EUVIMEDEUVIMED entdecken
Crossref

A supervised learning approach to estimating IRT models in small samples

Dmitry I Belov iD, Oliver Lüdtke, Esther Ulitzsch

British Journal of Mathematical and Statistical Psychology · 2026

DOI: 10.1111/bmsp.12396 ↗

Kurz gefasst

Fachartikel in der Zeitschrift British Journal of Mathematical and Statistical Psychology, 2026. Verfasst von Dmitry I Belov und 2 weiteren Autor:innen. Beteiligte Einrichtung: Law School Admission Council, Newtown, Pennsylvania, USA. Hauptthemen laut MEDLINE: Psychometrics, Supervised Machine Learning, Models, Statistical. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 64 Literaturangaben. Gefördert von: Norges Forskningsråd, Research Council of Norway through its Centres of Excellence scheme.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Existing estimators of parameters of item response theory (IRT) models exploit the likelihood function. In small samples, however, the IRT likelihood oftentimes contains little informative value, potentially resulting in biased and/or unstable parameter estimates and large standard errors. To facilitate small‐sample IRT estimation, we introduce a novel approach that does not rely on the likelihood. Our estimation approach derives features from response data and then maps the features to item parameters using a neural network (NN). We describe and evaluate our approach for the three‐parameter logistic model; however, it is applicable to any model with an item characteristic curve. Three types of NNs are developed, supporting the obtainment of both point estimates and confidence intervals for IRT model parameters. The results of a simulation study demonstrate that these NNs perform better than Bayesian estimation using Markov chain Monte Carlo methods in terms of the quality of the point estimates and confidence intervals while also being much faster. These properties facilitate (1) pretesting items in a real‐time testing environment, (2) pretesting more items and (3) pretesting items only in a secured environment to eradicate possible compromise of new items in online testing.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

PsychometricsSupervised Machine LearningModels, StatisticalHumansBayes TheoremLikelihood FunctionsComputer SimulationNeural Networks, ComputerMonte Carlo MethodMarkov ChainsLogistic ModelsSample Size

Stichwörter der Autor:innen

item response theorylikelihood‐free estimationneural networkssmall‐sample estimation

Autor:innen und Institutionen

  • Law School Admission Council, Newtown, Pennsylvania, USA.
  • IPN - Leibniz Institute for Science and Mathematics Education, Kiel, Germany.
  • Centre for International Student Assessment, Munich, Germany.
  • Centre for Educational Measurement, University of Oslo, Oslo, Norway.
  • Centre for Research on Equality in Education, University of Oslo, Oslo, Norway.

ORCID-Profile: Dmitry I Belov

Förderung

  • Norges Forskningsråd
  • Research Council of Norway through its Centres of Excellence scheme

Zitieren

APA

Belov, D. I., Lüdtke, O., & Ulitzsch, E. (2026). A supervised learning approach to estimating IRT models in small samples. British Journal of Mathematical and Statistical Psychology. https://doi.org/10.1111/bmsp.12396

Vancouver

Belov DI, Lüdtke O, Ulitzsch E. A supervised learning approach to estimating IRT models in small samples. British Journal of Mathematical and Statistical Psychology. 2026. doi:10.1111/bmsp.12396

BibTeX
@article{Belov2026,
  author = {Belov, Dmitry I and Lüdtke, Oliver and Ulitzsch, Esther},
  title = {A supervised learning approach to estimating IRT models in small samples},
  journal = {British Journal of Mathematical and Statistical Psychology},
  year = {2026},
  doi = {10.1111/bmsp.12396}
}
Kostenlos starten →