Inflated scores on the online exams during the COVID-19 pandemic school lockdown
(1) Liberal Arts and Science Academy, Austin, Texas, (2) Department of Chemistry, St. Edward's University, Austin, Texas
Due to the COVID-19 pandemic, colleges and universities in the US had to provide students with online-only instruction and assessment in the middle of the Spring 2020 semester. Since this was neither adequately planned nor thoroughly prepared due to little time, its effects on student academic performance are unknown. This study tested a hypothesis that students’ test scores were significantly higher on online exams during the COVID-19 school lockdown than those of the in-person exams before the lockdown. We compared the scores of the same students between one in-person exam and two online exams in a fundamental engineering class in the Spring 2020 semester. We further compared the scores between the students who took three in-person exams in the 2019 semesters and those in the Spring 2020 semester. The same instructor taught the class in both cases, and all exams had similar difficulties. We found that students achieved significantly inflated test scores when taking online exams, with the increase being in the range of 20% to 50%. Therefore, we have proven the hypothesis. Score inflation also depended on the level of prevention of student collaboration on online exams, with a loosely-controlled online exam’s scores increasing by 50% compared to exams with strict measures to prevent cheating. In comparison, students showed a similar performance at three in- person exams in the 2019 semesters, with differences ranging 6% to 8%. The current study could help show how the COVID-19 pandemic influences long-term academic performance of institutions, instructors, and more importantly, students.