University of Minnesota
University Relations

Minnesota Supercomputing Institute

Log out of MyMSI

Research Abstracts Online
January 2010 - March 2011

Main TOC

University of Minnesota Twin Cities
College of Education and Human Development
Department of Educational Psychology

PI: Ernest C. Davenport, Jr.

The Effects of Ability Level of the Examinees in Answer Copy Pairs on the Power and Type I Error Rate of GBT Index to Detect Answer Copying on Multiple-Choice Tests

The GBT index developed by Van der Linden & Sotaridona (2006) is the latest statistical method suggested to detect answer copying on multiple-choice tests and grounds on a strong statistical basis. However, no one has yet done research that has explored the empirical type I error rate and power of the GBT index. Three variables are hypothesized to have an effect on the statistical properties of the GBT index: the ability level of the source examinee (12 levels), the ability level of the cheater examinee (12 levels), and the amount of copying (10 levels). These researchers are exploring the effects of these variables in a simulation environment by manipulating these variables and calculating the empirical power and type I error rate of the GBT index under different conditions. The results of the study will reveal information about the statistical properties of the GBT index under different conditions for the people who may want to use this method in practice to detect answer copying or to provide additional evidence of answer copying.

Group Member

Cengiz Zopluoglu, Graduate Student