EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep Computerized Adaptive Testing

Jiguang Li iD, Robert Gibbons, Veronika Ročková

Psychometrika · 2026-06 · 91 (3) · 1127-1148

DOI: 10.1017/psy.2026.10106 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Psychometrika, online erschienen am 27.03.2026. Verfasst von Jiguang Li und 2 weiteren Autor:innen. Beteiligte Einrichtung: Econometrics and Statistics, https://ror.org/024mw5h28The University of Chicago Booth School of Business, USA. Die Arbeit nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NIA NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

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 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 (RL) 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 RL to deliver scalable, high-precision CATs.

Schlagwörter

Stichwörter der Autor:innen

Bayesian methodsadaptive data collectiondeep Q-learningmultidimensional item response theory

Autor:innen und Institutionen

  • Econometrics and Statistics, https://ror.org/024mw5h28The University of Chicago Booth School of Business, USA.
  • Department of Statistics, https://ror.org/024mw5h28The University of Chicago, USA.

ORCID-Profile: Jiguang Li

Förderung

  • NIA NIH HHS

Zitieren

APA

Li, J., Gibbons, R., & Ročková, V. (2026). Deep Computerized Adaptive Testing. Psychometrika, 91(3), 1127-1148. https://doi.org/10.1017/psy.2026.10106

Vancouver

Li J, Gibbons R, Ročková V. Deep Computerized Adaptive Testing. Psychometrika. 2026;91(3):1127-1148. doi:10.1017/psy.2026.10106

BibTeX
@article{Li2026,
  author = {Li, Jiguang and Gibbons, Robert and Ročková, Veronika},
  title = {Deep Computerized Adaptive Testing},
  journal = {Psychometrika},
  year = {2026},
  volume = {91},
  number = {3},
  pages = {1127--1148},
  doi = {10.1017/psy.2026.10106}
}
Kostenlos starten →