Spaced Versus Massed Testing in a College Class: An Explanatory Item Response Model
Many studies have shown that distributed study is more effective than massed study. In the present study we were interested in the effects of frequent testing, student ability, practice quizzing, and item form (multiple choice or short-answer) at the pre-tests and the post-test. Two sections of an undergraduate psychological methods class were taught by the same professor at similar times of day. In the frequent testing class, 8 midterm exams were spaced about 1 exam every 2 weeks. In the standard class, 2 midterm exams were spaced about 1 every 8 weeks. All exams, including the final exam, consisted of both multiple choice (MC) and short answer (SA) questions. The form of questions (MC or SA) during the midterm exams was called the pre-test method and the form on the final was called the post-test method. Both classes took the same comprehensive final exam. Only final exam data was analyzed in this study. An explanatory item response model (EIRM) was used to estimate the effects of the person predictors: student ability, testing frequency, and quiz assignment, and the item predictors: pre-test and post-test method on final exam performance. Not surprisingly, student ability and post-test method explained the most variance in item responses of any of the predictors in the model. Testing frequency also significantly influenced item responses. A marginally significant interaction between testing frequency and post-test method was also observed. We conclude that frequent testing (spacing) improves performance relative to massing, however, the largest benefits are expected to be seen on recall memory tasks. We also argue that the benefits of frequent testing generalize across populations and conditions.