Academic Dishonesty and Online Exams

• In Event: Accounting Education Best Practices

Abstract

The widespread and growing use of online courses coupled with concerns for academic dishonesty in the internet era, require an evaluation of how online introductory accounting courses are assessed. In this study I run an experiment that evaluates the effect on exam performance of different versions of online unproctored multiple choice exams that are modified to limit internet cheating. I find that mid-term exam results are significantly higher in unproctored versus proctored exams and I find that compared to using standard publisher-based test bank questions results are significantly lower when algorithmic, scrambled questions are used and lower again when re-worded questions are used, which suggests some impact on limiting online cheating. [Learning effects will be evaluated by comparing unproctored mid-terms with a proctored final exam]. Exam performance is positively correlated with GPA for face-to-face and online students, however age is positively correlated for face-to-face students, but negatively correlated for online students.

Authors

• David R Upton, UNC Greensboro

• Ayalew A Lulseged, University of North Carolina-Greensboro