Evaluating Fake Lecturers

Clearly, we need to repeat a recently-published set of experiments by inserting fake lecturers into the evaluation forms that we give to students.

Here is a quote from the paper's results blurb: "Without a portrait, 66% (183 of 277) of students evaluated the fictitious lecturer, but fewer students (49%, 140 of 285) did so with a portrait (chi-squared test, p < 0.0001)."

(Tip of the cap to the Improbable Research blog.)