In what might be classed as a stunningly unsurprising result Michael Hessler and his colleagues did an experiment to demonstrate that bribing students with cookies helps with evaluations of teaching. Broadly speaking I have no doubt that they are right. I have seen a few criticisms of the research method of this paper — would other more ‘respectable’ interventions do the same thing? — but I’d be surprised if the minor critiques matter that much. The authors have shown that human beings, here students, aren’t objective evaluators which has the advantage of agreeing with the findings of nearly every study ever, as well as what a lay person might think from their general experience of life.
I would guess it occurs to most teachers at some stage that their teaching ratings will be improved if they do certain things that have nothing to do with teaching quality. Teachers will react differently. At one extreme some (probably tenured faculty) will refuse steadfastly to wear clean clothes complaining that students shouldn’t be so superficial. At the other extreme teachers will frantically ingratiate themselves with treats, easy grades, cheap flattery, and content-less courses that everyone can exceed at without working. While most teachers are probably in the mid-range I would still think that we need some sort of school policy to stop things being a race to the bottom as professors compete to flatter and bribe their students.
In this research where I was a bit more disappointed was that the bigger questions weren’t really answered. How long is the “half-life” of a cookie treat? If I give cookies one week will the students still have their judgment influenced the next week? The month after? These sort of questions matter to use of student evaluations in the real world — at our school evaluations are taken at the beginning of class to limit some forms of gaming.
The authors conclude that: “These findings question the validity of SETs [student evaluations of teaching] and their use in making widespread decisions within a faculty” (Hessler et al, 2081, page 1064). Especially, when as they argue against, the tools are used “blindly”. This makes sense — people don’t respond in an unbiased manner to surveys, and following data blindly is often a problem. Yet showing that people aren’t the impartial arbiters of social science folklore is painfully easy. Given this we need a bit more detail on better methods. I, for one, worry about just ignoring student input both because a) students have a valid opinion but b) also because the alternatives all have significant problems too. Review of teaching by senior colleagues has a whole host of obvious problems to do with abuse of power. (And as a practical matter for junior faculty senior colleagues are probably more expensive to bride than undergraduates). We certainly need to do better in performance evaluation — in teaching and other fields. That is a question worthy of much more research.
Read: Michael Hessler, Daniel M Pöpping, Hanna Hollstein, Hendrik Ohlenburg, Philip H Arnemann, Christina Massoth, Laura M Seidel, Alexander Zarbock, and Manuel Wenk (2018) Availability of cookies during an academic course session affects evaluation of teaching, Medical Education, 50 (10) pages 1064-1072