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 insert ignore into journalissuearticles values(EFA); and Rasch modeling, the authors investigate the insert ignore into journalissuearticles values(a); an overall assessment of dimensionality using EFA, insert ignore into journalissuearticles values(b); a secondary assessment of dimensionality using a principal components analysis insert ignore into journalissuearticles values(PCA); of the residuals when the items are fit to the Rasch model, and insert ignore into journalissuearticles values(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.