Bayesian Estimation of Generalized Item Response Models with Flexible Tail Behavior

This dataset provides a comprehensive computational framework for the Bayesian inference of Generalized Item Response Theory (IRT) models. It features highly optimized algorithms for both the Generalized 2-Parameter Logistic Model (GL2PLM) and the Generalized Graded Response Model (GLGRM). The codebase is structured to address two primary estimation paradigms: (1) Known Shape Parameter: providing benchmark recovery analysis when $p$ is a pre-specified constant; and (2) Unknown Shape Parameter: implementing advanced MCMC strategies, including PG-augmented Gibbs sampling integrated with profile-likelihood grid search and slice sampling techniques. This repository serves as the official reproduction suite for the study titled "Bayesian Estimation of the Generalized 2PL and GRM Models".Institutional Affiliation: Minnan Normal University, School of Mathematics and Statistics, Fujian, China.