Proctored versus Unproctored Math Placement Tests: Does It Matter?

Abstract

Many institutions use placement tests as a method to assess students’ readiness for college-level coursework. With the increased use of technology in testing, many institutions have transitioned placement test administration to an online format in an unproctored setting. While unproctored placement tests may provide financial and logistical benefits for institutions and students, it is important to examine if there are differences in academic outcomes when tests are administered in this format. Guided by the literature on test administration modality (i.e., proctored versus unproctored examinations), I examined if there are differences in math course performance and college student enrollment persistence between students who completed a proctored or unproctored math placement test. To investigate these important educational outcomes, I analyzed data collected as part of a 2018 randomized control trial conducted at a large, urban public institution in the Mid-Atlantic region of the United States in which incoming, first-year students were randomly assigned to a proctored or unproctored group to complete a math placement assessment. The current study affirmed findings from a pilot study, which suggests that students tend to place approximately one course level lower when placed using a proctored math placement test compared to an unproctored placement test and that students tend to have higher final grades in their initial math course when placed by a proctored math placement test.The current study analyzed final grades in the second math course taken between proctored and unproctored groups as well as student persistence. Analyzing mean course grades, I found that differences in grades between students who take a proctored and unproctored math placement test continue in some cases into the second math course. The percentage of F’s and withdrawals for initial and second math course final grades between students who take a proctored and unproctored math placement assessment also show differences between groups. Applying hierarchical linear regression, suggests that test administration modality does not account for a significant amount of variance in course grades. When controlling for demographic characteristics and academic factors, performance from the initial course taken, was found to be the most significant factor of grades in second math course taken. Results from the current study suggests that test administration modality during math placement tests while not a statistically significant variable in academic performance in second math course taken, may still be helpful as it is a statistically significant variable in academic performance in the first math course taken. Since initial math course grades were statistically significant in explaining the difference in grades between groups, institutions should consider using proctoring during math placement tests as a practice. Further research should be conducted, however, to understand how test administration modality during placement tests affects students in different programs, including programs that do not require courses along the institution’s math course sequence. Additionally, further research on types of proctoring would provide a more comprehensive understanding of the proctoring options available to institutions and whether they result in the same outcomes for students.