A Bayesian Hierarchical Model for Detecting Aberrant Growth at the Group Level

Authored by: William P. Skorupski , Joe Fitzpatrick , Karla Egan

Handbook of Quantitative Methods for Detecting Cheating on Tests

Print publication date:  October  2016
Online publication date:  October  2016

Print ISBN: 9781138821804
eBook ISBN: 9781315743097
Adobe ISBN: 9781317588108

10.4324/9781315743097.ch12

 Download Chapter

 

Abstract

Cheating has been a problem in testing for as long as there have been high stakes associated with the results. Cheating on statewide assessments may have serious implications for the psychometric integrity of item parameters and test scores as well as the validity of those test scores. Cheaters cheat in lots of different ways, which can make the phenomenon difficult to detect. Over the last few decades, a number of approaches for detecting cheating have been suggested, such as identifying unusual similarity among response patterns (e.g., Wollack, 1997, 2003) and analyzing person-fit data (e.g., Drasgow, Levine, & Williams, 1985; Levine & Rubin, 1979). These techniques operationalize cheating as something individual test takers do, either by copying answers or by using illicit materials to enhance their scores. The stakes in testing, however, are not always high for only the student. In some cases, teacher merit pay is tied to test results. Under adequate yearly progress (AYP) standards associated with No Child Left Behind (NCLB), schools and even districts may be shut down or taken over by the state based on test results. In K-12 statewide assessments, teachers and administrators may be motivated to cheat because test results are often used for teacher, school, and district accountability. This cheating may range from subtle (e.g., teaching to the test) to blatant (e.g., changing student answer documents). The threat of not making AYP provides a large incentive for educators to cheat. In many states, AYP is based, in part, on the percentage of students reaching proficiency. Other states have started using growth models based on individual student growth across years when estimating AYP.

 Cite
Search for more...
Back to top

Use of cookies on this website

We are using cookies to provide statistics that help us give you the best experience of our site. You can find out more in our Privacy Policy. By continuing to use the site you are agreeing to our use of cookies.