A Rasch-Based Borderline Method for Standard Setting in an Objective Structured Clinical Examination Station

• In Event: 35.081 - Roundtable Session 13
  In Roundtable Session: 35.081-18 - Multidisciplinary Applications Examining Psychometric Properties With the Rasch Model

Abstract

The borderline regression method (BRM) is currently one of the most popular methods of setting a cut score in an objective structured clinical examination (OSCE) station. It assumes that the checklist items and the global rating item assess a common competency. However, these assumptions are seldom verified and the psychometric properties of the global rating scale are unknown. We tested a Rasch-based borderline method (Rasch-BM) of setting a cut score in an OSCE station that avoids these disadvantages. The study objectives were: (1) empirically test the proposed Rasch-BM with three OSCE stations and (2) compare the Rasch-BM cut-scores and failure rates to those of the BRM. Results show that the method has some advantages, but further testing is needed.

Authors

• Jean-Sebastien Renaud, Université Laval

• Gilles Raiche, L'Université du Québec à Montréal

• Eric Dionne, University of Ottawa

• François Ratté, Université Laval

• Julie F. Thériault, Université Laval