Abstract
Background: A large‐scale standardized patients‐based objective structured clinical examination (OSCE) was introduced in Québec in 1990 for licensing family physicians.
Purpose: This article presents the results of a construct validity study conducted on the Spring, 1993 OSCE, using the validation by the extremes method.
Methods: Three groups of participants (205 graduating residents, 21 graduating clerks, and 13 general practitioners) were submitted to the same exam. Differences in examination total scores, scores by cases, and scores by skills were analyzed using analysis of variance.
Results: The reliability of the examination was 0.80. Clerks obtained significantly lower total results (M = 58.8, SD = 5.6) than residents (M = 70.9, SD = 5.64, p < 0.001); no difference was found between residents’ and family physicians’ total scores (M = 71.3, SD = 4.34, p ‐ 0.662). Cases that were difficult for one group were as difficult for the two other groups.
Conclusions: The results support the construct validity of the Quebec licensing examination OSCE, and they confirm that a performance‐based examination for licensure should emphasize comprehension and problem‐solving throughout the assessment of integrative skills such as investigation, diagnosis, and treatment, and that trivialization of checklists should be avoided.