A Controlled Experiment on Two-Stage Exams in an Introductory Statistics Course

Work thumb

Views: 71


A two-stage collaborative group exam is an exam in which students first write the test individually, then rewrite the same or similar test in groups immediately after. To assess the effect of two-stage exams on student learning in an introductory statistics course, we used a randomized crossover design on three midterms of two parallel sections of an introductory statistics course. For each midterm, students wrote an individual full test (stage 1), immediately followed by an individual retest and a group retest (stage 2), and finally a learning check test in the form of a pop quiz four days later (stage 3). Student performance on the learning check test was then compared between group retest and individual retest conditions. This paper focuses on the results of Midterm 1 only. We found significant improvement in short-term content retention from writing two-stage exams, and significantly more content retention improvement from group retesting than individual retesting. As expected, we find performance improvement on group retests to be significant, with “lower” achieving students gaining the most. To assess student opinion on two-stage exams, a year after the initial experiment, we conducted an anonymous class survey on a different cohort enrolled in the same course. The survey shows that the majority of the students found writing two-stage exams less stressful than traditional exams, and the vast majority of students reported they learned more from writing two-stage exams and preferred two-stage exams to traditional exams.