The Drunkard's Search: Student Evaluation in Assessing Teaching Effectiveness
As social scientists, we understand the problem of the Drunkard's Search -- the lure and perils of using easy-to-obtain but irrelevant data -- yet we are employed by institutions that are clearly searching under the lamppost for data to use in employment decisions. Researchers from various disciplines have studied and lamented the biases inherent in student course evaluations. Studies have found that these evaluations show systematic bias against women and people of color. They also may mask poor teaching practices as they are better measures of popularity than teaching effectiveness. Each year another study is released, leading to momentary hand wringing about the weakness of course evaluations as a means of assessing faculty and warning against their use in tenure and promotion decisions. Rather than adopting alternative means of evaluating faculty teaching and thinking creatively about student feedback, administrators and faculty leaders default to these surveys, claiming there is no other way to collect data on teaching. These course surveys continue to be used despite the mounting evidence that they provide not only no evidence of teaching effectiveness, but bad evidence that does damage to faculty both directly and indirectly. In an effort to raise the profile of these discussions and push for tangible change, we offer a comprehensive literature review of the existing research on student course evaluations and their biases.
Siver, Christi and Haeg, G. Claire, "The Drunkard's Search: Student Evaluation in Assessing Teaching Effectiveness" (2018). Forum Lectures. 373.