NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED517450
Record Type: Non-Journal
Publication Date: 2011-Apr-8
Pages: 25
Abstractor: As Provided
Reference Count: 23
ISBN: N/A
ISSN: N/A
The Effects of Answer Copying on the Ability Level Estimates of Cheater Examinees in Answer Copying Pairs
Zopluoglu, Cengiz; Davenport, Ernest C., Jr.
Online Submission, Paper presented at the Annual Meeting of the American Educational Research Association (New Orleans, LA, Apr 8-12, 2011)
The purpose of this study was to examine the effects of answer copying on the ability level estimates of cheater examinees in answer copying pairs. The study generated answer copying pairs for each of 1440 conditions, source ability (12) x cheater ability (12) x amount of copying (10). The average difference between the ability level estimates before answer copying and after answer copying was examined. The results indicate that there is no gain on average from answer copying for the cheater examinee when both cheater and source examinees are in the same ability level interval. The gain from answer copying is higher as the source examinee's ability level increases. The results also indicate that the gain and loss from answer copying is not the same when the source and cheater examinee exchange their position in the answer copying pair. The results indicate that answer copying on multiple-choice tests may have an important effect on the observed ability level estimate of a cheater examinee even for the small amount of copying. Answer copying behavior causes invalid test scores when the difference between the ability levels of the source and cheater examinees get larger. Answer copying can invalidate the test scores and high stake decisions based on the standard test scores. The educators should be aware of this problem and they should look for solutions to prevent or decrease the answer copying behavior. One solution might be using analytical methods to detect answer copying pairs by calculating the probability of matching responses between response vectors. The literature has a variety of analytical methods such as the [omega] index (Wollack, 1996) and the GBT index (Van der Linden & Sotaridona, 2006) to detect answer copying pairs. However, the statistical power of these methods is still questionable to detect a low ability examinee who gains an important amount of benefit by copying answers from a high ability examinee (Zopluoglu & Davenport, 2010). So, the better way is to prevent answer copying before it occurs. (Contains 10 tables and 1 figure.)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: Elementary Secondary Education
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A