Differential Item Functioning Analysis Using a Mixture 3-Parameter Logistic Model With a Covariate on the TIMSS 2007 Mathematics Test

The purpose of this study was to explore what may be contributing to differences in performance in mathematics on the Trends in International Mathematics and Science Study 2007. This was done by using a mixture item response theory modeling approach to first detect latent classes in the data and then to examine differences in performance on items taken by examinees in the different latent classes. An exploratory mixture 3-Parameter Logistic model analysis detected two latent groups in the data. The model considered in this study used internet access as a covariate to illustrate the effect of the covariate on latent class membership.