Linking teachers’ facial microexpressions with student-based evaluation of teaching effectiveness: A pilot study using FaceReader™

This study seeks to investigate the potential influence of facial microexpressions on student-based evaluations and to explore the future possibilities of using automated technologies in higher education. We applied a non-experimental correlational design to investigate if the number of videotaped university lecturers’ facial microexpressions recognized by FaceReader™ serves as a predictor for positive results on student evaluation of teaching effectiveness. Therefore, we analyzed five videotaped lectures with the automatic facial recognition software. Additionally, each video was rated by between 8 and 16 students, using a rating instrument based on the results of Murray´s (1983) factor analysis. The FaceReader™ software could detect more than 5.000 facial microexpressions. Although positive emotions bear positive influence on the “overall performance rating”, “emotions” is not predicting “overall performance rating”, b = .05, t(37) = .35, p > .05. The study demonstrates that student ratings are affected by more variables than just facial microexpressions. The study showed that sympathy as well as the estimated age of the lecturer predicted higher student ratings.