Validating an Instrument for Student Evaluation of Teachers: Some Noteworthy by-Products

Abstract

We present strong evidence for the external validity, and some evidence for the construct validity, of a published instrument for student evaluations of teaching performance. For external validation, we replicated the instruments factor structure very closely, yielding five factors virtually identical to those of a previous study that used two different populations. For construct validation, we developed a preliminary theoretical model that predicts different effects of three sets of variables on student ratings of teachers. The model was partially supported. Student expectations about teachers had significant effects on all dimensions of teaching effectiveness. Class size and teacher experience had much weaker, and mostly negative, effects on the ratings. Student ratings of teaching effectiveness were driven more strongly by a student characteristic than by a teaching condition or a teacher characteristic.