50% Fewer Students Needed For Course Evaluations

A multi-site, multi-country study shows that far fewer students are needed to evaluate a course, if they are asked to predict the opinion of their classmates instead of give their own. The predictive model of evaluation required input from 50% fewer students to produce stable ratings compared to the traditional personal opinion-based evaluations. “Gender, estimated general level of achievement or satisfaction after having completed the final examination” were not significant factors in the findings. The predictive model of evaluation used in this study borrows from the political science literature on predicting election results, and is consistent with prior findings that a random sample of the class using only 30-40 respondents may adequately represent the general opinion of the class. The potential to reduce evaluation fatigue is a highly attractive feature of this model.—Sarang Kim, MD

Schönrock-Adema J, et al. ‘What would my classmates say?’ An international study of the prediction-based method of course evaluation. Med Educ. 2013 May;47(5):453-62.

Link To Article

This entry was posted in Undergraduate Medical Education and tagged , . Bookmark the permalink.