Academic integrity in on-line exams: Evidence from a randomized field experiment

We study academic integrity in a final exam of compulsory course with almost 500 undergraduate students at major Spanish university. Confinement and university closure due to Covid-19 took place by the end last lecture week. As consequence, usual classroom was turned into an unproctored on-line multiple-choice without backtracking. exploit different orders problems detailed data timestamps students’ integrity. First, taking average over questions that were part both earlier later “rounds”, we find number correct answers round 7.7% higher than round. Second, completion time 18.1% shorter Third, mere reminder university’s code ethics, which sent subgroup halfway through exam, did not affect cheating levels.