Higher Education End-of-Course Evaluations: Assessing the Psychometric Properties Utilizing Exploratory Factor Analysis and Rasch Modeling Approaches

This paper offers a critical assessment of the psychometric properties of a standard higher education end-of-course evaluation. Using both exploratory factor analysis (EFA) and Rasch modeling, the authors investigate the (a) an overall assessment of dimensionality using EFA, (b) a secondary assessment of dimensionality using a principal components analysis (PCA) of the residuals when the items are fit to the Rasch model, and (c) an assessment of item-level properties using item-level statistics provided when the items are fit to the Rasch model. The results support the usage of the scale as a supplement to high-stakes decision making such as tenure. However, the lack of precise targeting of item difficulty to person ability combined with the low person separation index renders rank-ordering professors according to minuscule differences in overall subscale scores a highly questionable practice.