Test Eşitlemede Grup Değişmezliğinin Farklı Simülasyon Koşulları Altında İncelenmesi

Abstract

The purpose of this study was to examine the impact of differential item functioning in anchor items on the group invariance in equating for different sample sizes. Within this scope, the factors that chosen to investigate the group invariance in equating were sample size, frequency of sample size of subgroups, differential form of differential item functioning, frequency of items in the anchor test with differential item functioning, directionality of differential item functioning and mean differences in subpopulation ability levels. The current study was conducted by using item response theory true score equating under equivalent groups anchor test design. REMSD, RESDk, RMSD(x) and RSDk(x) indexes were used for investigating group invariance in equating. This study was designed as a comparison of equating results on 96 simulation conditions. The R language and SPSS software was utilized for analysis and 100 replication were performed for each condition. The effect of the conditions held in the study on group invariance in equating were evaluated by taking average of REMSD, RESDk, RMSD(x) and RSDk(x). Also ANOVA was performed to determine significant effect of each of factor on group invariance in equating. In the scope of the results of the study, differential form DIF was the factor that had the most prominent impact on group invariance in equating. The findings of this study show group invariance was affected by factors of DIF were only in instances in which DIF in anchor items is differential across test forms. When DIF in anchor items differs across forms to be equated, direction of DIF has the largest effect on group invariance. When DIF in anchor items were been on each of forms to be equated, differences in group mean ability has the largest impact on group invariance.