Items Fit to the Three Parameter Logistic Model and Their Effect on the Factorial Structural of Scale Using Confirmatory Factor Analysis: Simulation Study

Abstract

The current study aimed to identify the effect of fitting the items to the three-parameter model in item response theory on the standard regression weights and residuals as indicators of the quality of the factor structure of the scale using confirmatory factor analysis in light of the sample size and number of items. The study sample consisted of different sizes that were generated using simulation. following the comparative descriptive approach and using appropriate statistical methods through R. SPSS. AMOS. The study reached many results. the most important of it: there are statistically significant differences in the standard regression weights due to the difference in the percentage of items that do not fit the three- parameter model. and the differences are in favor of models with the lowest percentage of non- fit items. and the highest in The number of items and the smaller sample size. and for the residuals. the differences were in favor of the model with the lower percentage of non- fit items. and the higher number of items and smaller sample size. There are also significant effects of the joint interactions between the study variables on the standard regression weights and the residuals. and according to the results. a set of recommendations and proposals were presented.