Predicting mastery level on a large-scale standardized patient test: a comparison of case and instrument score-based models using discriminant function analysis.

摘要: Clinical skills assessments have traditionally been scored via experts' ratings of examinee performance. However, this approach to scoring may be impractical in a large-scale context due logistical and cost considerations as well the increased probability rater error. The purpose investigation was therefore identify, using discriminant analysis, weighted score-based models that maximize accuracy with which mastery level can estimated for examinees taking nationally administered standardized patient test. Additionally, resulting classification functions applied predict cross-validation sample also examined. Results suggest it might feasible implement an automated procedure cost-effective manner while still retaining important facets decision-making process expert raters. Cost-benefit, test development psychometric implications these results are discussed full paper.