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

Heteroscedastic one-factor models and marginal maximum likelihood estimation.

David J Hessen1, Conor V Dolan

  • 1Department of Methodology and Statistics, Utrecht University, Utrecht, The Netherlands. D.J.Hessen@uu.nl

The British Journal of Mathematical and Statistical Psychology
|October 16, 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

Genetics of major depressive disorder in a homogeneous population with uniform phenotyping.

Molecular psychiatry·2026
Same author

Validation of the ED-15-NL, a Brief Session-by-Session Questionnaire to Measure Changes in Eating Disorder Cognitions and Behaviors.

The International journal of eating disorders·2026
Same author

Causation Between Smoking Quantity and Depressive Symptoms in Young Adults: Evidence From Novel Cross-Lagged Twin Models.

medRxiv : the preprint server for health sciences·2025
Same author

The Power to Resolve Cultural Transmission and Sibling Interaction Using Polygenic Scores.

Behavior genetics·2025
Same author

Genetic and environmental contributions to variation in plasma phosphorylated tau 217.

Alzheimer's & dementia : the journal of the Alzheimer's Association·2025
Same author

Measurement Error and Power in Family-Based Extensions to Mendelian Randomization.

Behavior genetics·2025
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 new heteroscedastic one-factor models and a likelihood ratio test for analyzing data with varying residual variances. The proposed methods demonstrate robustness and power in simulations, offering practical applications for real-world data analysis.

Area of Science:

  • Psychometrics
  • Statistical Modeling
  • Factor Analysis

Background:

  • Traditional one-factor models assume equal residual variances (homoscedasticity), which may not hold in practice.
  • Heteroscedasticity, where residual variances differ, can bias results in standard factor analysis.
  • There is a need for flexible models that account for varying residual variances in factor analysis.

Purpose of the Study:

  • To introduce a general class of heteroscedastic one-factor models.
  • To propose a marginal maximum likelihood (MML) procedure for parameter estimation in these models.
  • To develop and evaluate a likelihood ratio (LR) test for comparing homoscedastic and heteroscedastic one-factor models.

Main Methods:

  • Modeling residual variances as parametric functions of the one-dimensional factor score.

Related Experiment Videos

  • Employing MML estimation under assumptions of conditional multivariate normality and normality of the factor score.
  • Deriving an LR test to contrast homoscedastic and heteroscedastic models.

    • Main Results:

    • Simulation studies confirm the robustness and power of the LR test under small test length and sample size conditions.
    • The study identifies conditions influencing the power to detect heteroscedasticity.
    • The MML estimation and LR test procedures are demonstrated on real data.

    Conclusions:

    • The proposed heteroscedastic one-factor models and MML estimation provide a valuable extension to traditional factor analysis.
    • The LR test is a reliable tool for detecting heteroscedasticity in one-factor models.
    • The methods offer practical utility for analyzing complex data structures in various scientific fields.