The Comparison of Parameters Efficiency of M-GRM Model between Posterior Predictive Model Method and Likelihood ratio Method with Monte Carlo Simulation Method

Abstract

This research aims to compare the performance of M-GRM models parameter with Monte Carlo simulation based on approximation method of Posterior predictive model (when b = -2.5,  -2, 0, 1, 2, 2.5: c = 0.1,0.2, 0.3: = 0.3, 1.0, 1.7, = -3, -2, -1, 0, 1, 2, 3 and = 50, 100, 200, 400 with 1,764 situations). For determination the unidimentional property of M-GRM using R Program to replicate 10,000 recursions with 1,764 situations, the sample size of 50, 100, 200, 400 was used. The result shows that the b parameter from posterior predictive model has more performance than likelihood method where as c parameter, form likelihood method has more performance in vice versus.