Unsupervised online exams have minimal impact on cheating, according to a study conducted by researchers at Iowa State University. When the university shifted to remote learning due to the COVID-19 pandemic, psychology professor Jason Chan was concerned about the potential for rampant cheating during online exams. However, after analyzing test score data from nearly 2,000 students across 18 classes, Chan and PhD candidate Dahwi Ahn discovered that individual student scores remained consistent with their results from in-person exams. This outcome suggests that cheating was either not prevalent or ineffective at significantly boosting scores.
The researchers found similar results across various academic disciplines, class sizes, course levels, and test styles. Unsupervised online exams produced scores that were comparable to those of in-person, proctored exams, indicating that they can provide a valid and reliable assessment of student learning. These findings have led both researchers to continue administering online exams, even for in-person classes. Online exams offer more flexibility for students with part-time jobs or extracurricular commitments and provide additional options for teaching remote classes. Ahn even led her first online course over the summer.
The minimal impact of cheating on test scores can be attributed to several factors. Students who are more likely to cheat may already be underperforming in their classes and feel anxious about failing. They may have missed lectures, fallen behind in their studies, or struggle to ask for help. In addition, even with the option to search the internet during an unmonitored exam, students may struggle to find the correct answer if they don’t understand the content. Previous studies comparing open-book and closed-book exams have provided evidence to support this theory.
Academic integrity and a sense of fairness also play a role in deterring cheating. Many students who have worked hard and take pride in their grades are less likely to share their exam answers with peers they perceive as freeloaders. However, despite the positive findings, instructors should remain cautious of potential weaknesses in unsupervised online exams. For example, certain platforms allow students to view the correct answers immediately after selecting multiple-choice options, making it easier for them to share answers in group chats.
To counter cheating and promote academic integrity, instructors can:
While the study conducted by Chan and Ahn provides valuable insights into the validity of online exams for student evaluations, it does have some limitations. The researchers did not account for the potential impact of stress and other COVID-19-related factors on students, faculty, and teaching assistants. It is unclear whether instructors may have been more lenient with grading or provided longer windows of time to complete exams. Furthermore, the study did not investigate whether the difficulty level of the classes varied throughout the semester.
Despite these limitations, the study’s findings offer reassurance regarding the effectiveness of unsupervised online exams. The researchers suggest that further research is necessary to understand how the advent of AI writing tools, such as ChatGPT, will impact online exams and instructors’ evaluation methods. In the meantime, the study encourages instructors to embrace online exams as a valid and reliable assessment tool that provides flexibility for students and expands options for remote teaching.
In conclusion, the study reveals that unsupervised online exams do not significantly contribute to cheating. The findings demonstrate that such exams can accurately evaluate student learning and offer flexibility for both students and instructors. By recognizing potential weaknesses and using techniques to promote academic integrity, instructors can effectively administer online exams while maintaining the trust and fairness of the evaluation process.