Is test security a concern when OSCE stations are repeated across clerkship rotations?

PURPOSE: To examine the issue of test security when the same stations on an objective structured clinical examination (OSCE) are repeated across clerkship rotations. Specifically, is there a significant difference in students' scores on stations repeated in three or four rotations within a single academic year? METHOD: The sample consisted of 15 stations in the OSCE given at the end of the third-year surgery clerkship at the Southern Illinois University School of Medicine from 1989-90 through 1993-94. Each station was administered three or four times a year. One-way analyses of variance with contrast coding to test for linear trends were used. Results were considered significant at or below the .05 level. RESULTS: Only three of the 15 stations showed significant linear trends. A two-part couplet orthopedic station showed a significant decreasing linear trend (p=.0001). Two stations showed significant increasing linear trends: a general surgery couplet station (p=.0004) and a plastic surgery station with an essay question (p=.0253). CONCLUSION: There was no consistent evidence that students scored increasingly higher on OSCE stations repeated throughout the year. Thus, it would appear that a clerkship can repeat OSCE stations within an academic year without risk of a trend toward increasing scores.