The Impact of the Spacing Effect and Overlearning on Student Performance in Calculus

Date of Completion

January 2011


Education, Mathematics




In most mathematics textbooks, each set of homework problems is comprised entirely of problems consisting of material found in the previous lesson. Homework problems can be assigned in various ways, such as a massed or a cumulative distribution of problems. A massed distribution of homework problems is the traditional method of assigning problems which are blocked by type within a section. A cumulative distribution assigns problems from within a section and from previous sections. Research has shown that the cumulative assignment of mathematics homework problems can improve subsequent student test performance (Mayfield & Chase, 2002; Rohrer, 2009; Rohrer & Taylor, 2007). ^ Additionally, most mathematics textbooks rely on a format that emphasizes a common learning strategy called overlearning. By an overlearning strategy, a student masters a skill then continues to practice this skill. This learning strategy is common in mathematics since homework assignments tend to require students to solve many problems of the same type. ^ A quantitative research design was employed in a large lecture Calculus I course at the university level to investigate the effects of massed versus cumulative practice and the overlearning strategy on student performance. Data collection took place during an entire semester and included student scores on homework and exams. ^ Using an ANCOVA research design, the analysis sought to determine whether statistical differences existed on final exam scores when homework was assigned in a massed versus a cumulative format and also sought to determine whether differences existed on final exam scores when homework was assigned in a small homework (n=6) format versus a large homework (n=9) format. The findings showed that there were no statistical differences on final exam scores between the participants in the massed and cumulative groups. Additionally, the 50% increase in the number of homework problems per set did not show statistical differences in student performance as measured on final exam scores. ^