R Package mirt
Rudolf Debelak, Carolin Strobl, Matthew D. Zeigenfuse · 2022
The mirt package contains functionality for fitting the parameters of a variety of Item Response Theory (IRT) models - including the 2PL and 3PL models - using marginal maximum likelihood (MML) for the item parameters, as well as maximum likelihood (ML), weighted maximum likelihood (WLE) or Bayesian approaches (MAP and EAP) for the person parameters. This chapter shows how to perform item and person parameter estimation for the Rasch, 2PL and 3PL model, conduct model comparisons, and assess the goodness of fit with the mirt package. Like in the previous chapter, all R commands as well as the output are thoroughly explained and references to the theory part of the book are highlighted.