Student Perceptions of Fairness and Security in Versioned Programming Exams

The Bottom Line

Understanding the most effective and cost-efficient way to ensure fairness between exams will help test facilitators. If students’ perceptions of fairness impact individual test scores or overall class performance, then facilitators should focus on balancing fairness with test security.

Research Summary

As more exams move online, test facilitators are often asked to trade off on exam security (i.e. the prevention of cheating) for test fairness. Test fairness is measured by the similarity in difficulties across versioned exams. This paper examines 454 students’ perceptions of fairness of exams in a computer science course, as measured by:

1. exam security;
2. exam difficulty variance; and
3. pre-exam utility.

The study found that perceptions of fairness were not correlated with exam performance, but were significantly correlated with overall performance in the course. There are two potential explanations for this: 1) That students tend to view themselves as competent, thus may perceive an exam to be unfair if they performed poorly and, 2) low performing students may fixate on unfairness because they are more susceptible to it.‍

Key Takeaways

1. Students Perceptions Correlate to Their Overall Course Performance

Students’ perceptions of the fairness of variation between exams is correlated with their overall course performance, yet is not correlated with their most recent individual test score. Students differ significantly in their perceptions of fairness, with some thinking that cheating prevention is paramount while others believe reducing variation in difficulty between tests is more important.

2. Exam Versioning Is an Effective Tool to Prevent Cheating Online

Exam versioning is an effective tool to prevent cheating in online tests, as such, more research should investigate the continuum on which students perceive fairness.