GPIRT: A Gaussian Process Model for Item Response Theory
This addresses the challenge for researchers in psychometrics and social sciences who need accurate latent trait estimation without restrictive parametric assumptions, representing a novel method rather than an incremental improvement.
The authors tackled the problem of estimating latent traits from binary observed indicators in item response theory (IRT) when data deviates from parametric assumptions, by proposing GPIRT, a Bayesian nonparametric model using Gaussian process priors that simultaneously estimates flexible item response functions and recovers ability estimates for respondents.
The goal of item response theoretic (IRT) models is to provide estimates of latent traits from binary observed indicators and at the same time to learn the item response functions (IRFs) that map from latent trait to observed response. However, in many cases observed behavior can deviate significantly from the parametric assumptions of traditional IRT models. Nonparametric IRT models overcome these challenges by relaxing assumptions about the form of the IRFs, but standard tools are unable to simultaneously estimate flexible IRFs and recover ability estimates for respondents. We propose a Bayesian nonparametric model that solves this problem by placing Gaussian process priors on the latent functions defining the IRFs. This allows us to simultaneously relax assumptions about the shape of the IRFs while preserving the ability to estimate latent traits. This in turn allows us to easily extend the model to further tasks such as active learning. GPIRT therefore provides a simple and intuitive solution to several longstanding problems in the IRT literature.