Procedure of Standard Setting Using Borderline Regression Methods in Objective Structured Clinical Examination of Undergraduate Medical Education

Standard setting with borderline regression method (BRM) is a practical and defensible process to determine the cut-off point in Objective Structured Clinical Examination (OSCE). For standard setting method in OSCE, two different scores, the checklist scores and the global rating scores were recorded. These scores were used to apply standard setting using BRM to calculate the pass/fail cut-point in five stations OSCE in an undergraduate programme. The traditionally set pass marks at 50 % wasreestablished to be 54.98 % using BRM. To assess the quality of OSCE, the auto-generated correlation of determination (R2) ranging from 19.9 % to 56.6 % and slope of regression line B coefficient ranging from 4.00 to 10.32 points among the 5-stations OSCE was also considered. Root Mean Square Error (RMSE) was determined as an absolute indicator of reliability in a composite scaled score of multi-dimensional OSCE stations. BRM was found a less time consuming and practical method of standard setting for OSCE stations. It also has the advantage of providing data for quality assurance. RMSE is more appropriate to determine the fitness of model than Cronbachs alpha. R2and the intergrade discrimination provides the other relative indices of reliability in multi stations OSCE of students clinical performance.