the importance of qualifications

THE IMPORTANCE OF QUALIFICATIONS CREDENTIALISM IN THE 21ST CENTURY Final Report | April 2008

ii

Copyright in the Importance of Qualifications – Credentialism in the 21st Century is owned by the State of Queensland (acting through the Department of Education, Training and the Arts). While all care has been taken in preparing this publication, the State of Queensland (acting through the Department of Education, Training and the Arts) expressly disclaims any liability for any damage from the use of the material contained in this publication and will not be responsible for any loss, howsoever arising from use, of or reliance on this material.

iii

ABOUT EIDOS Eidos is an independent research institute and think tank (see www.eidos.org.au). Its objective is to generate new ideas and dialogue on good education, labour market and social public policy. We believe that engaged research collaboration and policy innovation contributes to a good society. Eidos is Greek for ideas. Our aim is to inspire, facilitate and support our members and partners to be more collaborative, effective and legitimate. Eidos members include universities and policy leaders. We draw the intellectual strength of our research community into an active dialogue with policy makers and practitioners. Within the Eidos membership there are more than 70 leading research and policy institutes and centres, and over 300 active senior and early career researchers.

ACKNOWLEDGEMENTS The research team included Associate Professor Joseph Zajda and Professor Elizabeth Warren from Australian Catholic University, and Professor Carmel Diezmann from Queensland University of Technology. Dr Peter Grimbeek was responsible for the quantitative analysis of survey data and contributed to other aspects of the report. Eidos Institute would like to acknowledge the significant contribution made by the referee and editor of this report. Eidos Institute facilitated this process, with design by Eidos Creative.

iv

CONTENTS ABOUT THIS REPORT........................................................................................................................................................................ 1 Key messages ................................................................................................................................................................................... 1 EXECUTIVE SUMMARY ...................................................................................................................................................................... 2 INTRODUCTION ................................................................................................................................................................................ 4 LITERATURE REVIEW ......................................................................................................................................................................... 6 METHODOLOGY................................................................................................................................................................................. 9 Research questions......................................................................................................................................................................... 9 Data collection ................................................................................................................................................................................ 9 Data analysis...................................................................................................................................................................................10 QUANTITATIVE OUTCOMES........................................................................................................................................................11 Demographic descriptives...........................................................................................................................................................11 Locality .....................................................................................................................................................................................11 Type of business .....................................................................................................................................................................11 Industry sectors......................................................................................................................................................................11 Associations between demographic variables (bivariate correlations) ......................................................................12 Employment descriptives.............................................................................................................................................................17 Associations between variables ..........................................................................................................................................20 Methodological issues related to inferential statistics...........................................................................................................20 THE IMPORTANCE EMPLOYERS PLACE ON QUALIFICATIONS WHEN RECRUITING ............................................22 Descriptive statistics ....................................................................................................................................................................22 Inferential statistics .......................................................................................................................................................................23 Formal qualifications..............................................................................................................................................................23 Part qualifications ...................................................................................................................................................................23 Interview and focus group data..................................................................................................................................................23 THE IMPORTANCE EMPLOYERS PLACE ON EXPERIENCE WHEN RECRUITING.......................................................25 Descriptive statistics ....................................................................................................................................................................25 Inferential statistics .......................................................................................................................................................................25 Interview and focus group data..................................................................................................................................................25 SUBSTITUTABILITY OF QUALIFICATIONS AND EXPERIENCE .........................................................................................27 Descriptive statistics ....................................................................................................................................................................27 Inferential statistics .......................................................................................................................................................................28 Interview and focus group data..................................................................................................................................................28 MEASURES TAKEN TO IMPROVE OR UPGRADE SKILLS FOR WORKERS WITHOUT QUALIFICATIONS .........30 Descriptive statistics ....................................................................................................................................................................30 Inferential statistics .......................................................................................................................................................................30 Interview and focus group data..................................................................................................................................................31 FACTORS THAT INFLUENCE EMPLOYERS’ OVERALL TRUST IN QUALIFICATIONS................................................32 Descriptive statistics ....................................................................................................................................................................32 Inferential statistics .......................................................................................................................................................................32 OTHER PROBLEMS AND ISSUES RELATED TO VOCATIONAL TRAINING ..................................................................34 Focus group data...........................................................................................................................................................................34 CONCLUSIONS..................................................................................................................................................................................35 REFERENCES........................................................................................................................................................................................36

v

TABLES Table 1. Bivariate associations between select participant characteristics ................................................................... 13 Table 2. Number of current employees in various categories (missing responses excluded)...................................... 17 Table 3. Number of current employees in various categories (missing responses coded as zero) ............................. 18 Table 4. Number of employees in various categories 12 months ago.......................................................................... 18 Table 5. Number of employees in various categories who left in last 12 months........................................................ 18 Table 6. Number of employees in various categories who were hired to replace staff in last 12 months .................. 19 Table 7. Number of employees in various categories who were hired in new positions in last 12 months................ 19 Table 8. Rated importance of formal qualifications across seven categories of positions ............................................ 22 Table 9. Rated importance of part qualifications across seven categories of positions................................................ 22 Table 10. Rated importance of experience across seven categories of positions ........................................................ 25 Table 11. Relative importance of qualifications across seven categories of positions .................................................. 27 Table 12. Whether organisations planned to upgrade workers across seven categories of positions ........................ 30 Table 13. Influence of various considerations on level of confidence in qualifications ................................................. 32 Table 14. Other considerations that influence level of confidence in qualifications ..................................................... 34

FIGURES Figure 1. Relative percentage of participants from eight industry sectors.................................................................... 12 Figure 2. Spatial representation of associations between selected participant demographic characteristics ............... 15 Figure 3. Relative percentage of participants from five composite industry sectors .................................................... 16 Figure 4. Number of participants that favour experience and qualifications equally .................................................... 28 Figure 5. Number of participants that favour qualifications over experience............................................................... 28

1

ABOUT THIS REPORT This project investigated the extent to which employers are substituting formal qualifications with part qualifications (modules) and/or experience when recruiting workers. It questions whether individuals need to hold certain qualifications or credentials in order to obtain particular kinds of employment by examining a) the recruitment behaviour of employers; b) employer attitudes and opinions about the importance of qualifications and experience; c) how companies improve the skills of workers without formal qualifications; and d) what factors influence employers’ confidence/trust in qualifications. Findings aim to inform the Queensland Department of Education, Training and the Arts (DETA) about whether it should set target levels for VET qualifications holdings in various occupational groups; and/or invest more in training that provides the skills, knowledge base and qualities that meet the needs of employers and workers but only leads to the part qualifications and the completion of modules. Key messages This study confirms the finding from previous research that employers highly value formal qualifications when recruiting. Employers rated the importance of formal qualifications most highly for people in trades and professional positions and least highly for those in labourer positions. Employers rated the importance of part qualifications most highly for people in trades positions and least highly for those in labourer positions. The importance placed by employers on experience, when recruiting, varied in different sectors and occupational groups. Employers rated experience most highly when recruiting people for trade, technical and professional positions and least highly for those in labourer positions. In general, employers highly value formal qualifications in conjunction with experience when recruiting. The extent to which employers were substituting qualifications with part qualifications and/or experience depended on the type of position (professional vs. labourer), position category in relation to size of the firm, and industry sector. Reasons given by those employers who substituted qualifications with part qualifications and/or experience included their general preference for experience over qualifications; concerns about access to, and quality of training; and the difficulty in recruiting qualified workers due to the tight labour market. Employers were more likely to train qualified and/or experienced workers such as those in trade, sales, and management and administration positions; and least likely to train unqualified workers such as labourers or technical and professional workers recruited on the basis of pre-existing formal qualifications. Large employers were more likely to train workers than medium-sized employers and small employers. Employers were more likely to trust qualifications that were issued in Australia, and if they had previously employed workers with the same type and level of qualification. Based on the finding that employers highly value qualifications when recruiting workers, the Queensland Government should continue its efforts to set target levels for various occupational groupings and improve the quality of, and access to training. Despite Government’s initiatives to train unqualified workers, this research indicates two challenges: employers in the three industry groupings tend not to train unqualified workers, and/or use accredited modules (that would lead to a qualification) when training workers in general. The first challenge warrants further research about to what extent employers are training workers who are underqualified, unqualified or do not have the required qualification for their job.

2

EXECUTIVE SUMMARY This report offers the latest findings on credentialism, skills and employment in Queensland in 2007 and early 2008. It examines the use made of qualifications, and relevant experience in recruitment processes by a sample of employers in Queensland. Researchers classified employers into three industries in Queensland: (a) Construction Trades; (b) Property and Business Services; and (c) Manufacturing. Employers responded to a range of questions relating to seven categories of employees: Technical; Trades; Clerical, sales and service; Plant/machine operators; Labourers; Management and administration; and Professionals. Researchers asked employers to comment on the importance their firms placed on formal qualifications and experience during the recruitment process, their preparation of workers for jobs, and factors that influenced their confidence in qualifications. Researchers used a formal questionnaire to collect quantitative responses from 180 employers out of a potential sample of 300 employers in Queensland. In addition, researchers conducted 28 interviews and 3 focus groups to collect qualitative responses. Based on survey data, the perceived importance of formal and part qualifications varied between sectors, according to size of firm (number of employees) and category of employee under consideration. The survey included an openended response from which it is clear that while employers did consider training and qualifications important they were also likely to state that experience matters or that they would only hire employees who were already suitably qualified. This variation in employer interest also expressed itself via the number of missing responses to survey questions related to specific categories of employees. Whereas up to two-thirds of employers were likely to answer questions about people in trades or clerical and sales or management and administration positions, typically only one-third answered questions about the importance of qualifications and experience for people in technical, plant & machine operator, or professional positions. A trivial interpretation of such “lop-sided” responsiveness is that those filling in the surveys were in management and administration positions and were likely to have risen from clerical and sales and trades positions. A more considered interpretation is that the pattern of missing responses identifies three groups of employees who differ in terms of the employer’s view of the importance of credentialism. The core group of interest is, as outlined above, those in trades or clerical and sales or management and administration positions, who are most likely to be the focus of on-the-job or off-site training (thus the relatively strong response rates for these categories). For instance, while tradespeople are employed on the basis of existing trade qualifications, they are also employed as apprentices. Likewise, those in sales, clerical, administrative, and management positions benefit from further training, where the training is both an investment in human capital and an advantage (along with experience) when applying for promotional positions. That is, for these three groups, enhanced credentials allow them to progress along career pathways. Employers were less likely to answer questions about the remaining four job categories. Consistent with the above, three of these job categories, comprising those in technical positions, plant and machine operators, and professionals, typically tend to be employed based on prior qualifications. The fourth job category about which employers were less likely to answer questions is that of labourers. The lack of employer interest in people who work as labourers is evident in a number of ways. The turnover for labourers far exceeds that for people in other job categories. Employers were less unlikely to answer questions about labourers related to the importance of qualifications or for that matter the importance of experience. Likewise, they tended

3

not to answer questions about labourers when indicating whether they were planning to upgrade worker qualifications. What follows from the above is that problems and issues related to vocational training that might be of interest include the extent to which current training favours those in trades or clerical and sales or management and administration positions. Related to this is the extent to which the largely disinterested employer attitudes to those in labourer positions impede any investment in human capital for those in this part of the workforce. In terms of reliability of outcomes, the apparent lack of interest in providing comprehensive responses has complicated survey analyses and has undoubtedly given rise to inconsistencies in terms of estimating employer interest in formal qualifications vs. experience for those in varying industry sectors or in companies with differing numbers of employees. An instance of such inconsistencies is that at one point analytic outcomes are consistent with smaller firms not valuing qualifications for professionals while at another point outcomes are consistent with smaller firms valuing qualifications for clerical and sales staff. A major theme in responses to the open-ended question about credentials is that larger organisations are more likely to provide training. This seems unsurprising insomuch as larger organisations are more likely to have the scarce resources to do so and the ability to deliver cost-effective training due to economies of scale. Those employers who provided responses emphasised the value of previous work experience and experience in their industries. With the exception of labourer positions, relevant industry experience was held to be an important attribute during the recruitment process. This was the case for most positions across the industry groups surveyed, with the exception of jobs that required formal qualification, registration and accreditation. Experience was rated highly for most positions in the industry. Employers perceive general work experience to be a reliable criterion for potential employment. These results indicate on-the-job training to be one of the most popular pathways in manufacturing and property and business services, with construction the only industry to provide both on-the-job training and leave granted to complete qualifications. Survey questions also evaluated the degree of employers’ trust in qualifications. The two most highly factors affecting employer confidence or trust in qualifications included their personal experience of employees, and whether or not employee qualifications had been issued in Australia. Finally, interviews and focus groups encouraged respondents to speak frankly and freely on issues that concerned them, filling the gaps in understanding related to trends and peculiarities in the survey data. The key concern that arose was that employers across the industry groups did not always value VET qualifications. They criticised the VET system for not providing the necessary skills and experience that industry needs; ensuring access, availability and quality of training courses; and addressing skill shortages in the tight labour market.

4

INTRODUCTION The report presents an overview of the views of employers on the need for formal qualifications as opposed to skills and experience when employing skilled workers. The core research question addressed the potential substitutability between qualifications and modules and/or experience. The three specific objectives were to: discover the views of employers on the need for formal qualifications as opposed to skills/experience when employing skilled workers; evaluate the measures taken by employers to improve technical and work-related training; and identify any problems and issues related to vocational training and employment facing employers due to economic, or policy factors that affect access to education and training. Increasing reliance on formal qualifications by employers, the multitude of degrees possessed by some job seekers (double degrees, double Masters etc), and the hierarchy of desirable qualifications when recruiting, have created the phenomenon which has been labelled as ‘credentialism’ (Dore (1976, 1997). Haynes (2006) defines credentialism as the necessity for individuals to hold certain educational credentials, usually formal degrees or certificates, in order to obtain particular kinds of employment. To some extent, this study questions the above notion of credentialism by examining the degree of substitutability between formal qualifications, skills, and experience of employed workers in Queensland. It examines on the nexus between formal qualifications (secondary and post-secondary certificates, diplomas and degrees) and employers’ criteria of employability (including experience and relevant skills); which are used when searching and selecting employees for specific jobs in the following three major sectors of the economy in Queensland: (a) construction, (b) property and business services, and (c) manufacturing groups. By doing so, the study seeks to provide advice to policymakers that can assist them in deciding whether they should focus on providing the necessary work skills and other desired competencies through modules, or on ensuring that qualifications are completed when necessary. Answering the core question of substitutability involved identifying employers’ views on the value of formal qualifications, VET modules and short courses, versus specific competencies and skills at the entry point of a given workplace environment. This study uses the Department of Education, Employment and Industrial Relations’ (2008) definition of ‘qualification certification’ to define formal qualifications: “A qualification awarded to a person on successful completion of a course in recognition of having achieved particular knowledge, skills or competencies”. Qualifications are those included in the Australian Qualifications Framework1. Those people who partially complete formal qualifications undertake an accredited module(s), which the Department defines as “a unit of education or training completed on its own or as part of a course. Modules may also result in the attainment of one or more units of competency”. The hierarchy of occupations assumes that people with formal qualifications fill all skilled jobs and many semi-skilled jobs. Anecdotal evidence suggests however, that there are significant numbers of under-qualified or unqualified people working as tradespersons and associate professionals. This may be due, in part, to the presence of older workers who entered the labour market at a time when informal training was more common; or it may reflect a

1

See http://www.aqf.edu.au/aboutaqf.htm

5

current intake of under-qualified or unqualified people into specific occupations where shortages of qualified job applicants exist (for example, in the construction and manufacturing industries). The presence of significant numbers of such workers in skilled occupations seems at odds with an apparent trend towards credentialism both in Australia and globally. Currently in Australia, as elsewhere in developed countries, the majority of labour market entrants now possess secondary and post-secondary qualifications, and the percentage is increasing (OECD, 2007). This study has demonstrated that employers highly value previous work experience and experience in their industries. They highly value qualifications when recruiting applicants for jobs that require formal qualifications, registration and accreditation. However, some employers did not always value VET qualifications because of a lack of congruence between industry needs and government policy reforms. In all cases, employers want appropriately skilled workers. Where there is difficulty in obtaining workers with the necessary experience and/or qualifications, some industries have no alternative but to take in unqualified and unskilled workers and to provide on-the-job training. This creates a niche for the provision of general employability training. A major policy question from an economic perspective is to what extent there may be ‘over-education’ in the Queensland’s VET sector. If employers are happy to hire under-qualified and unqualified people with necessary skills and experience, then additional training content to make up a full qualification may represent an inefficient use of resources. Alternatively, the additional knowledge and skills that workers get from qualifications may produce gains to both individuals and employers. Therefore, an important issue for policymakers is whether employers are placing more emphasis on qualifications or more emphasis on experience and skills when filling jobs that normally require vocational qualifications; taking into account low unemployment and high labour demand in some industries and in some geographic areas. This report now moves to the Literature Review section that explains ‘credentialism’ and presents what is known about the recruitment behaviour of employers in Australia. The Quantitative Outcomes section presents overall results, followed by sections that draw together findings from the survey, interviews and focus groups to address each research question. The Conclusion section summarises these findings in a way that provides timely advice to policymakers about the potential substitutability between qualifications and modules and/or experience. A separate document of appendices accompanies this report. This supporting document includes data collection tools, results tables for the descriptive and inferential analysis by research question, and more detailed outcomes from the interviews and focus groups.

6

LITERATURE REVIEW The literature relating to human capital theory demonstrates that education consistently emerges as the prime human capital investment (Sweetland, 1996). Human capital refers to the productive capacities of human beings as income producing agents in the economy. Human capital research has found that education and training raises the productivity of workers by imparting useful knowledge and skills; improves a worker’s socio-economic status, career opportunities and income (Becker, 1964, 1994; Carnoy, 2000; Levin, 1987; Psacharopoulos, 1994; Schultz, 1971; Saha, 2005); and plays a significant role in driving overall economic performance. Until recently, there has been consensus that workers need formal qualifications (credentials) to secure particular kinds of employment and employers predominately use credentials when recruiting workers. Dore (1976) and Collins (1979) laid the foundation for the current credentialist framework. In his Diploma Disease, Dore (1976) examined the emergence of credentialism from the 1950s onwards, while Collins (1979) in his The Credential Society, offered his sociological theory of credentialism. Collins (1979) refers to credentialism as the need for highly qualified professionals in certain industries, labelling the phenomena as ‘credential capitalism’ (Collins, 1979). It is widely believed that employers use education credentials to screen and select workers because they believe that better-educated workers can be trained for specific jobs more quickly and at a lower cost than their less-educated peers (Throw, 1975; Davis, 1981; Brown, 2001; Walter, 2004). Collins (1979) differentiated between ‘credential capitalism’, which is characterised by the traditional laissez-faire attitude toward individual competition in the credential marketplace, and ‘credential Keynesianism’, which posits that education creates an artificial, yet economically useful credential currency - to offset deficiencies of aggregate demand. Credential capitalism assumes that one should get as much education as possible in order to facilitate one’s career advancement, by out competing others, at whatever level it takes. Credential Keynesianism, on the other hand, assumes that both investment in education and training and the credentialing of occupations would be encouraged. In markets where job skills are preferred to qualifications, Collins (1979) called for new types of credentials in order to adapt to market forces and accept employers’ on-the-job training programs: New types of credentials are proposed because the public has lost confidence in the value of the old types. Hence the inflationary struggle for credentials seems to be building up in new directions…we also hear of a massive expansion of internal and external credentialing in the business school sector. Skilled trades, contractors, and realtors continually establish more restrictive licensing programs, usually based on their own formal training requirements. Business corporations have established their own training programs. 2 Brown (2001) found that in the 1970s economists had developed two parallel theories within the human capital framework. Educational signalling theories saw students “as engaging in ‘defensive’ degree posturing in labour market queues; that is, they sought to keep from losing ground relative to degree holders” (p. 22). ‘Filtering’ and ‘screening’ theories focused on employers’ needs to narrow down large pools of applicants and make predictions about future employees’ capacity for productivity. Brown (2001) re-examined credentialism through the expansion of access to higher education and the proliferation of formal degree requirements for entry to employment. He linked credentialism to the relationship of educational expansion to economic growth and the relative importance of technical skills versus occupational status-group cultures in degrees and recruitment. He also assessed the trajectory of credentialism in light of potential policy reforms, market crises, and State interventions in the future.

2

http://home.earthlink.net/~fheapblog/id14.html

7

Guile (2003) argued that the challenge of the knowledge economy is to produce new knowledge and skills. Overcoming current credentialist approaches involves rethinking what is meant by ‘learning’. He pointed to the limitations of the notion of the ‘learning society’ and the emphasis on the accumulation of credentials as a strategy to respond to global economic and technological development. Walters (2004) found that debates about credentialism has focused largely on issues such as earnings, underemployment, and productivity, while less empirical treatment had been given to the issue of skill utilisation. The global economy, market forces, competition, and increasing completion rates of secondary and postsecondary education in developed countries may well have created reliance on qualifications and the qualification escalation phenomenon of credentialism. Keating et al. (2005) suggest employers are motivated by ‘human capital objectives’, ‘generic skills’ and ‘specific knowledge and skills’ denoted by academic qualifications: …employers are motivated by human capital objectives, particularly the potential for high levels of labour productivity. These objectives are mitigated by market interventions in the form of regulations, licenses, and industrial awards. They can also be mitigated by other objectives, such as status, especially for high-level qualifications…The holders of qualifications may be regarded as having qualities such as a work ethic, a capacity to complete tasks over a long period, and reliability and punctuality. Qualifications can signal generic skills as well as a capacity to learn. They also can signal more specific knowledge and skills (p. 36). The increased interest in the role and types of qualifications needed by countries to remain competitive and relevant in a global market led to many countries not only introducing formal qualifications for a range of industries but also establishing accreditation bodies to monitor standards. At the international level, there has been a high level of interest with regard to qualifications and training for the VET sector, with many European countries establishing relevant national accreditation agencies to monitor qualification and standards (Zajda, 2007). Australia has invested in national frameworks and accreditation processes and progressed down the path of endeavouring to ensure the proportion of the Australian population with qualifications is ever increasing. In the late 1980s and early 1990s, the Commonwealth Government introduced competency-based training. Competency-based training, delivered through industry-endorsed and nationally accredited modules, aims to equip people with the necessary skills and experience to perform an occupation. The underlying assumption is that what constitutes effective performance in the workplace can be modulised, codified and transmitted to learners. Keating et al. (2008) recognised the need to re-examine not only the significance of vocational qualifications but also the nexus between qualifications and the job market in the global economy. There has been relatively little research into the recruitment behaviour of employers in Australia (Keating et al., 2005), apart from Wooden et al. (1997), Foster et al. (2007), and Richardson (2007), which offer comprehensive analysis of qualification skills and experience. Wooden et al. (1997) placed ‘education’ after ‘attitude’, ‘skills’, ‘experience’ and ‘appearance’, as essential selection criteria. Existing recent studies in Australia provide some insight into attitudes to, and level of, knowledge about qualifications, skills and VET (Davies et al., 2001; Haukka et al., 2004; Keating et al., 2005; NCVER, 2008). In their report, Qualifications Use for Recruitment in the Australian Labour Market, Keating et al. (2005) demonstrated that VET qualifications are commonly recognised by Australian employers, with 50% of those employers surveyed using qualifications as both a sorting and screening mechanism. The report suggested that most of these employers prefer vocational education and training to university qualifications; and that VET qualifications appear to score more highly in the categories of job stability, reduced training, job skills and general skills (p. 27).

8

The 2007 Survey of Employers’ Use and Views of the VET System conducted by NCVER (2008) measured the extent of employer engagement and satisfaction with VET in meeting the skill needs of their workforce in Australia. Around one-third (33.3%) of those employers surveyed had jobs that require vocational qualifications, with many employers taking on people with these qualifications because they need to meet legislative, regulatory or licensing requirements, or to provide skills required for the job. Large employers were much more likely to take on people with VET qualifications than medium-sized employers and small employers. Only a small proportion of employers were dissatisfied with vocational qualifications, citing reasons such as not providing employees with the skills required for the job; not enough hands on/practical skills were being taught; and training is of poor quality and standard. The survey found that 22.1% of employers used nationally recognised training, and almost one-half (49%) used unaccredited training to provide the skills required for the job or maintain professional or industry standards. Reasons given by employers who chose unaccredited training over nationally recognised training were that unaccredited training was more cost-effective and the time to it was more convenient and flexible. Employment levels, economic growth and demand for skilled labour were high (and still are) when employers completed the above survey. Because of the economic climate in Queensland, many under-qualified, unqualified and unskilled workers are obtaining employment in jobs that normally require vocational qualifications. Some were working as unqualified tradespersons and associate professionals (Adult Skills and Knowledge for the Smart State, 2005; Queensland Skills Plan, 2006). In the case of Queensland’s construction industry, the high demand for skilled labour has contributed to an increase in apprenticeships (Trendle, 2007). However, apprentices are being hired more to fulfil current production requirements than they are to satisfy anticipated future demand for tradespersons. Many adults are beginning apprenticeships, providing a pool of unskilled workers for the construction and other industries. In summary, the existing research about credentialism indicates that employers, in general, primarily use qualifications to recruit workers. In Australia, employers who recruit workers with VET qualifications highly value such qualifications. However, in tight labour markets employers have little choice but to take on workers without the necessary qualifications. As a result, employers are increasingly recruiting workers based on their experience and skills, and in some situations, taking on workers with no relevant experience and skills.

9

METHODOLOGY This study employed quantitative and qualitative methods consisting of an empirical survey, individual interviews and focus groups. All three methods sought to gather data on the recruitment behaviour of employers; employer attitudes and opinions about the importance of qualifications and experience; how companies improve the skills of workers without formal qualifications; and what factors influence employers’ confidence/trust in qualifications. Research questions The study investigated six specific questions: 1. How much importance do employer place on formal qualifications and part qualifications (TAFE modules and short courses) when recruiting? 2. How much importance do employers place on experience when recruiting? 3. To what extent are qualifications and experience substitutes in the eyes of employers? 4. What measures do employers take to improve and/or upgrade skills of workers without formal qualifications? 5. What factors influence employers’ overall trust in qualifications? 6. What other problems and issues related to vocational training and employment do employers face? Data collection Methods included: •

a review of existing research on credentialism and the recruitment behaviour of employers;

•

a survey of 180 employers in Queensland; and

•

28 depth interviews and 3 focus groups.

Participants came from Queensland’s Construction Trades, Manufacturing, and Property and Business Services industries/sectors. This study retained the older 1993 ANZSIC classifications for Construction Trade Services and Property and Business Services. The 2006 version of ANZSIC removed Construction Trade Services as a sector and rearranged Property and Business Services into three new divisions/industries in the 2006 version. Although the purposive sampling approach for the survey aimed to achieve a quota of 300 employers in the above industries/sectors, only 180 employers responded to the survey. The study targeted seven occupations/occupational categories: technical, trades, clerical, sales and service, plant/machine operators, labourers, management and administration, and professional employees. Researchers assumed that most people employed in these occupations/categories are likely to benefit from vocational qualifications, and provide a good representation of people holding qualifications across all AQF levels. To test the relationship between formal qualifications, skills and work-related experience, a 15-item survey instrument (along with a 5-point Likert-scale for some key questions) was developed (Appendix A). The instrument included certain questions that were designed to test and analyse the employers’ attitudes to qualification and experience when recruiting; and their attitudes about part qualifications and modules that are considered relevant

10

and essential to the position. Other questions explored employers’ confidence and trust in qualifications, their support for on-the job training, and the extent of support given for employees to complete formal qualifications in certain areas. Researchers used semi-structured telephone interviews to explore responses by 28 employers to survey questions and to gain feedback beyond these questions (see Appendix B for interview questions and Appendix C for feedback). Each individual telephone interview took approximately 18 minutes to complete. The participants were employers and supervisors from 15 trades in Queensland’s Construction Trades, Property and Business Services, and Manufacturing industries/sectors (indicated by the letters A to T in Table 2 shown on the next page). Participants from the Construction Trades came from plumbing, painting, concreting, tiling, bricklaying, carpentry, structural steel erection, electrical installation services and air-conditioning installation services. Participants from Property and Business Services included architects, drafters, surveyors, engineers, computer maintenance, computer consultants, advertising, signwriting, graphic designers and business consultants. From the Manufacturing sector, participants came from machinery and equipment manufacturers, electronic equipment manufacturers, and food manufacturers. Participants were located across Queensland, including the Brisbane CBD, Fortitude Valley and Cannon Hill, and in the more regional/rural districts of Emerald, Dysart, Roma and Charters Towers. Researchers conducted three (3) focus groups in the Brisbane CBD, with 12 employers/supervisors participating. The intention of the focus groups was to provide additional information not readily available in the empirical survey, especially where the open-ended final question was optional, and in many cases not attempted by participants. The interviews were valuable in eliciting further information, however, in the focus groups the free-ranging discussion between similarly qualified (in terms of industry group involvement) individuals, allowed the participants to share and compare their experiences and concerns. Participants were from both small and large companies, and most were senior executives or business owners who were involved in recruitment and/or recruitment policy. Data analysis Multiway Frequency Analysis (MFA), followed by post hoc two-way Cross-Tabulations tables were used to examine the effect of type of industry and number of employees on items related to five of the six research questions (sixth research question an open-ended question capable of descriptive analysis only). MFA is a nonparametric test that can be used to examine relations among three or more discrete (i.e., nominal, qualitative, categorical) independent variables with two or more levels. MFA is a simplified version of Loglinear regression analysis and Logit analysis. MFA is an extension of the chi-square for goodness-of-fit technique. The goal of MFA is to discover whether there is an association among two or more discrete variables. A salient issue related to the application of MFA analysis to the current data set is that as with other chi-square statistic analytic procedures, sample size is important. A rule of thumb is that the sample size should approximate five times the number of cells in consideration. Further, a minimum cell size of five is required for reliable estimates. In the case of the present data set, the number of missing responses per item necessitated some collapsing of variables to meet the requirements. A particularly cogent consideration here is that approximately 40 participants responded to questions about plant and machine operators and about 50 to questions about the importance of labourers. A workable scenario for MFAs that resulted in viable cell counts involved collapsing across response categories, a methodological issue canvassed in more detail in the body of the result section. The qualitative data analysis process focused on identifying themes and patterns within the data and establishing the spread of responses. Individual participant data was evaluated, and then collated together to establish the scope of each theme and provide some indication of its strength. Findings supported the quantitative outcomes.

11

QUANTITATIVE OUTCOMES Demographic descriptives Locality Of the 180 participants, almost two-thirds (N=119) came from Queensland Metropolitan localities, another 30.5% (N=55) from Queensland regional localities, and the other 3.3% (N=6) from Queensland rural localities. To facilitate subsequent analyses, participants from regional and rural localities were merged to form a single category consisting of just over one-third (N=61) of the participants. Type of business Over 90% (N=167) of the participants came from the private (for profit) sector, rendering this variable unsuitable for cross-classificatory analysis. Number of employees Participants were asked to nominate the number of employees in their organisations, with response categories including 1-9, 10-19, 20-49, 50-99, and 100 plus employees. Given that only 5% (N=9) came from organisations with 50-99 employees, and less than 7% (N=12) came from organisations with a 100 or more employees, these two categories were merged to create a super-category of organisations with 50 or more employees Of the 180 participants, 54% (N=98) came from organisations with 1-9 employees, another 20% (N=36) came from organisations with 10-19 employees, 14% (N=25) came from organisations with 20-49 employees, and the remaining 12% (N=21) came from organisations with 50 or more employees. In order to facilitate Multiway Frequency Analysis, it became necessary to collapse these four categories into two comprised of those with 1-9 employees vs. those with 10 or more employees. Industry sectors Participants came from eight major industry sectors. As illustrated in Figure 1 on the next page, of the 180 participants, the largest percentages (33%; N=59) came from the construction and trades and business services (19%, N=33) sector. Smaller percentages came from the computer services (13%, N=23) and machinery and equipment manufacturing (10%, N=18) sectors. Finally, the smallest percentages of participants came from the technical services (3%, N=6), food manufacturing (7%, N=12) and electronic equipment manufacturing (6%, N=10) sectors. After entry into Optimal Scaling (nonparametric factor analysis), these eight industry sectors were merged to form five sectors, and prior to commencing MFA these five sectors were merged to form three super-sectors.

12

40.0%

Percent

30.0%

20.0%

33.1%

10.0%

18.5% 12.9% 10.1%

9.6% 6.7%

5.6%

3.4% 0.0% Construction trade

Business services

Technical services

Computer services

Marketing business

Food manufacturing

Machinry equip man

Electrnic equip man

Industry sector

Figure 1. Relative percentage of participants from eight industry sectors Associations between demographic variables (bivariate correlations) As indicated in the (nonparametric) Spearman’s Rho matrix of correlations below (Table 1 on the next page), 13 of the 48 correlations were statistically significant: Many of these trivially reflected the reality that participants from one industry were unlikely to report about other industry sectors. More interestingly, company location was negatively associated with computer services (Rho=-0.168) and positively with food manufacturing (Rho=0.235), consistent with the former being sited in metropolitan and the latter in regional and rural localities. Also, the number of employees correlated negatively with computer services (Rho=-0.214) and positively with food manufacturing (Rho=-0.191) consistent with computer services reporting fewer employees and food manufacturing more employees.

13

Table 1. Bivariate associations between select participant characteristics Company Construction location (Qld trade services 2gps) Company location (Qld 2gps)

Construction trade services

Business services

Technical services

Computer services

Business services

Technical services

Computer services

Correlation Coefficient

1.000

Sig. (2-tailed)

.

N

180


-.123

1.000

Sig. (2-tailed)

.101

.

N

178

178


.119

-.336(**)

1.000

Sig. (2-tailed)

.115

.000

.

N

178

178

178


-.133

-.132

-.089

1.000

Sig. (2-tailed)

.076

.080

.237

.

N

178

178

178

178


-.168(*)

-.271(**)

-.184(*)

-.072

1.000

Sig. (2-tailed)

.025

.000

.014

.340

.

N

178

178

178

178

178

Marketing & Food Bus Mngt manufacturing services

Machine & Electronic Equip manu Equip manu

Number of employees (4gps)

14

Marketing & Bus Mngt services

Food manufacturing

Machine & Equip manu

Electronic Equip manu

Company Construction location (Qld trade services 2gps)

Business services

Technical services

Computer services

Marketing & Food Bus Mngt manufacturing services


.051

-.229(**)

-.155(*)

-.061

-.125

1.000

Sig. (2-tailed)

.496

.002

.039

.421

.096

.

N

178

178

178

178

178

178


.235(**)

-.189(*)

-.128

-.050

-.104

-.087

1.000

Sig. (2-tailed)

.002

.011

.088

.506

.169

.246

.

N

178

178

178

178

178

178

178


.116

-.236(**)

-.160(*)

-.063

-.129

-.109

-.090

1.000

Sig. (2-tailed)

.124

.002

.033

.406

.086

.148

.231

.

N

178

178

178

178

178

178

178

178


-.071

-.172(*)

-.116

-.046

-.094

-.079

-.066

-.082

1.000

Sig. (2-tailed)

.348

.022

.122

.546

.212

.293

.384

.278

.

N

178

178

178

178

178

178

178

178

178

.068

.117

-.032

-.072

-.214(**)

-.035

.191(*)

.047

-.043

1.000

Sig. (2-tailed)

.363

.119

.669

.343

.004

.640

.011

.530

.571

.

N

180

178

178

178

178

178

178

178

178

180

Number of employees (4gps) Correlation Coefficient

* Correlation significant at 0.05 level (2-tailed) ** Correlation significant at 0.01 level (2-tailed)

Machine & Electronic Equip manu Equip manu

Number of employees (4gps)

15

Optimal scaling The associations between locality, number of employees, and industry sector were examined with SPSS Optimal Scaling. Multinominal1 Optimal scaling does Homogeneity analysis, which quantifies nominal (categorical) data by assigning numerical values to the cases (objects) and categories. The goal of Homogeneity analysis is to describe the relationships between two or more nominal variables in a low-dimensional space (i.e., minimum of two dimensions) containing the variable categories as well as the objects in those categories. Objects within the same category are plotted close to each other, whereas objects in different categories are plotted far apart. Each object is as close as possible to the category points for categories that contain that object. Homogeneity analysis can be viewed as principal components analysis of nominal data. Homogeneity analysis is preferred over standard principal components analysis either when linear relationships between the variables do not hold or when variables are measured at a nominal level. 2. Industry sector Company location (Qld 2gps) Number of employees (4gps)

3

2 10-19 empls Electronic Equipment 1

50+ empls

Technical Construction/Trade

Qld metro Machinery/Equipment 0

Qld reg/rural 20-49 empls 1-9 empls Business Computer Marketing/Business

-1

Food Manufacturing

-2 -2

-1

0

1

2

3

Figure 2. Spatial representation of associations between selected participant demographic characteristics As illustrated in Figure 2, participants were recruited from eight industries in two major locations, and came from firms with varying numbers of employees. The Dimension 1/Dimension 2 labels associated with the horizontal and vertical axes relate to the plotting of this nonparametric factor solution in two-dimensions. These preset dimension labels are not necessarily relevant to specific analyses. The simplest option is to hide them (See above). The left and right halves of the illustration related to Queensland Metropolitan and Queensland rural localities. The four quadrants are segregated in terms of the number of employees. So, the lower left hand quadrant groups industry sectors in the metropolitan region with 1-9 employees. These include participants from computer services, business services, and marketing and business management services. The top left hand quadrant groups industries in the metropolitan region with 10-19 employees. These include participants from electronic equipment manufacturers, from technical services, and from construction and trades. The lower right quadrant groups industry sectors in the

16

regional/rural sector with 20-49 employees. These include participants from food manufacturing. Finally, the upper right quadrant groups industry sectors in the regional/rural localities with 50 or more employees. These include participants from machinery and equipment manufacturing. Optimal Scaling analyses provides a rationale for collapsing across low frequency response categories. Researchers collapse categories within variables as a prelude to entering demographic variables as variables in univariate or multivariate parametric analyses. They do so to minimise the likelihood of such analyses becoming unstable (and, hence, unreliable) because of the presence of one or more cells (a cell being equivalent to a single or conjoined response category such as older females) with fewer than 6-10 respondents or one or more cells with fewer than 10% of participants. Hence, collapsing across response categories can be a desirable intermediary step after preliminary categorical level analyses, aided by frequency and contingency table analyses of participant numbers. In this case, the close spatial proximity of the technical services vs. electronic equipment manufacturing sectors suggests that these could be merged to form a composite variable for the purpose of subsequent analysis. Likewise, the spatial proximity of business services vs. marketing and business management services sectors suggests that these two sectors could be merged. Finally, the relative spatial proximity of food manufacturing vs. machinery and equipment manufacturing sectors suggest that these two sectors could be merged.

40.0%

Percent

30.0%

20.0%

33.1% 28.1%

10.0%

16.9% 12.9% 9.0%

0.0% Construction

Bus & Mktg/Mngt

Tech & elect equip

computer

Food & mach/equip

Industry sectors (5gps)

Figure 3. Relative percentage of participants from five composite industry sectors As illustrated in Figure 3, the outcome redistributed the percentages of participants per industry sector such that the smallest industry sector now boasts 9% (N=16) participants, and the four other sectors include percentages of participants that range from 13-33%. This outcome is more conducive to reliable (stable) inferential outcomes for analyses involving the three composite variables (Business services & Business Marketing & Management; Food manufacturing & Machinery/Equipment manufacturing; Technical services & Electronic equipment manufacturing).

17

At the point of undertaking MFA (See below) it was deemed necessary to collapse these five sectors into three supersectors, comprising construction, computers and electronics, and other manufacturing and business services and management organisations (See below). Employment descriptives Participants were asked about the number of people employed across a range of work categories: employed currently employed 12 months ago who left in last 12 months hired as replacements hired to fill fresh positions. These employment related variables were regressed on one another systematically to identify multivariate outliers. Two cases were identified as potential outliers but not excluded from analyses. Table 2. Number of current employees in various categories (missing responses excluded)

Technical

Trades

Clerical & sales

Plant & machine operators

Labourers

Valid responses

100

136

134

81

98

140

93

Missing

80

44

46

99

82

40

87

4.45

8.56

5.26

10.16

7.36

6.99

6.55

12.637

20.465

12.856

59.488

25.512

24.080

19.670

Mean SD3

Management & Professionals administration

As shown in Table 2, the large number of responses classified as missing raises a more serious issue in analysing the responses to these items. Given that an appropriate positive response to each of these employment related categories comprises a positive integer, one simplifying assumption is that missing responses amount to a statement that no employees were employed in those categories. With that in mind, missing responses were recoded as zero for subsequent analyses. This transform decreased the average number of employees reported per category (because of the large number of participants not reporting employees in specific categories but it did make all 180 cases available for multiple variable analyses. This procedure was followed for each of the five major contexts for reporting number of employees in each of the seven listed categories (Technical, Trades, Clerical and sales, Plant and machine operators, Labourers, Management and administration, Professionals).

3

Standard deviation

18

A rationale for the transform is that it increases the likelihood of obtaining not only more conservative (due to decreased variance) but also more reliable inferential outcomes. Table 3. Number of current employees in various categories (missing responses coded as zero)

Technical

Trades

Clerical & sales


Labourers

180

180

180

180

180

180

180

0

0

0

0

0

0

0

Mean

2.47

6.47

3.92

4.57

4.01

5.43

3.38

SD

9.656

18.151

11.318

40.091

19.136

21.419

14.479

Valid responses Missing


Table 3 shows that participants were most likely to report that people were currently employed in management and administration and positions in the trades. They were least likely to be employed in technical positions. Table 4. Number of employees in various categories 12 months ago

Valid responses Missing Mean SD

Technical

Trades

Clerical & sales


Labourers

Management Professionals & administration

180

180

180

180

180

180

180

0

0

0

0

0

0

0

2.66

4.32

3.38

3.76

3.41

4.69

3.11

15.742

12.839

8.858

30.050

17.268

20.793

14.258

Table 4 shows that participants were most likely to report that people were employed 12 months ago in management and administration and positions in the trades. They were least likely to be employed in technical positions. Table 5. Number of employees in various categories who left in last 12 months

Technical

Trades

Clerical & sales


Labourers

180

180

180

180

180

180

180

0

0

0

0

0

0

0

Mean

1.02

1.21

.78

1.49

8.12

.54

.40

SD

7.667

3.584

4.178

15.022

68.596

2.393

1.928

Valid responses Missing

Management Professionals & administration

19

Table 5 shows that participants were most likely to report that labourers had left in the last 12 months. Employees who had left were least likely to have been employed in management and administration or as professionals. Table 6. Number of employees in various categories who were hired to replace staff in last 12 months


Technical

Trades

Clerical & sales


Labourers


180

180

180

180

180

180

180

0

0

0

0

0

0

0

.46

1.01

.71

1.48

8.47

.44

.38

2.009

3.483

4.160

15.022

68.751

2.270

1.918

Table 6 shows that participants were most likely to report that labourers had been hired to replace staff in last 12 months. They were least likely to have been hired to replace staff in technical positions, management and administration, or in professional positions. Table 7. Number of employees in various categories who were hired in new positions in last 12 months


Technical

Trades

Clerical & sales


Labourers


180

180

180

180

180

180

180

0

0

0

0

0

0

0

.51

1.03

.82

1.25

.70

.78

.57

3.325

4.639

4.574

11.586

4.886

3.975

3.336

Table 7 shows that participants were most likely to report that employees had been hired to fill new positions in the trades or as plant and machine operators in the last 12 months. They were least likely to have been hired to fill new technical or professional positions. It is clear from the above that a fairly homogenous pattern prevails across the five contexts for reporting number of employees. That is, people employed as labourers were most likely to leave or be replaced. People employed in trades or management and administration positions were most likely to be employed currently or in 12 months ago. The low turnover in people employed in technical positions was also notable. This pattern was examined further via a series of Spearman’s Rho correlation matrices (See Appendix B for matrices).

20

Associations between variables Summary of bivariate correlations outcomes Based on statistically significant correlations, those who currently employed plant and machine operators also employed people in technical, trades, and clerical and sales positions. Those who employed labourers also employed trades and plant and machine operators. Those who employed people in management and administration positions also employed people in trades, clerical and sales, plant and machine operators, and labourers. Finally, those who currently employed professionals also employed people as plant and machine operators and in management and administration positions (See Appendix B for more detail). Those who employed people in clerical and sales positions 12 months ago also employed people in trades positions. Those who employed plant and machine operators 12 months ago also employed people in technical and trades positions. Those who employed labourers 12 months ago also employed people in technical, trades, clerical and sales, and plant and machine operator positions. Those who employed people in management and administration positions also employed people in trades, clerical and sales, plant and machine operator, and labourer positions. Finally, those who employed people in professional positions were more likely to employ people technical, plant and machine operator, labourer, and management and administration positions (See Appendix B for more detail). Those reporting people who left in any specific category of position in the last 12 months were also likely to report people leaving from every other category as well (See Appendix B for more detail). Those reporting people who replaced in any specific category of position in the last 12 months were also likely to report people replaced in every other category as well (See Appendix B for more detail). Those reporting people hired to fill new positions in any specific category of position in the last 12 months were also likely to report people hired into new positions in every other category as well (See Appendix B for more detail). One way to summary the above series of bivariate matrices, is that patterns of employment aren’t specific to categories of positions. Methodological issues related to inferential statistics Given the types of variables under consideration (mostly categorical or ordinal variables), Loglinear regression, and more specifically Multiway Frequency Analysis (MFA) seems an appropriate methodology for inferential analyses. As stated previously, MFA (SPSS) does stepwise (backwards) multivariate (Loglinear) regression where the multivariate IVS are categorical. According to SPSS, it can be used as a screening tool via Model Selection to identify significant variables and interactions for subsequent contingency table (chi square) analyses. The result section provides an indication of which variables and interactions form part of the final model, together with various measures of the relative strength of effects. In this case, the MFA seemed an appropriate procedure for evaluating the significance of associations between industry sector, size of industry, and the importance of qualifications for specific position categories as per the survey questionnaire.

21

Accordingly, MFAs together with follow-up contingency tables were used to examine the significance of interactions related to the importance of qualifications as examined via each of five research questions.4 These analyses focused on survey questions 7 and 8 for outcomes related to research Q1 (importance of qualifications), survey question 9 for outcomes related to research Q2 (Importance of experience), survey question 10 for outcomes related to research Q3 (Substitutability), survey question 11 and 12 for outcomes related to research Q4 (Measures to provide training), and survey question 13 for outcomes related to research Q5 (Employer trust in qualifications).. A salient issue related to the application of MFA analysis to the current data set is that as with other chi-square statistic analytic procedures, sample size is important. A rule of thumb is that the sample size should approximate five times the number of cells in consideration. Further, a minimum cell size of five is required for reliable estimates. In the case of the present data set, this necessitated some collapsing of variables to meet the requirements. Had the 180 participants provided responses for every variable of interest that ideal response rate would have made it possible to undertake MFAs for a maximum of 36 cells. However, it is clear that combining eight industry sectors with five response categories related to the importance of qualifications (Q7-Q9) produces a 40 cell matrix, which requires a larger sample (i.e., 200 plus participants) than at present would be available even for an 100% response rate. A particularly cogent consideration here is that only approximately 40 participants responded to questions about plant and machine operators and about 50 to questions about the importance of labourers. At the other end of the range, around 120 participants responded to questions about the importance of qualifications for clerical and sales positions. A workable scenario for MFAs that results in viable cell counts involves collapsing: the eight industry sectors into one original plus two merged sectors (construction, computers/electronic equip, & manufacturing (food, machinery, business services, Marketing & management) the original five response categories for number of employees into two groups (1-9, 10+) the 5-response category Likert scales for estimating the importance of qualifications into two dichotomous categories, as follows: Less important (Not required, Very Little, Moderate) vs. More important (A great deal, Mandatory). These drastic transforms lower the required number of responses per variable to 60 (3*2*2=12). While two of the position categories (plant & machine operators, labourers) don’t quite meet this criterion, their cell size sufficiently approximate the minimum of 60 for the purpose of MFA analysis. MFA generates a saturated model, which includes all interactions and main effects. From the perspective of the present investigation into the importance of qualifications, 3-way and 2-way interactions related to importance of qualifications are of particular interest. Accordingly, the focus of reporting in the result section was on statistically significant interactions that involved estimates of the importance of qualifications (e.g., industry sector * importance or number of employees * importance) With this in mind, statistically significant interactions with specific measures of the importance of qualifications were reported and further examined if significant via contingency tables (chi-square statistics). The next sections report MFA outcomes of interest together with the detail provided by follow-up contingency tables. Findings from the interviews and focus groups support these outcomes. It is worth noting that the locus of effects of interest as per the follow-up contingency table was identified by examining the standardised residual per cell.

4

An exception to the above is that outcomes based on responses to the open-ended survey Q14 that canvas responses to research Q6 (Other problems and issues) were only reported descriptively.

22

THE IMPORTANCE EMPLOYERS PLACE ON QUALIFICATIONS WHEN RECRUITING Descriptive statistics Participants were asked to rate the importance placed on formal qualifications by their organisations by ticking one of five Likert scale response categories (Not required (1), Very little (2), Moderate (3), A great deal (4), Mandatory (5). Table 8. Rated importance of formal qualifications across seven categories of positions

Technical

Trades

Clerical & sales


Labourers


Valid responses

76

104

119

41

52

111

64

Missing

104

76

61

139

128

69

116

Mean

3.25

3.98

2.63

3.07

1.83

3.13

3.73

SD

1.396

1.182

1.088

1.694

.901

1.214

1.483

It is worth noting that, as shown in Table 8, only a minority of the 180 participants answered the question in regard to four (i.e., technical, plant & machine operators, management and administration, and professionals) of the seven position categories. Even where a majority answered the question, one-third to one-quarter participants did not provide a response. The unusual level of missing responses creates problems for subsequent inferential analyses. These missing scores can neither readily be replaced nor ignored. If replaced, say, by the mean or by the minimum score, then the resulting distributions would become highly kurtotic and tend to lack variability, which distorts analytic outcomes. If ignored, then it’s likely that a rather small subset of cases would remain in any analysis that included one or more of these variables, with negative effects on the reliability and generalisability of statistical outcomes. For those who did respond, as indicated above, the average scores (Table 8) were computed using Likert scale responses encoded via the numbers 1-5 (1=not required and 5=Mandatory). The importance of formal qualifications was rated most highly for people in trades and professional positions (Moderate to A great deal) and least highly for those in labourer positions (Not required to Very little). Table 9. Rated importance of part qualifications across seven categories of positions

Technical

Trades

Clerical & sales


Labourers


Valid responses

74

103

113

41

50

105

57

Missing

106

77

67

139

130

75

123

Mean

2.89

3.39

2.35

2.34

1.82

2.64

2.49

SD

1.320

1.206

.981

1.296

.896

1.128

1.338

23

As shown in Table 9, the importance of part qualifications was rated most highly for people in trades (Moderate to A great deal) and least highly for those in labourer positions (Not required to Very little). In preparation for MFA, the five response categories for level of importance of formal and part qualifications were collapsed two form a level, Not important (Not required, very little, Moderate) vs. Important (A great deal, mandatory) response scale ( See Appendix B for Frequencies). Inferential statistics Formal qualifications MFA analysis indicated the interaction between: industry sector and the importance of qualifications for plant and machine operator positions to be statistically significant (p