Description
Th.J.B. Kropmans; D. Cunningham; K.Kennedy; J.A. Patterson (2014) Introduction Skills examinations use a cut-off score of 50% to separate good and bad performing students. However, a fixed cut-off score is only associated with students ability to pass exams. The difficulty of the exam or variability between examiners in marking student�s performance is not being taken into account. The Online Marking Tool developed within our School of Medicine incorporates various Global Rating Scales (GRS) marking professional competence as Pass, Borderline, Fail, Good and Excellent. Material Borderline Regression Analysis is used to determine a flexible cut-off score. The Online Marking Tool contains of an OSCE Management Tool to plan and execute Objective Structured Clinical Examinations (OSCE), a OSCE station bank and a Result Analysis tool including a fully fledged Borderline Regression Analysis pack (Excel export pack). Individual (blinded) scores at item level are correlated with the GRS of Fail, Borderline, Pass, Good and Excellent or variations of this GRS (Borderline Fail and Borderline Pass; Fail, Borderline, Pass). Results Borderline Group Average, the average mean score of those students being marked as �Borderline� is compared with Borderline Regression Method 1 and 2 (Borderline Fail e.g Pass) to determine a flexible cut-off score. Twelve prestigious universities are currently using this unique online marking tool for clinical skills assessments. Over 200 OSCE were successfully administered and analysed using this software solution. Conclusion Costs of the labour intense paper trail of regular OSCE is reduced by 70% and the error reduction in the paper trail is reduced by 30%. Approximately 19% more students fail after introducing Borderline Regression Analysis, together with a minimum amount of stations that need to be passed [Chaired Session] Irish Network of Medical Educators University College Dublin, 13 02 2014-14 02 2014.| Period | 13 Feb 2014 → 14 Feb 2014 |
|---|---|
| Event type | Conference |