Comparing Models of Change to Estimate the Mediated Effect in the Pretest–Posttest Control Group Design

Matthew J. Valente, David Mackinnon

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Models to assess mediation in the pretest–posttest control group design are understudied in the behavioral sciences even though it is the design of choice for evaluating experimental manipulations. The article provides analytical comparisons of the four most commonly used models to estimate the mediated effect in this design: analysis of covariance (ANCOVA), difference score, residualized change score, and cross-sectional model. Each of these models is fitted using a latent change score specification and a simulation study assessed bias, Type I error, power, and confidence interval coverage of the four models. All but the ANCOVA model make stringent assumptions about the stability and cross-lagged relations of the mediator and outcome that might not be plausible in real-world applications. When these assumptions do not hold, Type I error and statistical power results suggest that only the ANCOVA model has good performance. The four models are applied to an empirical example.

Original languageEnglish (US)
Pages (from-to)428-450
Number of pages23
JournalStructural Equation Modeling
Volume24
Issue number3
DOIs
StatePublished - May 4 2017

Fingerprint

Analysis of Covariance
Estimate
Group
Type I error
Model
Statistical Power
Mediation
Design
behavioral science
Mediator
Real-world Applications
mediation
manipulation
Confidence interval
Manipulation
Coverage
confidence
coverage
Simulation Study
Specification

Keywords

  • longitudinal mediation
  • mediation
  • pretest–posttest design
  • two waves

ASJC Scopus subject areas

  • Decision Sciences(all)
  • Modeling and Simulation
  • Sociology and Political Science
  • Economics, Econometrics and Finance(all)

Cite this

Comparing Models of Change to Estimate the Mediated Effect in the Pretest–Posttest Control Group Design. / Valente, Matthew J.; Mackinnon, David.

In: Structural Equation Modeling, Vol. 24, No. 3, 04.05.2017, p. 428-450.

Research output: Contribution to journalArticle

@article{a2b6ceb9f5cb427c90e85314b1b28aec,
title = "Comparing Models of Change to Estimate the Mediated Effect in the Pretest–Posttest Control Group Design",
abstract = "Models to assess mediation in the pretest–posttest control group design are understudied in the behavioral sciences even though it is the design of choice for evaluating experimental manipulations. The article provides analytical comparisons of the four most commonly used models to estimate the mediated effect in this design: analysis of covariance (ANCOVA), difference score, residualized change score, and cross-sectional model. Each of these models is fitted using a latent change score specification and a simulation study assessed bias, Type I error, power, and confidence interval coverage of the four models. All but the ANCOVA model make stringent assumptions about the stability and cross-lagged relations of the mediator and outcome that might not be plausible in real-world applications. When these assumptions do not hold, Type I error and statistical power results suggest that only the ANCOVA model has good performance. The four models are applied to an empirical example.",
keywords = "longitudinal mediation, mediation, pretest–posttest design, two waves",
author = "Valente, {Matthew J.} and David Mackinnon",
year = "2017",
month = "5",
day = "4",
doi = "10.1080/10705511.2016.1274657",
language = "English (US)",
volume = "24",
pages = "428--450",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "Psychology Press Ltd",
number = "3",

}

TY - JOUR

T1 - Comparing Models of Change to Estimate the Mediated Effect in the Pretest–Posttest Control Group Design

AU - Valente, Matthew J.

AU - Mackinnon, David

PY - 2017/5/4

Y1 - 2017/5/4

N2 - Models to assess mediation in the pretest–posttest control group design are understudied in the behavioral sciences even though it is the design of choice for evaluating experimental manipulations. The article provides analytical comparisons of the four most commonly used models to estimate the mediated effect in this design: analysis of covariance (ANCOVA), difference score, residualized change score, and cross-sectional model. Each of these models is fitted using a latent change score specification and a simulation study assessed bias, Type I error, power, and confidence interval coverage of the four models. All but the ANCOVA model make stringent assumptions about the stability and cross-lagged relations of the mediator and outcome that might not be plausible in real-world applications. When these assumptions do not hold, Type I error and statistical power results suggest that only the ANCOVA model has good performance. The four models are applied to an empirical example.

AB - Models to assess mediation in the pretest–posttest control group design are understudied in the behavioral sciences even though it is the design of choice for evaluating experimental manipulations. The article provides analytical comparisons of the four most commonly used models to estimate the mediated effect in this design: analysis of covariance (ANCOVA), difference score, residualized change score, and cross-sectional model. Each of these models is fitted using a latent change score specification and a simulation study assessed bias, Type I error, power, and confidence interval coverage of the four models. All but the ANCOVA model make stringent assumptions about the stability and cross-lagged relations of the mediator and outcome that might not be plausible in real-world applications. When these assumptions do not hold, Type I error and statistical power results suggest that only the ANCOVA model has good performance. The four models are applied to an empirical example.

KW - longitudinal mediation

KW - mediation

KW - pretest–posttest design

KW - two waves

UR - http://www.scopus.com/inward/record.url?scp=85011838352&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85011838352&partnerID=8YFLogxK

U2 - 10.1080/10705511.2016.1274657

DO - 10.1080/10705511.2016.1274657

M3 - Article

VL - 24

SP - 428

EP - 450

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 3

ER -