Multiple forms of an exam are frequently used as a means of reducing the likelihood of cheating in large classes. However, questions have been raised regarding whether the order of test items influences student performance. Exams that are arranged in the order in which material was presented in class might improve student performance by providing the student with a logically ordered series of prompts. In contrast, an exam in which items are randomly ordered might be more confusing and, as such, might have adverse effects that result in a reduced level of performance. Viewed from the perspective of assessment, this type of effect is undesirable. Previous studies (Gohmann and Spector 1989; Monk and Stallings 1970) have concluded that the sequence of exam items does not have a statistically significant effect on the mean level of student performance. However, these studies do not address a number of other relevant questions, including possible differences in the distribution of scores on each form of the exam, effects on the validity of individual exam items, and effects on the reliability of the exam instrument. The purpose of this article is to address these perceived shortcomings. The results provide evidence that the distribution of exam scores may, in fact, be influenced by the order of exam items. However, item validity and instrument reliability appear to be unaffected.
