Investigating of the accuracy of estimating a Person’s ability by using methods in the program Bilog_Mg according to One- Parameter Model Rasch Model

Abstract

This study set out to examine how accurately ability of persons can be estimated using three commonly applied statistical methods: Maximum Likelihood (ML), Expected a Posteriori (EAP), and Maximum A Posteriori (MAP). These methods, implemented through the Bilog_MG software, were assessed within the framework of the Rasch model, also known as the One-Parameter Model. To carry out the analysis, simulated data were generated to produce binary responses (0 or 1) for a 20-item test, based on three sample sizes—200, 500, and 1000—using the WinGen program. Each estimation method was then applied to these datasets to estimate ability of persons and evaluate the precision of these estimates through standard errors. The results showed that the MAP method provided the most precise ability estimates across all sample sizes, followed by EAP and then ML