Skip to content

Graphical model framework for MML estimation of multidimensional IRT models

Research Method · All research methods
Evidence · 3 claims (3 for) · 1 study (1 theoretical), q2 · 0 of 1 report an effect size · 3 claims rest on one study

Description

The article presents a framework in which a multidimensional IRT model is represented as a directed acyclic graph whose nodes are random variables and whose edges encode conditional (in)dependence relations. The DAG is moralized, triangulated, and transformed into a junction tree, and the E-step of the EM algorithm is carried out via local computations on the tree. A stated advantage is that "these subsets of variables can be derived in a fully automatic way by applying algorithms to the graphical representation of the model", making the approach generally applicable and rendering hand derivations obsolete.

Accounts

  • Graphical model framework for deriving efficient MML estimation schemes for multidimensional IRT models: The article presents a framework in which a multidimensional IRT model is represented as a directed acyclic graph whose nodes are random variables and whose edges encode conditional (in)dependence relations. The DAG is moralized, triangulated, and transformed into a junction tree, and the E-step of the EM algorithm is carried out via local computations on the tree. A stated advantage is that "these subsets of variables can be derived in a fully automatic way by applying algorithms to the graphical representation of the model", making the approach generally applicable and rendering hand derivations obsolete. (Rijmen (2009))
  • Graphical model framework for exploiting conditional independence to make multidimensional IRT estimation tractable: The article presents a framework in which a statistical model is represented as a directed acyclic graph, moralized, triangulated, and converted into a junction tree whose cliques define conditionally independent subsets of latent variables. The author states that "the computational complexity of a multidimensional model is inversely related to the number of conditional independence relations one is willing to assume." The framework provides algorithms to partition the joint latent space so that brute force integration can be replaced by local computations on cliques, applied to a family of multidimensional IRT growth models. (Rijmen (2010))

Claims

-

Key Sources

  • Rijmen, F. (2009). Efficient Full Information Maximum Likelihood Estimation for Multidimensional IRT Models. ETS Research Report RR-09-03. http://www.ets.org/research/contact.html. doi:10.1002/j.2333-8504.2009.tb02160.x
  • Rijmen, F. (2010). Measuring Multidimensional Latent Growth. ETS Research Report RR-10-24. https://www.ets.org/research/contact.html
Page metadata
Field Value
type research-method
id graphical-model-framework-for-mml-estimation-of-multidimensional-irt-models
aliases graphical-model-framework-for-junction-tree-estimation-of-multidimensional-irt-growth-mode
title Graphical model framework for MML estimation of multidimensional IRT models
description The article presents a framework in which a multidimensional IRT model is represented as a directed acyclic graph whose nodes are random variables and whose edges encode conditional (in)dependence relations.
status review
moved_from theory/graphical-model-framework-mml-irt-estimation
generated by process:sweep-kinds on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 2 — rijmen-2009, rijmen-2010

Frontmatter is the page's machine-readable record — see the Schema & Guide.