Abstract
Practice exams are a type of deliberate practice that have been shown to improve student course performance. Deliberate practice differs from other types of practice, because it is targeted, mentally challenging, can be repeated, and requires feedback. Providing frequent instructor feedback to students, particularly in large classes, can be prohibitive. A possible solution is to have students grade practice exams using an instructor-generated rubric, receiving points only for completion. Students can either grade their own or a peer’s work. We investigated whether peer or self-grading had a differential impact on completion of practice exam assignments, performance on practice exams or course exams, or student grading accuracy. We also investigated whether student characteristics mattered. We found that 90% of students took all practice exams or only missed one and that there was no difference on practice or course exam performance between the peer and self-graders. However, in the peer-grading treatment, students with lower incoming grade point averages and students identified as economically or educationally disadvantaged were less accurate and more lenient graders than other students. As there is no clear benefit of peer grading over self-grading, we suggest that either format can solve the challenge instructors face in giving frequent personalized feedback to many students.
