Reliability of Medical Student Clinical Assessments

• In Event: Educational Measurement Issues in the Health Care Professions

Abstract

Accurate assessments of medical students’ performance are important for student learning, improvement, and advancement decisions. Assessment in the clerkship space are subject to variability based on the student, the patient context, assessor factors, and clinical context. G theory was used to assess the reliability of Workplace Based Assessments (WBA) and to estimate the number of WBAs necessary to accurately estimate student competency on the group of Entrustable Professional Activities (EPA) being piloted at a large, Mid-Atlantic medical school. 9790 total assessments using the Modified Ottawa coactivity scale were analyzed with results indicating students’ explain between 1.9% and 9.1% of variance in assessment scores and that 24 or more assessments would be necessary to produce reliable results based on individual EPAs.

Authors

• JK Stringer, Rush Medical College

• Michael Ryan, Virginia Commonwealth University

• Robert Perera, Virginia Commonwealth University

• Yoon Soo Park, University of Illinois at Chicago

• Sally Santen, Virginia Commonwealth University