Do streaks matter in multiple-choice tests?

Success in life is determined to a large extent by school performance, which in turn depends heavily on grades obtained in exams. In this study, we investigate a particular type of exam: multiple-choice tests. More concretely, we study if patterns of correct answers in multiple-choice tests affect performance. We design an experiment to study if students who take a multiple-choice test with streaks of identical answers earn fewer points than students taking a test with a variable pattern of answers, and we study several potential mechanisms. We do not find a clear treatment effect, as multiple-choice test points were not significantly lower in a consistent manner in the treatment groups. We observe that in a streak, the likelihood to choose a different and incorrect answer than the previous ones is higher in the treated group. The number of identical previous answers does not seem to explain these choices of incorrect answers.