Coursework versus examinations in end-of-module assessment: a literature review

Abstract

In the UK and other countries, the use of end-of-module assessment by coursework in higher education has increased over the last 40 years. This has been justified by various pedagogical arguments. In addition, students themselves prefer to be assessed either by coursework alone or by a mixture of coursework and examinations than by examinations alone. Assessment by coursework alone or by a mixture of coursework and examinations tends to yield higher marks than assessment by examinations alone. The increased adoption of assessment by coursework has contributed to an increase over time in the marks on individual modules and in the proportion of good degrees across entire programmes. Assessment by coursework appears to attenuate the negative effect of class size on student attainment. The difference between coursework marks and examination marks tends to be greater in some disciplines than others, but it appears to be similar in men and women and in students from different ethnic groups. Collusion, plagiarism and personation (especially ‘contract cheating’ through the use of bespoke essays) are potential problems with coursework assessment. Nevertheless, the increased use of assessment by coursework has generally been seen as uncontentious, with only isolated voices expressing concerns regarding possible risks to academic standards.

Keywords:

• coursework
• degree class
• ethnicity
• examinations
• gender

Acknowledgements

I am grateful to Michael Bassey, Guillaume Cabanac, Jude Carroll, Geoffrey Crisp, Åge Diseth, Graham Gibbs, David Gijbels, James Hartley, Jón Torfi Jónasson, Sigurður Jónsson, Marcin Kozak, Erika Löfström, Lisa Lucas, Stephen Newstead, Anna Ólafsdóttir, David Putwain, Bart Rienties, Chris Smith, Maria Hvid Stenalt, Keith Trigwell and Belinda Tynan for their advice and suggestions and to Paul Ginns, Wilbert McKeachie, Chris Rust and Denise Whitelock for their comments on previous versions of this paper.

Notes

1. There is an extensive body of research on the impact of anxiety in examinations and other test situations (see Zeidner 1998, 2007). Meta-analytic reviews have clearly established that such ‘test anxiety’ is negatively correlated with attainment in higher education (e.g. Hembree 1988; Richardson, Abraham, and Bond 2012). However, Richardson et al. commented that there was insufficient data for them to compare its impact in coursework and examinations, and so the idea that it is reduced with coursework remains to be tested. Ryle (1969, 103) argued that ‘any alternative system of examining is likely to produce … its own varieties of stress’. In secondary education, the workload and deadlines involved in coursework can indeed result in a different kind of stress (Putwain 2009).

2. International students differ from home students in cultural and linguistic terms as well as ethnicity. Two studies examined international students taking MBA programmes at UK universities. Smith (2011) compared students with English as their first language (mostly from the UK) and students with English as a second language (all international students). The former did better than the latter in examinations but not in coursework. Kelly and Moogan (2012) also found that UK students did better than international students, but the difference was mainly in their coursework marks rather than in their examination marks. The discrepancy in the results of these two studies may be due to contextual factors or to the specific forms of assessment that were used.

3. Gender and ethnicity are the main demographic characteristics that have been discussed in the literature on coursework vs. examinations, but there is a suggestion that age is another factor that might be considered. At the UK Open University, the Older Students Research Group examined academic attainment among students who had been registered in the undergraduate programme in 1982 (Clennell 1984; Ryan 1985). The pass rates in continuous assessment were 75% for students under the age of 60, 78% for those aged 60 and over and 77% for those aged 70 and over; but the pass rates in the end-of-module examinations were 90% for students under the age of 60, 85% for students aged 60 and over, and 77% for those aged 70 and over. In short, older students appear to encounter some difficulty when assessed by examinations, but they do slightly better than younger students when assessed by coursework.

4. There is a gap in the literature regarding the advantages, disadvantages and psychometric properties of particular assessment tasks (essay questions, short answers, multiple-choice questions, etc.) when used in coursework as opposed to examinations. Heywood (1977, 1999), Brown (1997), Wakeford (1999, 2003) and Light, Cox, and Calkins (2009, 200–236) all reviewed these topics in some detail, but the classic studies to which they referred were mainly carried out in the context of traditional examinations. None of these authors considered whether the reliability, validity or appropriateness of these tasks was different when they were used in coursework rather than in examinations.

5. An anonymous reviewer of this article argued that ‘plagiarism is predominantly a US/UK rather than a North European phenomenon’. According to colleagues (and colleagues of colleagues), however, plagiarism in higher education appears to be a general problem in most countries in the European Union (see Glendinning 2013, forthcoming), including Belgium, Denmark, Finland (Löfström and Kupila 2013; Tuhkanen 2014), France (Maad 2013), Germany (Webber-Wulff 2014), Lithuania, the Netherlands (Rienties and Arts 2004), Poland and Sweden, as well as Australia (Gullifer and Tyson 2010, forthcoming), Hong Kong, Iceland, New Zealand (Walker 2010), Norway and Singapore, and anti-plagiarism software is widely used. In most countries, the extent of the problem is not formally documented. However, in Sweden, the University Chancellors’ Board (2013, 8) reported that 373 students had been excluded from Swedish universities in 2012 due to plagiarism.