Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Generalized latent variable models with non-linear effects.

Dimitris Rizopoulos1, Irini Moustaki

  • 1Erasmus University Medical Center, Rotterdam, The Netherlands. d.rizopoulose@erasmusmc.nl

The British Journal of Mathematical and Statistical Psychology
|May 31, 2007
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Bayesian multivariate linear mixed-effects models with varied association structures.

Statistical methods in medical research·2026
Same author

A comparative analysis of corticosteroids and exclusive enteral nutrition induction therapy in children with small bowel Crohn's disease: results of two prospective cohorts.

European journal of pediatrics·2026
Same author

Clinical correlates for major bleeding during veno-arterial extracorporeal membrane oxygenation support: a time-dependent perspective.

Critical care (London, England)·2026
Same author

Plasma proteins and mechanisms involved in the evolvement of cardiac function after myocardial infarction.

Scientific reports·2026
Same author

DIF Analysis with Unknown Groups and Anchor Items.

Psychometrika·2026
Same author

Comparing the effectiveness of prostate cancer screening protocols: European Association of Urology- and European Randomized Study of Screening for Prostate Cancer-based strategies.

International journal of cancer·2026
Same journal

Proficiency order invariance of MLE, MAP, EAP, and WLE in item response theory.

The British journal of mathematical and statistical psychology·2026
Same journal

Bias and precision in true-score estimation.

The British journal of mathematical and statistical psychology·2026
Same journal

Polychoric correlations under the assumption of elliptical latent traits.

The British journal of mathematical and statistical psychology·2026
Same journal

Regularized reduced rank regression for mixed predictor and response variables.

The British journal of mathematical and statistical psychology·2026
Same journal

A multiple-choice SDT model for cognitive diagnosis models.

The British journal of mathematical and statistical psychology·2026
Same journal

Modular item response and structural equation modelling via measurement and uncertainty preserving parametric modelling.

The British journal of mathematical and statistical psychology·2026
See all related articles

This study introduces a new latent variable model framework for mixed responses, incorporating non-linear latent and covariate effects. This enhances item response theory models for complex real-world data analysis.

Area of Science:

  • Psychometrics
  • Statistical modeling
  • Latent variable analysis

Background:

  • Traditional item response models (e.g., factor analysis, logistic, multinomial) primarily account for main effects of latent variables.
  • These models often do not accommodate interaction or polynomial latent variable effects, limiting their application in scenarios with non-linear relationships.
  • Existing methods for non-linear latent terms are predominantly situated within the structural equation modeling framework.

Purpose of the Study:

  • To develop a flexible latent variable model framework capable of handling mixed response types (metric and categorical).
  • To incorporate non-linear latent variable effects and covariate effects within this unified framework.
  • To propose methods for parameter estimation and factor score computation for the developed non-linear model.

Related Experiment Videos

Main Methods:

  • A latent variable model framework is proposed for mixed response data.
  • Full maximum likelihood estimation is employed, utilizing a hybrid integration-maximization algorithm.
  • A factor score estimation method based on multiple imputation is introduced for the non-linear model.

Main Results:

  • The developed framework successfully integrates non-linear latent and covariate effects for mixed response data.
  • The hybrid integration-maximization algorithm provides a viable method for parameter estimation.
  • The multiple imputation approach offers a robust way to obtain factor scores for the non-linear model.

Conclusions:

  • The proposed latent variable model framework extends traditional item response theory by accommodating non-linear relationships and mixed response types.
  • The estimation and factor score methods provide practical tools for analyzing complex data structures.
  • This research offers advancements in statistical modeling for psychometric and related fields.