Abstract:
Multiple choice exams have the advantage of being easily graded by computers so that students may obtain test results quickly. Many instructors with large classes rely heavily upon multiple choice exams to reduce the hours normally spent grading examinations. Multiple choice exams, if not administered properly, have the disadvantage of providing an easy a venue for students to cheat. Cheating incidents have reportedly increased from 23% of all students in 1941 to over 75 % of all students in 1980. The instructor is faced with the problem of detecting cheating, and once detected, proving its existence. Two statistical methods that calculate the probability of cheating on multiple choice exams are evaluated. Both methods make assumptions which weaken their use under actual classroom testing situations. Based on these weaknesses the authors concluded that no satisfactory method exists for proving cheating through statistical analyses. The recommended approach for instructors is to design exams and classroom settings that discourage cheating.
Keywords:
cheating, multiple choice exams
 Download Article | [ ] | 895 kB |
