lamle

CRANlamleMaRDI QIDQ5982715

Maximum Likelihood Estimation of Latent Variable Models

Björn Andersson, Shaobo Jin

Last update: 25 August 2023

Software version identifier: 0.3.1

Copyright license: GNU General Public License, version 3.0, GNU General Public License, version 2.0

Description

Approximate marginal maximum likelihood estimation of multidimensionallatent variable models via adaptive quadrature or Laplace approximations to the integrals in the likelihood function, as presented for confirmatory factor analysis models in Jin, S., Noh, M., and Lee, Y. (2018) <doi:10.1080/10705511.2017.1403287>, for item response theory models in Andersson, B., and Xin, T. (2021) <doi:10.3102/1076998620945199>, and for generalized linear latent variable models in Andersson, B., Jin, S., and Zhang, M. (2023) <doi:10.1016/j.csda.2023.107710>. Models implemented includethe generalizedpartial credit model, the graded response model, and generalized linear latent variable models for Poisson, negative-binomial and normal distributions. Supports a combination of binary, ordinal, count and continuous observed variables and multiplegroup models.

Cites work

• H-Likelihood Approach to Factor Analysis for Ordinal Data
• Estimation of Latent Regression Item Response Theory Models Using a Second-Order Laplace Approximation
• Fast estimation of multiple group generalized linear latent variable models for categorical observed variables