Bias, Type I Error Rates, and Statistical Power of a Latent Mediation Model in the Presence of Violations of Invariance

Margarita Olivera-Aguilar, Samuel H. Rikoon, Oscar Gonzalez, Yasemin Kisbu-Sakarya, David Mackinnon

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

When testing a statistical mediation model, it is assumed that factorial measurement invariance holds for the mediating construct across levels of the independent variable X. The consequences of failing to address the violations of measurement invariance in mediation models are largely unknown. The purpose of the present study was to systematically examine the impact of mediator noninvariance on the Type I error rates, statistical power, and relative bias in parameter estimates of the mediated effect in the single mediator model. The results of a large simulation study indicated that, in general, the mediated effect was robust to violations of invariance in loadings. In contrast, most conditions with violations of intercept invariance exhibited severely positively biased mediated effects, Type I error rates above acceptable levels, and statistical power larger than in the invariant conditions. The implications of these results are discussed and recommendations are offered.

Original languageEnglish (US)
JournalEducational and Psychological Measurement
DOIs
StateAccepted/In press - Jan 1 2017

Fingerprint

Statistical Power
Mediation
Type I Error Rate
Statistical Models
Invariance
Measurement Invariance
mediation
Mediator
trend
Intercept
Factorial
Biased
Recommendations
Simulation Study
Model
Unknown
Testing
simulation
Invariant
Estimate

Keywords

  • latent variables
  • measurement invariance
  • statistical mediation model

ASJC Scopus subject areas

  • Education
  • Developmental and Educational Psychology
  • Applied Psychology
  • Applied Mathematics

Cite this

Bias, Type I Error Rates, and Statistical Power of a Latent Mediation Model in the Presence of Violations of Invariance. / Olivera-Aguilar, Margarita; Rikoon, Samuel H.; Gonzalez, Oscar; Kisbu-Sakarya, Yasemin; Mackinnon, David.

In: Educational and Psychological Measurement, 01.01.2017.

Research output: Contribution to journalArticle

@article{5be39c49fa554818be5152392a22e8d4,
title = "Bias, Type I Error Rates, and Statistical Power of a Latent Mediation Model in the Presence of Violations of Invariance",
abstract = "When testing a statistical mediation model, it is assumed that factorial measurement invariance holds for the mediating construct across levels of the independent variable X. The consequences of failing to address the violations of measurement invariance in mediation models are largely unknown. The purpose of the present study was to systematically examine the impact of mediator noninvariance on the Type I error rates, statistical power, and relative bias in parameter estimates of the mediated effect in the single mediator model. The results of a large simulation study indicated that, in general, the mediated effect was robust to violations of invariance in loadings. In contrast, most conditions with violations of intercept invariance exhibited severely positively biased mediated effects, Type I error rates above acceptable levels, and statistical power larger than in the invariant conditions. The implications of these results are discussed and recommendations are offered.",
keywords = "latent variables, measurement invariance, statistical mediation model",
author = "Margarita Olivera-Aguilar and Rikoon, {Samuel H.} and Oscar Gonzalez and Yasemin Kisbu-Sakarya and David Mackinnon",
year = "2017",
month = "1",
day = "1",
doi = "10.1177/0013164416684169",
language = "English (US)",
journal = "Educational and Psychological Measurement",
issn = "0013-1644",
publisher = "SAGE Publications Inc.",

}

TY - JOUR

T1 - Bias, Type I Error Rates, and Statistical Power of a Latent Mediation Model in the Presence of Violations of Invariance

AU - Olivera-Aguilar, Margarita

AU - Rikoon, Samuel H.

AU - Gonzalez, Oscar

AU - Kisbu-Sakarya, Yasemin

AU - Mackinnon, David

PY - 2017/1/1

Y1 - 2017/1/1

N2 - When testing a statistical mediation model, it is assumed that factorial measurement invariance holds for the mediating construct across levels of the independent variable X. The consequences of failing to address the violations of measurement invariance in mediation models are largely unknown. The purpose of the present study was to systematically examine the impact of mediator noninvariance on the Type I error rates, statistical power, and relative bias in parameter estimates of the mediated effect in the single mediator model. The results of a large simulation study indicated that, in general, the mediated effect was robust to violations of invariance in loadings. In contrast, most conditions with violations of intercept invariance exhibited severely positively biased mediated effects, Type I error rates above acceptable levels, and statistical power larger than in the invariant conditions. The implications of these results are discussed and recommendations are offered.

AB - When testing a statistical mediation model, it is assumed that factorial measurement invariance holds for the mediating construct across levels of the independent variable X. The consequences of failing to address the violations of measurement invariance in mediation models are largely unknown. The purpose of the present study was to systematically examine the impact of mediator noninvariance on the Type I error rates, statistical power, and relative bias in parameter estimates of the mediated effect in the single mediator model. The results of a large simulation study indicated that, in general, the mediated effect was robust to violations of invariance in loadings. In contrast, most conditions with violations of intercept invariance exhibited severely positively biased mediated effects, Type I error rates above acceptable levels, and statistical power larger than in the invariant conditions. The implications of these results are discussed and recommendations are offered.

KW - latent variables

KW - measurement invariance

KW - statistical mediation model

UR - http://www.scopus.com/inward/record.url?scp=85034589241&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85034589241&partnerID=8YFLogxK

U2 - 10.1177/0013164416684169

DO - 10.1177/0013164416684169

M3 - Article

AN - SCOPUS:85034589241

JO - Educational and Psychological Measurement

JF - Educational and Psychological Measurement

SN - 0013-1644

ER -