Perceptions of desirable graduate competencies for science and technology new graduates

Abstract

Work‐integrated learning (WIL) programmes that combine on‐campus classroom‐based study with off‐campus authentic work experience are a growing area of interest internationally. Despite widespread practice of WIL, there are few reports that shed light on appropriate pedagogies for the work experience in particular. As with any form of education, providers hold certain views as to desirable outcomes in terms of graduate profiles and of desirable graduate competencies. A complication for multi‐party WIL programmes is that educational stakeholders (e.g., staff working in tertiary education provider institutions and employers) may hold different views as to desirable graduate competencies. Here we argue that an understanding of stakeholder views of desirable graduate competencies is an essential prerequisite of pedagogical design. The research reported here is an intrinsic case study and comprised an investigation of perceptions of 24 desirable graduate competencies for new science and technology graduates entering the workforce both today, and in ten years’ time. Stakeholders for four sector stakeholder groups (n = 458): undergraduate students (n = 71), recent graduates (n = 143), employers of graduates (n = 172), and faculty (n = 72), were surveyed using a previously reported and validated instrument. The research findings suggest that science and technology stakeholders see all 24 competencies as desirable, and see the importance of all skills and some skills in particular as likely to increase in ten years’ time. Despite emphasis on cognitive and technical skills (often termed ‘hard’ skills), the single most desirable skill is ability and willingness to learn, a behavioural skill (often termed ‘soft’ skills). It is proposed that classroom‐based instruction is unlikely to produce graduates with the desired skills, and that work‐integrated learning may have a role to play in the development of graduate competencies.

Notes

1. The terms ‘soft’ and ‘hard’ are perhaps more common than cognitive and behavioural, but also a little colloquial. For the sake of consistency we use cognitive and behavioural throughout this paper.

2. Given the scale of data obtained here (see below), non‐parametric statistical tests would technically be more appropriate. Here we have chosen to use the parametric equivalent tests. Parametric statistics work on the assumption that the data fall into a normal‐shaped type distribution. The statistical tools used for non‐parametric tests are generally less powerful (for example, they may be less sensitive in detecting differences in means). For these reasons most education (and social science) research falls into the category of non‐parametric statistics, but nonetheless employs parametric statistics. Our stance is that this finer level of detail enables researchers to better evaluate any educational significance of quantitative data (see Carver, 1978, 1993); it is up to the reader to judge if we have achieved that aim.

3. As these data are ordinal level, only estimated means can be computed, and the results can be used to only show order of ranking of competencies.