AUTHOR=Liu Tianci , Wang Chun , Xu Gongjun 

TITLE=Estimating three- and four-parameter MIRT models with importance-weighted sampling enhanced variational auto-encoder

JOURNAL=Frontiers in Psychology

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2022.935419

DOI=10.3389/fpsyg.2022.935419

ISSN=1664-1078

ABSTRACT=Multidimensional Item Response Theory (MIRT) is widely used in educational
and psychological assessment and evaluation. With the increasing size of modern
assessment data, one major challenge arising in MIRT applications is that many existing
estimation methods become computationally demanding and not scalable to such big
data. To address this issue, this work proposes an importance sampling enhanced
Variational Autoencoder (VAE) approach for the estimation of the three- and four-
parameter MIRT Models. The key idea is to adopt a variational inference procedure
in machine learning literature to approximate the intractable marginal likelihood, and
further use importance sampling to boost the trained VAE with a better log-likelihood
approximation. Simulation studies are conducted to demonstrate the computational
efficiency and scalability of the new algorithm in comparison to the popular alternative
algorithms, i.e., Monte Carlo EM and Metropolis-Hastings Robbins-Monro methods.
The good performance of the proposed method is also illustrated by a NAEP multistage
testing dataset.