The Behavior of the Information Function Related to the Percentages of Missing Response and Sample Size Depending on Item Response Theory

This study aims at verifying the behavior of the information function related to percentages missing response and sample size depending on item response theory. To achieve the aim of the study, generating (1000) and (2000) examinees response according to three parameter logistic model using WINGEN program on 50 dichotomous items. Using (SPSS) program was to obtain data include missing responses rates (15%) and (30%), after confirming Unidimensionalityresponses by using several indicators. The responses of the examinees were analyzed by BILOG– MG3 program to estimate the item parameters, information function and standard error for the entire file. The results showed that, the amount of estimates of information function varies increasingly according to the decreasing missing rate of percentages and increasing sample size. The results show that the standard error in estimating the information function varies decreasingly according to the decreasing missing rate of percentages and increasing sample.