Academic success and failure : Student characteristics and broader implications for research in higher education

The paper reports on the findings of a funded research project which investigated factors related to academic success and failure in a Faculty of Arts. The project initially aimed to explore failure which resulted when students remained enrolled but submitted no items for assessment. The initial aim was to investigate whether the problem resided essentially in characteristics of non-completing students, or whether there were institutional factors associated with courses which facilitated failure. However, the wealth of data made available to the researchers also permitted the analysis of factors related to academic success. The paper discusses the student characteristics which were found to be clearly related to academic success and failure – including gender, university entry score, and mode of enrolment. The paper also explores implications of the findings for “mass” higher education. It concludes that a profitable approach to research on student success and failure might be to investigate the strategies which students develop to effectively cope (or not cope) with competing life demands. Background Initially, this study was undertaken to explore a particular type of academic failure in the Faculty of Arts at a large commuter university. It had been noted that a proportion of students recorded grades of FNS – that is, these students enrolled in a course but submitted no work for assessment and hence failed. A particular aim of the project was to investigate whether the problem resided essentially in personal characteristics of noncompleting students, or whether there were institutional factors associated with courses or programs that facilitated failure, or a combination of both. The project soon broadened to incorporate other forms of academic failure because early analyses suggested that differing demographic and course factors might be associated with different types of failure. In addition to FNS, factors associated with the grades F (“normal” failure) and WF (failure resulting from late course withdrawal) were also explored. It was apparent that these combinations of fail grades at times could swell the proportion of students failing in courses to serious levels (20%-25%). The wealth of data made available to the researchers also meant that factors related to academic success could be explored and we report briefly on those findings in the paper. Nevertheless, academic failure has remained the focus of the project. Actually, it is difficult to identify a body of extant research that deals principally with student failure in university settings. It is as if the topic is something of an embarrassment for institutions – and so a mainstream focus of scholarly attention has not developed. This despite the ubiquity of failure, as Peelo (2002, p.7) has noted; While ‘failing’ or ‘failure’ is part of everyday experience in universities, it hardly seems to matter in the education literature. There have been few attempts to understand ‘failing’ as an ever-present phenomenon in higher education. In fact, as she also goes on to note, the extensive literature on student retention, either implicitly or explicitly, does place academic success/failure at the centre of research and theorising about the problem of attrition. In the attrition literature, the focus has been either on the failure of students to successfully integrate themselves into the institution (Tinto, 1987) or, on the other hand, on the failure of the institution to integrate diverse student populations, their lifestyles and subcultures (Braxton, 2000; Peelo & Wareham, ETL Conference, 2004, Logan Campus, Griffith University: Kerry Wimshurst & Richard Wortley 2002; Yorke, 1999). In the recent literature there also has been a tendency to collapse so-called “involuntary attrition” (institutional exclusion arising from academic failure) into the same category as “voluntary attrition” (student-initiated withdrawal from their programs). At the same time, a parallel area of research and theorising has drawn attention to the multiple transitions that, it is said, young people increasingly must negotiate. The argument is that the majority of young people, more than at any other time in modernity, now feel compelled to balance commitments to employment, higher education, and other personal and relationship matters. At times, the stresses and strains of coping with multiple transitions and their crosscurrents impact adversely upon students, resulting in academic failure. For broad discussions of this apparent sea change in the outlook of young people, see the youth studies of Dwyer & Wyn (2001); Ball, Maguire and Macrae (2000); and for higher education, McInnis, James and Hartley (2000). Even so, it is not clear how helpful such conflations of perhaps very different processes and experiences might be in the long run in unravelling the problem of academic failure. In this paper, because of limited space, we concentrate on student factors and plan to disseminate elsewhere the findings which relate to institutional factors. Nevertheless, in the meantime, it should be noted that our findings confirm the need to explore both student and institutional factors. As Yorke’s (1999) work in the UK has shown, student attrition – whether voluntary or involuntary is rarely the result of one single factor, but rather failure (broadly defined) tends to result from a combination of factors. The research reported below identified demographic and biographical factors that were significantly related to academic success and failure; however, there was also evidence of pronounced differences in grading practices between different components within the same Faculty (programs, courses and schools). Method Data covering all students who were enrolled in Faculty of Arts courses (subjects) offered by any of the five Schools for the three years 1998-2000 (inclusive) were entered into files for statistical analysis. The five Schools offering courses were: Arts (ART), Criminology and Criminal Justice (CCJ), Film, Media and Culture (FMC), Humanities (HUM) and Languages and Linguistics (LAL). The data files covered biographical details, academic results, program information and course data for each student. Initially, this translated into a data set of 54,590 cases – where one case represented one student grade for a course. After data cleaning, this resulted in a data set of 11,264 students enrolled in 221 undergraduate courses that formed the basis of the analysis. The statistical procedure used was a multivariate analysis of variance (MANOVA). The dependent variables for the analyses of both the student and course data were the proportion of grades FNS, F, WF, PC, P, C, D, and HD. The independent variables consisted of a range of student and course related characteristics available from the standard university records. The usefulness of the MANOVA procedure is that, in the analyses, all effects are adjusted for one another. That is, the results reflect the unique variance of each predictor or variable. If a number of student characteristics emerge as significant in terms of predicting student grades (as discussed below), then each is significant in itself, independent of any compounding interactions between the variables. In addition, we sought feedback from course convenors on the findings of the statistical analyses. While the staff survey was a relatively structured instrument, there was the opportunity for them to provide more open and reflective comments. Twenty-four convenors replied which constituted a 46.0% response rate from those who were contacted.