What question did this study set out to answer?

This research aims to enhance computerized adaptive testing by integrating Bayesian methods and reinforcement learning techniques.

March 29, 2026Open Access

Deep Computerized Adaptive Testing

Key Points

This research aims to enhance computerized adaptive testing by integrating Bayesian methods and reinforcement learning techniques.
Developed a new CAT system utilizing Bayesian multivariate IRT.
Replaced MCMC simulations with direct sampling methods for efficiency.
Employed a double deep Q-learning algorithm to optimize item selection policy offline using a calibrated item bank.
Conducted simulation and real-data studies to assess performance.
Evaluated posterior variance reduction and test termination efficiency.
Achieved significant acceleration in item selection methods compared to traditional approaches.
Enhanced item selection through a reinforcement learning strategy, leading to faster test terminations.
Demonstrated consistent performance improvements in posterior variance reduction.

Abstract

Computerized adaptive tests (CATs) play a crucial role in educational assessment and diagnostic screening in behavioral health.Unlike traditional linear tests that administer a fixed set of pre-assembled items, CATs adaptively tailor the test to an examinee's latent trait level based on their previous responses.We introduce a novel CAT system that builds on recent advances in Bayesian multivariate IRT.Our approach leverages direct sampling from the latent factor posterior distributions, significantly accelerating existing information-theoretic item selection methods by eliminating the need for computationally intensive Markov Chain Monte Carlo (MCMC) simulations.To address the potential suboptimality of one-step-ahead item selection rules, we also develop a double deep Q-learning algorithm that efficiently learns an optimal item-selection policy offline using a calibrated item bank.Through simulation and real-data studies, we demonstrate that our approach not only accelerates existing item selection methods but also highlights the potential of reinforcement learning in CATs.Notably, our Q-learning-based strategy consistently achieves the fastest posterior variance reduction, leading to earlier test termination.These results demonstrate the promise of combining exact posterior sampling with reinforcement learning to deliver scalable, high-precision CATs.

Ask AI

Helpful

Bookmark

View Full Paper