Contributions: Abstract

Using the R-FIT computer program to assess item-fit in the Rasch model

Item-fit is of major concern when applying item response models. Several indices if item-fit have been proposed for the Rasch item response model, such as a) the cube root transformation of the mean square statistic (Smith, 1982; Wright & Stone, 1977), b) covariance based item fit indexes such as ECI2 and ECI4 (Tatsuoka, 1984), and c) likelihood function based Lz (Levine & Rubin, 1979).

The purposes of this paper are: a) to show a GAUSS computer program to calculate item-fit indexes in Rasch model, and b) to examine the type I error rates and the power obtained under different tests