Effects of Differing Item Parameters on Closed-Interval DIF Statistics

Abstract

The closed-interval signed area (CSA) and closed- moderate to large differences between the groups than interval unsigned area (CUA) statistics were studied by the CUA. However, the effect of the pseudo-guessing monte carlo simulation to detect differential item func- parameter varied depending on item discriminations. tioning when the reference and focal groups had differ- Index terms: closed-interval measures, differential item ent parameter distributions. When the-pseudo-guessing functioning, item response theory, monte carlo simula- parameter was varied, the CSA was better able to detect tion, signed area measures, unsigned area measures.