DFIT provides a set of functions to calculate the noncompensatory (NCDIF), compensatory (CDIF) and test level (DTF)
differential functioning indices for items and tests under Raju's (Raju, et al. 1995) DFIT framework.
It also provides functions for obtaining cut-off points for identifying differential functioning for these indices
following the Monte Carlo Item Parameter Replication approach proposed by Oshima et al. (2006).

This package also improves upon available DFIT software by allowing the covariance matrices for both focal and reference
groups to be used. This improves the obtained cut-off points, which result in type I error rates at the nominal level,
and increased power, when compared to the cut-off points obtained when using only the focal group item parameter
estimates and their estimate covariances (Cervantes, 2012). Furthermore, this package includes functions for obtaining
the asymptotic covariance matrices of item parameter estimates (currently only for dichotomous IRT models) and for
calculating the DFIT indices base on the focal group distribution as well as ability estimates for a sample from the
focal population are included; these enable ad hoc and a priori power calculations for given item parameters and sample
sizes to be possible with this package.
