose life-styles that allow him/herself to live the present as well as to promote his/her personal development
i:08_7
-0,91
project himself/herself in the future, effective and affectively relating to self, others and the world
i:05_7
-1,09
anticipate resources required to cope with future scenarios beyond termination
i:01_7
-1,11
realize that he/she can cope autonomously with life’s future challenges
104
Factor Analysis of the GSI To further assess the construct validity and explore the underlying factor structure of the GSI, exploratory factor analysis was used. Note that item generation for this new measure was guided by the goal of a seven factor solution (according to the sequencial component of Paradigmatic Complementarity) as can be seen in Figure 1. Nonetheless, I chose to use exploratory factor analysis (EFA) rather than confirmatory factor analysis (CFA) for this purpose given the newness of the construct being investigated and the lack of solid theory and empirical research on the construct; CFA is most useful in later stages of measure development to refine and improve measures (Kahn, 2006). Hence, EFA guided by a seven factor solution was used. While acknowledging the possibility of a different factor structure emerging, I used EFA primarily as a tool to assist in selecting the best and most representative items for the proposed subscales of general strategies with the aim of empirically deriving an abbreviated measure. At least two ways of conducting exploratory factor analysis on the GSI are possible, each with advantages and disadvantages. The first way involved conducting EFA on the entire sample of 227 participants, while the second way involved splitting the data into two halves, with one half being the replication/confirmation sample used to check the factor stability of the factors identified in the other half. The advantage of using the first method (EFA on the whole sample) is the larger sample size, which decreases the possibility of finding relations in the data by chance; on the other hand, using only one sample for EFA without any confirmation or replication of findings can also result in findings by chance alone and the stability of the factor structure would be unknown. Results would then await confirmation in future research conducted using the GSI measure. The advantage of the split half method, therefore, is the increased
105
confidence in findings if they replicate from the first sample to the second (confirmation) sample. However, the smaller sample sizes of each of the split halves (n of 102) could be problematic, as some recommendations call for at least 200/250 participants for EFA, or a minimum ratio of 5 cases per item. In sum, one could argue for either of the two methods being more appropriate for the present study. I decided on the first method, conducting EFA on the entire sample because of its greater statistical power, while acknowledging the need for confirmation of the factor stability of the GSI in future research. Before conducting the factor analysis, all 35 items were examined and deemed appropriate for factor analysis on the basis of means, standard deviations, kurtosis, and skewness. A principal components analysis (PCA), as opposed to principal axis factor (PAF) analysis, was used to explore the factor structure of the GSI. PAF is most commonly used when the goal is to understand the relations among a set of measured variables (items) in terms of a smaller number of underlying latent variables (factors) (Kahn, 2006), but I wanted to maximize the number of factors. Rotated solutions are usually preferable because they create a more even distribution of the variance accounted for among factors, increase the interpretability of factors, and make variables load highly on as few factors as possible (Kahn, 2006). After considering whether to use an orthogonal (e.g., varimax) or oblique (e.g., promax) rotation, a varimax rotation was selected. Different rotations may provide slightly different results but the differences are not usually dramatic; such was the case in the present study as I found very similar results when comparing promax to varimax rotations. Only varimax rotation results are presented. Therapist’s strategy-related operations. To identify the factor structure of the Therapist’s Strategy-Related Operations Form of the GSI, a principal component factor 106
analysis with varimax rotation was conducted on the 35 questionnaire items. Since the inverted items loaded on the same factor they were removed from further analysis and the regular 28 items were reentered for a new analysis.
In this second analysis,
sampling adequacy (Kaiser-Meyer-Olkin) indicates a relative compact pattern of correlations (0.84) and analysis of sphericity indicates a strong relationship between the items (df = 378, p < .000), both of which are indicators that factor analysis is appropriate for this measure. Factors were extracted on the basis of eigenvalue greater than 1, scree testing (Figure 2), proportion of variance accounted, percentage of variance explained by each factor, number of items with significant factor loadings, and factor interpretability (Kahn, 2006) and closest resemblance to the theoretical proposal of Paradigmatic Complementarity sequencing model with seven dimensions (Vasco, 2006).
Figure 2: Scree plot for factor analysis of the 28 Items from GSI-Top/t Considering all these criteria, the best solution found was one of five main factors, accounting for 54.22% of the variance in scores. Results are presented in table 11. 107
Table 11 Factor Loadings Of The 28 Items Of The GSI-Top/t F1
F2
F3
F4
F5
i29:1
feel a collaborative relationship with the therapist
.85
i16:1
experience the setting as safe and trust (in) therapist’s ability to help him/her
.85
i14:1
feel hopeful that he/she might improve with the help of psychotherapy
.81
i07:1
feel motivated to approach his/her problems from a psychological standpoint
.74
i20:2
feel curious and interested in observing how he/she treats himself/herself
.46
i09:2
translate problematic experience into its elements (e.g. cognition, emotion, behaviour)
.31
i03:2
become aware of parts or needs (of)within himself/herself which are in conflict
.69
i02:3
make sense of his/her problematic experiences in terms of past or present circumstances
.68
i17:2
increase awareness of how he/she relates to others and others relate to him/her
.59
i22:3
identify existing patterns in his/her usual ways of functioning
.76
i13:4
understand when and how he/she contributes to his/her usual difficulties
.73
i11:3
understand what he/she is trying to attain when behaving in hindering ways
.73
i23:4
recognize himself/herself as the only agent of his/her own choices
i21:3
overcome the processes that hinder awareness of experience or helpful meaning making
i27:5
stand up for her/himself by dealing with internal or external blocks to the expression of his/her identity
.72
i30:4
commit to respecting and validating his/her needs, regardless of their correspondence to others’ expectations
.71
i12:5
ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs
.69
i25:5
effectively deal with situations and respect his/her needs at the same time
.68
i28:5
choose life-styles that allow him/herself to live the present as well as to promote his/her personal development
.62
i33:4
take responsibility for taking care of her/himself by tentatively mobilizing internal or external resources toward that goal
.57
i32:6
develop relationships/situations that support his/her choices
.56
.49
i15:6
emotionally nurture and nourish himself/herself regarding the expression of his/her identity and growth
.50
.30
i34:6
deal with internal or external obstacles to the consolidation of his/her identity and growth
.48
.35
i06:6
generalize the expression of his/her identity to the different domains of his life
.36
.53
i01:7
realize that he/she can cope autonomously with life’s future challenges
.78
i08:7
project himself/herself in the future, effective and affectively relating to self, others and the world
.77
i05:7
anticipate resources required to cope with future scenarios beyond termination
.67
i35:7
strengthen the feeling of self-coherence and purpose in his/her life
.57
108
.42
.32 .31
.46 .39
.43 .40
.36 .32
In all subsequent tables revealing results from factor analysis, loading values below .30 will not be presented for the sake of clarity in terms of visual analysis. The first factor included predominantly items representing the original dimension 5, implementation of reparing actions together with a few representatives of the adjacent dimensions 4 and 6 and explained 15% of the variance (eigenvalue = 4.20). The second factor accounting for 11.3 % of the variance (eigenvalue = 3.17) included all items representing the original dimension 7, projection in the future, together with a single representative of the adjacent dimensions 6. The third factor included predominantly items representing the original dimension 1, hope, safety motivation and structure building explaining 11% of the variance (eigenvalue = 3.10). The fourth factor accounting for 9.8 % of the variance (eigenvalue = 2.75) included mixed items representing the original dimension 3, new meaning making regarding self and experience, and dimension 4, regulation of responsibility.
Finaly the fifth factor
included items from the original dimension 2, increasing awareness of experience, together with an item from the adjacent dimension 3, and it explained 7.25 % of the variance (eigenvalue = 2.02). Despite the ideal structure of 7 dimensions as theoretically proposed in Paradigmatic Complementarity was not fully validated by factorial results, I could now try to develop a short valid measure to test not a 7 dimension sequential model, but a 5 dimension sequential model, that still respected the theoretical model. I mentioned “not fully”, because I consider the fact that the somewhat mixed factors included items predominantly from adjacent dimensions, of acceptable validity, given the clinical reality where therapists tend to merge dimensions. Indeed, this result would allow me to use the four-item seven scales to attempt a replication of previous results (Vasco, 2006; Conceição, 2005), as I did, with identical main results, although they are not presented
109
here. My choice to further the depuration and shrinkage of the measure, in this research project, is influenced by higher validity and reliability demands with such fuzzy concepts as general strategies (always difficult to operationalize, and even more so in a naturalistic context), as well as the importance of creating practice and research friendly measures. Therefore, my next goal in item reduction was to retain the items that most strongly and purely represented each factor (with as little loading on the other factor as possible), and to narrow down the item pool to 15 items, a minimum number that would permit acceptable reliability. Items lacking substantial loading on either factor or loading on more than one factor were immediately eliminated. I then tried to identify the best 15 items using a combination of the following criteria: (a) highest factor loadings onto one factor and (b) respect for the theoretical structure as postulated in the seven-dimension sequencing model of Paradigmatic Complementarity. The reduction from 7 to 5 dimensions was expected to allow testings of a similar enough sequential model of general strategies, even if slightly reinterpretations of the dimensions were needed. Three iterations with the five factor solution were run, using the same analysis, deleting items each time according to the above criteria and conducting factor analysis again on the remaining subset of items. These criteria allowed for the identification of the best 3 items for each factor. To provide some statistical basis for selecting the third item in some of the factors, a reliability analysis of all factors was conducted, selecting the items with the highest item total correlations and largest decrease in “alpha-if-item deleted” were retained. Scree plot is presented in figure 3.
110
Figure 3: Scree plot for factor analysis of the 15 selected items from GSI-Top/t. The final five factor solution retaining 15 items accounted for 66.4 percent of the variance, KMO of .75 and analysis of sphericity indicating a strong relationship between the items (df = 105, p < .000). Each of the 5 rotated factors accounted for 14.50 to 11.14 percent of the variance. I cheked the internal consistency of the short scales, some of them renamed, and all had acceptable to good reliabilities (alpha = .78; .60; .78; .72; and .71 respectively; see Table 12), even consisting of only three items, which is known decrease the probability of high reliabilities.
111
Table 12 Factor Loadings Of The 15 Items Of The GSI-Top/t F1
F2
F3
F4
F5
Motivation, Safety and Hope building (! = .78) i07:1 I16:1 i14:1
feel motivated to approach his/her problems from a psychological standpoint experience the setting as safe and trust (in) therapist’s ability to help him/her feel hopeful that he/she might improve with the help of psychotherapy
.79 .85 .84
Increasing Awareness and New meanings (! = .60) i02:3 i03:2 i17:2
make sense of his/her problematic experiences in terms of past or present circumstances become aware of parts or needs (of)within himself/herself which are in conflict increase awareness of how he/she relates to others and others relate to him/her
.78 .75 .58
Regulation of Responsibility (! = .78) i22:3 i13:4 i11:3
identify existing patterns in his/her usual ways of functioning understand when and how he/she contributes to his/her usual difficulties understand what he/she is trying to attain when behaving in hindering ways
.73 .83 .82
Regulation of Needs (! = .72) i30:4 i12:5 i25:5
commit to respecting and validating his/her needs, regardless of their correspondence to others’ expectations ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs effectively deal with situations and respect his/her needs at the same time
.79 .73 .74
Consolidation and Projection of Self (! = .71) i06:6 i08:7 i01:7
generalize the expression of his/her identity to the different domains of his life project himself/herself in the future, effective and affectively relating to self, others and the world realize that he/she can cope autonomously with life’s future challenges
112
.65 .81 .84
Patient’s strategy-related change processes. To identify the factor structure of GSI-Pcp/t, a principal component factor analysis with varimax rotation was conducted on the 35 questionnaire items. Since the inverted items also loaded on the same factor they were removed from further analysis. The regular 28 items were reentered for a new analysis. In this second analysis of sampling adequacy (Kaiser-Meyer-Olkin) indicates a relative compact pattern of correlations (0.96), and analysis of sphericity indicates a strong relationship between the items (df = 378, p < .000), both of which are indicators that factor analysis is appropriate for this measure. Factors were again extracted on the basis of eigenvalue greater than 1, scree testing (Figure 4), proportion of variance accounted, percentage of variance explained by each factor, number of items with significant factor loadings, and factor interpretability (Kahn, 2006).
Figure 4: Scree plot for factor analysis of the 28 items from GSI-Pcp/t. Considering all these criteria, the best solution was one of three main factors, accounting for 69% of the variance in scores (see Table 13).
113
Table 13 Factor Loadings Of The 28 Items Of The GSI-Pcp/t F1
i29:1 i14:1 i16:1 i07:1 i20:2 i03:2 i11:3 i13:4 i22:3 i02:3 i17:2 i23:4 i09:2 i06:6 i28:5 i35:7 i08:7 i05:7 i34:6 i27:5 i01:7 i15:6 i25:5 i32:6 i12:5 i33:4 i30:4 i21:3
feel a collaborative relationship with the therapist feel hopeful that he/she might improve with the help of psychotherapy experience the setting as safe and trust (in) therapist’s ability to help him/her feel motivated to approach his/her problems from a psychological standpoint feel curious and interested in observing how he/she treats himself/herself become aware of parts or needs (of)within himself/herself which are in conflict understand what he/she is trying to attain when behaving in hindering ways understand when and how he/she contributes to his/her usual difficulties identify existing patterns in his/her usual ways of functioning make sense of his/her problematic experiences in terms of past or present circumstances increase awareness of how he/she relates to others and others relate to him/her recognize himself/herself as the only agent of his/her own choices translate problematic experience into its elements (e.g. cognition, emotion, behaviour) generalize the expression of his/her identity to the different domains of his life choose life-styles that allow him/herself to live the present as well as to promote his/her personal development strengthen the feeling of self-coherence and purpose in his/her life project himself/herself in the future, effective and affectively relating to self, others and the world anticipate resources required to cope with future scenarios beyond termination deal with internal or external obstacles to the consolidation of his/her identity and growth stand up for her/himself by dealing with internal or external blocks to the expression of his/her identity realize that he/she can cope autonomously with life’s future challenges emotionally nurture and nourish himself/herself regarding the expression of his/her identity and growth effectively deal with situations and respect his/her needs at the same time develop relationships/situations that support his/her choices ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs take responsibility for taking care of her/himself by tentatively mobilizing internal or external resources toward that goal commit to respecting and validating his/her needs, regardless of their correspondence to others’ expectations overcome the processes that hinder awareness of experience or helpful meaning making
114
F2
F3
0,80 0,80 0,80 0,72 0,43
0,50
0,76 0,74 0,73 0,72 0,67 0,62 0,48
0,56 0,50
0,83 0,82 0,81 0,81 0,80 0,80 0,79 0,78 0,77 0,76 0,76 0,74 0,74 0,72 0,62
0,53
0,45
The first factor included predominantly items representing the original dimension 5, 6 and 7, implementation of reparing actions, consolidation and projection in the future, together with a few representatives of dimension 4 and explained 36% of the variance (eigenvalue = 10.10). The second factor accounting for 19.46 % of the variance (eigenvalue = 5.45) included items representing the original dimension 2 and 3, increase awareness regarding self and experience and new meaning making regarding self and experience and a few from dimension 4, regulation of responsibility. The last factor included predominantly items representing the original dimension 1, hope, safety motivation and structure building explaining 13.79% of the variance (eigenvalue = 3.86). Next, I repeated the analysis including only the equivalent 15 items from GSITop/t, and a three factor solution (see Figure 5) accounted for 70.9 percent of the variance, KMO of .93 and analysis of sphericity indicating a strong relationship between the items (df = 105, p < .000).
Figure 5: Scree plot for factor analysis of the 15 selected items from GSI-Pcp/t. 115
As can be seen in table 14, in comparison with the analysis of the strategyrelated therapist operations, in this analysis, the first factor explaining 29,36 % of the variance (eigenvalue = 4.40) included axactly factors 3 and 4, regulation of needs, and consolidation and projection of self in the future. The second factor explaining 26,36 % of the variance (eigenvalue = 3.95) included the items from the previous factor 1 and 5 Regulation of Responsibility and Increasing Awareness and New meanings. The third factor explaining 15,16 % of the variance (eigenvalue = 2.27) included all the items from dimension 1, Motivation, Safety and Hope, as factor 2 in the analysis of the therapist’s strategy-related operations. Because of the argument that when some correlation is expected among factors, using orthogonal rotation may result in a loss of valuable information if the factors are correlated (Costello & Osborne, 2005), I also tested an oblique rotation. Nevertheless, principal-axis factoring followed by promax (an oblique rotation) yield exactly the same results. Internal consistency analysis yield good to excellent results, for both the five mirrored short scales (alphas = .78, .84, .85, .88, and .88 respectively; see table 14), and the factorial 3 scales that resulted from the analysis. The first factor had internal consistency values of .90, the second factor .93 and the third factor .78. To determine the reliability of the GSI-Pcp/T, as a global scale of outcomes comprised of the selected 15 items, an analysis of the internal consistency of the measure was performed through the computation of Cronbach's alpha coefficient showing exceptionally high internal consistency: .93; item-total correlations ranged from .65 to .76, with nine values above .70, with the exception of the three items represesenting the first dimension, namely, hope, safety and motivation building (probably closest to the alliance construct), where the values were .58, .50 and .37 for items 16, 14 and 7 respectively.
116
Table 14 Factor Loadings Of The 15 Selected Items Of The GSI-Pcp/T F1 !=.90
F2 !=.93
F3 !=.78
Motivation, Safety and Hope building (! = .78) 07_1 16_1 14_1
feel motivated to approach his/her problems from a psychological standpoint experience the setting as safe and trust (in) therapist’s ability to help him/her feel hopeful that he/she might improve with the help of psychotherapy
.73 .77 .86
Increasing Awareness and New meanings (! = .84) 02_3 03_2 17_2
make sense of his/her problematic experiences in terms of past or present circumstances become aware of parts or needs (of)within himself/herself which are in conflict increase awareness of how he/she relates to others and others relate to him/her
.71 .77 .66
Regulation of Responsibility (! = .85) 22_3 13_4 11_3
identify existing patterns in his/her usual ways of functioning understand when and how he/she contributes to his/her usual difficulties understand what he/she is trying to attain when behaving in hindering ways
.74 .73 .77
Regulation of Needs (! = .88) 30_4 12_5 25_5
commit to respecting and validating his/her needs, regardless of their correspondence to others’ expectations ecologically act in ways where he/she expresses him/herself with clarity and congruence regarding his/her needs effectively deal with situations and respect his/her needs at the same time
.74 .77 .79
Consolidation and Projection of Self (! = .88) 06_6 08_7 01_7
generalize the expression of his/her identity to the different domains of his life project himself/herself in the future, effective and affectively relating to self, others and the world realize that he/she can cope autonomously with life’s future challenges
.82 .82 .81
Table 15 reveals the maximum, minimum, mean and standard deviation values of the empirically derived scales of strategy-related therapist’s operations and patient’s change processes, characterizing the whole sample.
117
Table 15 Maximum, Minimum, Means And Standard Deviations Of The Empirically Derivated Scales Of Strategies N
Minimum
Maximum
Mean
Std. Deviation
Skewness
Kurtosis
Top1:SHM Top2:IAM Top3:RAR Top4:RSN Top5:CPS
227 227 227 227 227
1 1 1 1 1
7 7 7 6.7 7
3.94 4.59 4.25 4.12 3.52
1.60 1.26 1.44 1.38 1.45
-.20 -.75 -.41 -.36 .18
-.98 .28 -.55 -.51 -.74
Pcp1:SHM Pcp2:IAM Pcp3:RAR Pcp4:RSN Pcp5:CPS
220 220 220 220 220
2.7 1 1 1 1
7 7 7 6.3 7
5.44 4.62 4.48 3.68 3.52
1.03 1.25 1.26 1.27 1.39
-.43 -.46 -.43 .04 .13
-.37 -.34 -.21 -.45 -.68
Note. Top, Therapist’s Strategy-Related Operations; Pcp, Patient’s Strategy-Related Change Processes; SHM, Motivation, Safety and Hope building; IAM, Increasing Awareness and New meanings; RAR, Regulation of Responsibility; RSN, Regulation of Needs; CPS, Consolidation and Projection of Self. Skewness Std. Error = .16; Kurtosis Std. Error = 3.2
Despite the operations from the middle of the therapeutic process (regarding increasing awareness and insight, responsibility/agency regulation and regulation of satisfaction of needs) have slightly higher values comparing to the operations at the extremes of the process (safety, hope and motivation building and consolidation and projection of the self), there is no linear trend. In contrast, the patient’s strategy-related change processes appear to follow a negative linear trend, with the ones typical of the beginning of the process demonstrating higher values and the ones typical of the end of the process demonstrating lower values (see figure 6). These results replicate the ones from the previous study by Conceição (2005).
118
(" '" &" %"
23456789:;" 0.001; N= 220). As postulated in Conceição (2005) these higher values constitute evidence of change processes following a sequencing incremental logic with previous ones anticipating the next ones, or next ones including previous ones. As shown in table 17, the pattern of correlations between subscales of therapist’s strategy-related operations and the subscales of patient’s strategy related change processes are again replicated as compared to Conceição (2005), in that there is a negative significant correlation between the operations from dimension one and all the patient change processes (from dimension one to five).
120
Table 17 Correlations Between Subscales Of Therapist Change Processes (Upper Right) And Between Subscales Of Patient Change Processes (Lower Left) Of The General Strategies Inventory – Short Form/Therapsit Perspective Patient Strategy Related Change Processes
Therapist Strategy Related Operations
1:SHM
2:IAM
3:RAR
4:RSN
5:CPS
1:SHM
-.26**
-.29**
-.28**
-.30**
-.27**
2:IAM
.00
.14*
.05
-.03
-.07
.03
.19
**
.07
.07
.32
**
3:RAR
**
4:RSN
.21
5:CPS
.16*
.38**
.30
**
.36
**
.36**
**
.25**
.47**
.55**
.32
Note. SHM, Motivation, Safety & Hope building; IAM, Increasing Awareness & New meanings; RAR, Regulation of Responsibility; RSN, Regulation of Needs; CPS, Consolidation & Projection of Self. ** p< 0.01 (2-tailed); * p < 0.05 (2-tailed). N=227 for Therapist’s Strategy-Related Operations; N= 220 for Patient’s Strategy-Related Change Processes.
Only operations from dimension four and five are consistently and significantly positive correlated with all the patient’s change processes (from dimension one to five). The pattern of correlations between the therapist’s operations from dimension two and three and all the patient’s change processes is more blurred but still allows the general inference found in Conceição (2005) that, according to a postulated sequence, earlier patient change processes correlate negatively with therapist operations of all kind and later patient change processes correlate positively with therapist operations of all kind, with higher values in later or more advanced dimensions. Considering the correlations of subscales of general strategy-related therapist’s operations and patient’s change processes and other process variables it was found (as in Conceição, 2005) that the dimensions of patient change processes behave like intermediate results in that they maintain positive significant correlations with other outcome measures and with therapeutic alliance (see table 18). This pattern also happens with therapist’s operations. In contrast to a previous study were correlations
121
with OQ were quite high (Conceição, 2005), the correlations with BSI-GSI were inexistent in both forms, with the exception of the fourth and fifth more advanced dimensions. Table 18 Correlations Between Subscales Of General Strategies And Other Process And Outcome Variables 1:SHM
2:IAM
3:RAR
4:RSN
5:CPS
Therapist Strategy-Related Operations WAI-s SOS EIS BSI-GSI
-.16 a .04 -.16 a -.02
.46** .26* .58** -.09
.49** .19 b .52** -.04
.48** .26* .48** -.29*
.42** .22* .50** -.19b
Patient Strategy-Related Change Processes WAI-s SOS EIS BSI-GSI
.62** .49** .47** -.17a
.46** .26* .58** -.09
.49** .19 b .52** -.04
.48** .26* .48** -.29*
.42** .22* .50** -.19b
Note. N=91, N=74 for BSI; ** (p
