Emoticons as Performance Feedback for College Students: A Large-Classroom Field Experiment

We use a field experiment in a large college classroom to study the use of written representations of emotion (“emoticons”) as a form of feedback on exams. We find that treatment improved student attendance, quiz scores, homework scores, and test scores. We document heterogeneous impacts based on the type of emoticon received and gender. Improvements were driven by those who received disapproving emoticons but scored above the mean. Female students were responsible for the improved homework and test scores, while men drove improvements in attendance. Both genders improved similarly on quiz scores. We discuss two psychological theories supporting a treatment effect.