Is exam performance influenced by whether the proctor format is in-person or webcam?

The demand for webcam proctoring services is fueled by the large and growing share of online courses, and the need for integrity of assessments. The webcam proctor format is consistent with the attributes of online courses being free of time and geographic constraints while in-person proctoring is not. Two critically relevant factors between the proctor format environments of in-person and webcam are opportunities for cheating, and anxiety induced by the proctoring format. In-person has the benefits of inhibiting cheating opportunities and verification of the identity of the test taker. On the other hand, the webcam format is potentially more secure because of the one-on-one monitoring but may increase anxiety by the off-putting big brother format of the web cam. This study investigates whether exam performance depends on the proctor format of the assessment and whether cheating and anxiety play a relevant role.

This study uses data from 3 principles of microeconomics courses. The students are undergraduates. The course is the full 16-week fall/spring term. The course is taught by the same instructor and in consecutive semesters using the same course materials. The experiment has the approval of the Institutional Review Board. For a high stakes multiple-choice exam, students are assigned to three proctoring formats: no proctor, in-person, webcam. Learning outcomes are predicted from a model with independent variables of student characteristics common to the literature. The authors expect the prediction model to have the same explanatory power for the proctored exam formats. If there is a significant difference in explanatory power of the two models, this can be attributed to unobserved factors such as cheating, or anxiety related to proctoring format. The empirical results are the exam scores are lowest for the webcam proctor, higher for the in-person proctor, and highest for the un-proctored exam format. The result of the webcam proctor to the in-person proctor is consistent with an interpretation that the webcam proctor is more effective in limiting cheating opportunities than the in-proctor format. It is also consistent with the interpretation that the one-on-one surveillance of the webcam format creates more test anxiety then the in-person proctor format. Post-exam survey responses confirm the webcam promotes exam anxiety more than the in-person format.