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

Testing parameters in structural equation modeling: every "one" matters.

R Gonzalez1, D Griffin

  • 1Department of Psychology, University of Michigan, Ann Arbor, Michigan 48109, USA. gonzo@umich.edu

Psychological Methods
|September 26, 2001
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

Quantifying the computational capability of a nanomagnetic reservoir computing platform with emergent magnetisation dynamics.

Nanotechnology·2022
Same author

Outcomes of Patients with a pH<7.0 Presenting to the Emergency Department (OPpHED Study).

Irish medical journal·2022
Same author

Highly realistic cadaveric trauma simulation of the multiply injured battlefield casualty: an international, multidisciplinary exercise in far-forward surgical management.

Injury·2020
Same author

Observations on Functional Disorders of the Spinal Cord, and Their Connexion with Hysterical, Nervous, and Other Diseases. Illustrated by Cases, Selected Chiefly from the Reports of the Pallas, Kenry, and Currah Dispensaries.

The London medical and physical journal·2018
Same author

Observations on Functional Disorders of the Spinal Cord, and Their Connexion with Hysteric, Nervous, and Other Diseases; Illustrated by Cases, Selected Chiefly from the Reports of the Pallas-Kenry and Currah Dispensaries.

The London medical and physical journal·2018
Same author

Observations on Functional Disorders of the Spinal Cord, and Their Connexion with Hysterical, Nervous, and Other Diseases; Illustrated by Cases, Selected Chiefly from the Reports of the Pallas-Kenry and Currah Dispensaries.

The London medical and physical journal·2018
Same journal

Bayesian evaluation for latent variable models: A tutorial on computing information criteria and bayes factors with the r package bleval.

Psychological methods·2026
Same journal

A stochastic block prior for clustering in graphical models.

Psychological methods·2026
Same journal

Three-level vector autoregressive models.

Psychological methods·2026
Same journal

Scaling cognitive modeling to big data: A deep learning approach to studying individual differences in evidence accumulation model parameters.

Psychological methods·2026
Same journal

Best practices in multilevel modeling for within-cluster group comparisons: An evaluation of coding strategies reflecting group composition and heterogeneity.

Psychological methods·2026
Same journal

A unified framework for psychometrics in experimental psychology: The standardized generalized hierarchical factor model.

Psychological methods·2026
See all related articles

Standard errors in structural equation modeling can vary based on model identification, affecting significance tests. The likelihood ratio test offers an invariant alternative for testing parameter significance.

Area of Science:

  • Statistics
  • Psychometrics
  • Quantitative Psychology

Background:

  • Structural equation modeling (SEM) is widely used in various scientific disciplines.
  • Standard error estimation in SEM can be problematic and sensitive to model identification choices.
  • Inconsistent standard errors can lead to differing conclusions about parameter significance.

Purpose of the Study:

  • To highlight the issue of non-invariant standard errors in structural equation modeling.
  • To propose an alternative method for robust parameter significance testing.
  • To ensure consistency in statistical inference across different model identifications.

Main Methods:

  • Analysis of standard error estimation in common structural equation modeling software.
  • Comparison of parameter significance tests across alternative model identification strategies.

Related Experiment Videos

  • Evaluation of the likelihood ratio test as an invariant alternative.

    • Main Results:

    • Standard errors and associated Z-tests for parameters are shown to be sensitive to model identification.
    • Equivalent model identifications can produce different conclusions regarding parameter significance.
    • The likelihood ratio test demonstrates invariance across different identification methods.

    Conclusions:

    • The choice of model identification strategy can arbitrarily affect parameter significance in SEM.
    • Researchers may draw different conclusions from statistically equivalent models.
    • The likelihood ratio test is recommended for invariant and reliable testing of parameter significance in SEM.