Cumulative Achievement Testing IN MEDICAL SCIENCE EDUCATION

David Swanson¹, Kathy Holtzman¹, Aggie Butler¹ and the Case Western Reserve University School of Medicine Cumulative Achievement Testing Study Group

¹National Board of Medical Examiners (NBME), Philadelphia, PA 19104 U.S.A.

PURPOSE: This project explores use of cumulative achievement tests in basic science education. These tests encourage students to retain previously covered material and attend to relationships among topics by systematically retesting previously covered topics on end-of-unit tests.

METHODS: The systems-based curriculum at Case Western Reserve University (CWRU) is divided into six units. Using web-based test construction software developed by the NBME, CWRU faculty members and NBME staff developed six end-of-unit subtests customized to unit content from NBME item pools. Six test forms were assembled from these subtests by combining items across units. Each form consisted of 100 items on the unit just covered (primary focus) and blocks of 20 items from units covered earlier (secondary focus). To build secondary subtests, items from each primary subtest were assigned to five 20-item blocks, students were randomly assigned to five groups, and, across administrations, student groups were rotated across secondary subtests such that 1) all 100 items from each primary subtest were seen by 20% of students and 2) no student saw the same 20-item block more than once.

RESULTS: Following each (web-based) test administration, students received score reports depicting areas of strength and weakness for the unit just covered and for each 20-item block drawn from earlier units. Because all items were used for each test administration, items served as their own controls, and performance trends on individual items and topics were studied by analyzing shifts in mean scores.

CONCLUSION: Cumulative achievement tests appear promising for curriculum evaluation and study of student retention.