The purpose of this research was to develop and test computerized methods of detecting cheating on multiple-choice examinations. Two computer programs which compare the response vector of every examinee in a class to that of every other examinee were utilized to calculate two indices, both of which are based on identical incorrect responses of two or more examinees. The Wrong-in-common (WIC) index is the proportion of two times the number of identical incorrects to the Total Wrong for Both (TWB). The Run index (RUN) is the proportion of two times the length of the longest 'run' two examinees share to TWB. A run is defined as the number of items answered incorrectly and identically by both examinees within a succession of items which are marked or unmarked identically. The computer programs are written in Microsoft BASIC and run on the IBM-PC and compatible microcomputers as stand-alone executable files Achievement test item responses were obtained from a large public school system. From these data, 10 'honest' groups were created, and distributions of the RUN and WIC indices were plotted for four intervals of TWB. Average .1% probability levels were computed from the distributions for each of two subtests. These analyses indicated that cutoff points, index levels above which a pair might be considered suspicious, must vary contingent on TWB. Results suggested that it may be possible to detect certain types of 'test irregularities', and that irregularities occurred in 5 to 8% of the 170 fifth-grade classrooms studied. Directions for future research are discussed / acase@tulane.edu
Identifer | oai:union.ndltd.org:TULANE/oai:http://digitallibrary.tulane.edu/:tulane_26129 |
Date | January 1986 |
Contributors | Hartman, Alison Gwen (Author) |
Publisher | Tulane University |
Source Sets | Tulane University |
Language | English |
Detected Language | English |
Rights | Access requires a license to the Dissertations and Theses (ProQuest) database., Copyright is in accordance with U.S. Copyright law |
Page generated in 0.0021 seconds