Selection Bias and Systematic Sorting in Higher Education? Sampling- and Measurement Error in Students‘ Evaluation of Teaching
PhD Candidate: Edgar Treischl
Description: Student evaluation of teaching (SET) is the standard tool to measure teaching effectiveness from students‘ perspective. SET and other forms of empirical research are often faced with problems stemming from selection or systematic sorting of participants. In order to provide unbiased estimators of teaching quality SET needs to account for methodological concerns due to systematic sorting. In his dissertation Edgar Treischl demonstrates empirically that systematic sorting can lead to a sampling- and measurement error in SET. The first case implies that systematic sorting can lead to a selective sample of students enrolled in class and, ultimately participating in SET. In the last part of the dissertation the author asks about the consequences of systematic sorting in terms of validity and emphasizes that SET needs to account for student selection effects in order to provide unbiased indicators of teaching effectiveness.