COVID‐19‐Related Changes Impair Student Performance in a Subset of Students

Abstract

COVID‐19 forced radical changes to the delivery of college curricula world‐wide during the spring of 2020 with little to no warning. Commonly, classes that had been delivered face to face were moved to a virtual delivery format; a practice that continued during the fall 2020 semester. The purpose of this investigation was to perform an interim analysis of COVID‐related changes on student performance. We hypothesized that despite adequate preparation time for conversion to online delivery and a student body comprised largely of digital natives, the collective challenges associated with COVID‐19 would negatively impact student performance. To address this hypothesis we compared midterm data from fall 2020 to a comparison population comprised of midterm data pooled from the 2015‐2019 fall semesters from a large, public, Midwest university. Midterms are assigned as letter grades to students receiving a C‐ or below in a class. Letter grades were tallied for each term and then converted to a numerical score using a five‐point scale (C‐ = 1, F = 5) and summed across all midterm grades within a term to allow quantitation and statistical comparison. We discovered that the percent of students with a midterm (<C in a class) was ~15% lower in 2020 compared to the comparison group (p<0.0001) and that this rate has been declining since 2015 (p<0.0001). However, of the students that received midterms, there was a slight but significant increase in the number of midterms/student (p<0.01) as well as in the variation of the midterms/student (p<0.001) in 2020 compared to the comparison group. Moreover, students had lower grades reported (0.39 summed points; p<0.0001) and a greater amount of variance among those reported grades in the 2020 midterm group compared to the comparison group (p<0.0001). Finally, within the 2020 dataset, STEM majors had 10% more midterms/student (p<0.05) and these midterms were more severe (0.52 summed points; p<0.05) compared to non‐STEM majors. These data indicate that, surprisingly, and despite COVID‐19 related challenges, the percent of total students with midterms was decreased compared to previous semesters. Importantly though, despite significant efforts of faculty and universities, there is a population of students with exaggerated performance deficits compared to previous semesters. Given our finding of a smaller population of students with midterms, the exaggerated performance deficits in students with midterms may be due to factors external to instructional quality and point to social‐, economic‐, health‐, and internet access‐related barriers, which themselves maybe independent or interdependent, as well as the suitability and/or preparedness of the student for online instruction. Identification and remediation of these factors is essential given the likelihood of continued and expanded virtual learning, particularly for STEM students.