Bayesian Versus Frequentist Estimation for Structural Equation Models in Small Sample Contexts

A Systematic Review

Sanne C. Smid, Daniel McNeish, Milica Miočević, Rens van de Schoot

Research output: Contribution to journalReview article

Abstract

In small sample contexts, Bayesian estimation is often suggested as a viable alternative to frequentist estimation, such as maximum likelihood estimation. Our systematic literature review is the first study aggregating information from numerous simulation studies to present an overview of the performance of Bayesian and frequentist estimation for structural equation models with small sample sizes. We conclude that with small samples, the use of Bayesian estimation with diffuse default priors can result in severely biased estimates–the levels of bias are often even higher than when frequentist methods are used. This bias can only be decreased by incorporating prior information. We therefore recommend against naively using Bayesian estimation when samples are small, and encourage researchers to make well-considered decisions about all priors. For this purpose, we provide recommendations on how to construct thoughtful priors.

Original languageEnglish (US)
JournalStructural Equation Modeling
DOIs
StatePublished - Jan 1 2019

Fingerprint

Structural Equation Model
Bayesian Estimation
structural model
Small Sample
Literature Review
Small Sample Size
Prior Information
Maximum Likelihood Estimation
Biased
Recommendations
Simulation Study
Maximum likelihood estimation
Alternatives
trend
Context
Review
Systematic review
Bayesian estimation
Structural equation model
Small sample

Keywords

  • informative priors
  • Small samples
  • structural equation models
  • systematic review

ASJC Scopus subject areas

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

Cite this

Bayesian Versus Frequentist Estimation for Structural Equation Models in Small Sample Contexts : A Systematic Review. / Smid, Sanne C.; McNeish, Daniel; Miočević, Milica; van de Schoot, Rens.

In: Structural Equation Modeling, 01.01.2019.

Research output: Contribution to journalReview article

@article{c6e32ab96e854aa7aaf8d4d0e8eb106d,
title = "Bayesian Versus Frequentist Estimation for Structural Equation Models in Small Sample Contexts: A Systematic Review",
abstract = "In small sample contexts, Bayesian estimation is often suggested as a viable alternative to frequentist estimation, such as maximum likelihood estimation. Our systematic literature review is the first study aggregating information from numerous simulation studies to present an overview of the performance of Bayesian and frequentist estimation for structural equation models with small sample sizes. We conclude that with small samples, the use of Bayesian estimation with diffuse default priors can result in severely biased estimates–the levels of bias are often even higher than when frequentist methods are used. This bias can only be decreased by incorporating prior information. We therefore recommend against naively using Bayesian estimation when samples are small, and encourage researchers to make well-considered decisions about all priors. For this purpose, we provide recommendations on how to construct thoughtful priors.",
keywords = "informative priors, Small samples, structural equation models, systematic review",
author = "Smid, {Sanne C.} and Daniel McNeish and Milica Miočević and {van de Schoot}, Rens",
year = "2019",
month = "1",
day = "1",
doi = "10.1080/10705511.2019.1577140",
language = "English (US)",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "Psychology Press Ltd",

}

TY - JOUR

T1 - Bayesian Versus Frequentist Estimation for Structural Equation Models in Small Sample Contexts

T2 - A Systematic Review

AU - Smid, Sanne C.

AU - McNeish, Daniel

AU - Miočević, Milica

AU - van de Schoot, Rens

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In small sample contexts, Bayesian estimation is often suggested as a viable alternative to frequentist estimation, such as maximum likelihood estimation. Our systematic literature review is the first study aggregating information from numerous simulation studies to present an overview of the performance of Bayesian and frequentist estimation for structural equation models with small sample sizes. We conclude that with small samples, the use of Bayesian estimation with diffuse default priors can result in severely biased estimates–the levels of bias are often even higher than when frequentist methods are used. This bias can only be decreased by incorporating prior information. We therefore recommend against naively using Bayesian estimation when samples are small, and encourage researchers to make well-considered decisions about all priors. For this purpose, we provide recommendations on how to construct thoughtful priors.

AB - In small sample contexts, Bayesian estimation is often suggested as a viable alternative to frequentist estimation, such as maximum likelihood estimation. Our systematic literature review is the first study aggregating information from numerous simulation studies to present an overview of the performance of Bayesian and frequentist estimation for structural equation models with small sample sizes. We conclude that with small samples, the use of Bayesian estimation with diffuse default priors can result in severely biased estimates–the levels of bias are often even higher than when frequentist methods are used. This bias can only be decreased by incorporating prior information. We therefore recommend against naively using Bayesian estimation when samples are small, and encourage researchers to make well-considered decisions about all priors. For this purpose, we provide recommendations on how to construct thoughtful priors.

KW - informative priors

KW - Small samples

KW - structural equation models

KW - systematic review

UR - http://www.scopus.com/inward/record.url?scp=85066149725&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066149725&partnerID=8YFLogxK

U2 - 10.1080/10705511.2019.1577140

DO - 10.1080/10705511.2019.1577140

M3 - Review article

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

ER -