Frontiers in Education (Apr 2022)
Extended Exam Time Has a Minimal Impact on Disparities in Student Outcomes in Introductory Physics
Abstract
Disparities in student outcomes, including gendered performance differences, are widespread in introductory physics and other STEM courses. STEM education researchers have investigated many course and student factors that could contribute to these inequities, including class size, test formats, assignment weightings, and students’ sense of belonging. These inequities are often largest in the timed, multiple-choice, high-stakes exams that characterize so many traditional introductory STEM courses. Time pressure is widely believed to influence student outcomes on these exams, reducing overall performance and perhaps exaggerating widespread group performance disparities. Reducing time pressure for students by providing more test-taking time is a small, structural change that could have large impacts on student performance and could differentially affect students. To explore this possibility, we offered all 596 students in our introductory physics course a 50% extension in test-taking time and collected data on exam performance, student demographics, and the time each student took to complete each exam. We made careful comparisons of student performance to historical data, across demographic groups, and across time usage on the exams using both raw exam scores and a “Better Than Expected” measure that compares student performance in the course under study to their own performance in other courses. While students overall scored slightly higher with extended exam time, we found that extended time did not reduce the well-established disparities in student outcomes categorized by sex, race/ethnicity, or college generation status present in our introductory physics course. These findings both indicate that extending exam time is not a simple fix for disparities in student outcomes and reinforce that systemic changes towards more authentic assessments of STEM knowledge and capabilities are imperative.
Keywords