Evaluating Stan’s Variational Bayes Algorithm for Estimating Multidimensional IRT Models

Bayesian estimation of multidimensional item response theory (IRT) models in large data sets may come with impractical computational burdens when general-purpose Markov chain Monte Carlo (MCMC) samplers are employed. Variational Bayes (VB)—a method for approximating the posterior distribution—poses a potential remedy. Stan’s VB algorithms have drastically improved accessibility methods wide psychometric audience. Using marginal maximum likelihood (MML) and MCMC as benchmarks, present simulation study investigates utility built-in function estimating IRT between-item dimensionality. yielded marked speed-up comparison to MCMC, but did not generally outperform MML terms run time. estimates were trustworthy only difficulties, while bias discriminations depended on model’s Under realistic conditions non-zero correlations between dimensions, correlation subject severe bias. The practical relevance performance differences is illustrated from PISA 2018. We conclude that its current form, algorithm does pose viable alternative models.