Instructor Attire and Voluntary Student Participation in Higher Education: Evidence from a Randomized Class-Section Experiment

Students' engagement with a course is often measured through grades and attendance, yet much of what happens in a classroom is voluntary. This article studies one visible feature of teaching practice--instructor attire--and asks whether it changes that voluntary margin. The evidence comes from six first-year ``Information (Data Science)'' sections at Chiba University of Commerce. The same instructor taught all sections, using a common syllabus, schedule, and grading scheme; three sections were randomly assigned to formal attire and three to casual attire. The course is not an economics course, but the design speaks to a question familiar in economics education: whether observable teaching practices affect forms of student engagement that standard outcome measures miss. Formal attire is not associated with higher attendance or final exam scores. The participation margin looks different. Students in formal-attire sections submitted fewer ungraded reaction papers and wrote fewer words per submitted reaction paper. Exact randomization inference and wild-cluster bootstrap checks point in the same direction, while also showing how much uncertainty remains with only six randomized sections. A supplementary sentiment measure, coded from the original Japanese reactions, is too imprecise to support a separate conclusion. The main lesson is therefore about measurement as much as attire: voluntary-effort outcomes can reveal classroom responses that are not visible in grades or attendance alone.