Testing multilevel mediation using hierarchical linear models

Problems and solutions

Zhen Zhang, Michael J. Zyphur, Kristopher J. Preacher

Research output: Contribution to journalArticle

457 Citations (Scopus)

Abstract

Testing multilevel mediation using hierarchical linear modeling (HLM) has gained tremendous popularity in recent years. However, potential confounding in multilevel mediation effect estimates can arise in these models when within-group effects differ from between-group effects. This study summarizes three types of HLM-based multilevel mediation models, and then explains that in two types of these models confounding can be produced and erroneous conclusions may be derived when using popularly recommended procedures. A Monte Carlo simulation study illustrates that these procedures can underestimate or overestimate true mediation effects. Recommendations are provided for appropriately testing multilevel mediation and for differentiating within-group versus between-group effects in multilevel settings.

Original languageEnglish (US)
Pages (from-to)695-719
Number of pages25
JournalOrganizational Research Methods
Volume12
Issue number4
DOIs
StatePublished - Oct 2009

Fingerprint

Testing
Hierarchical linear models
Mediation
Hierarchical linear modeling
Mediation effect
Confounding
Monte Carlo simulation
Simulation study

Keywords

  • Hierarchical linear models
  • Mediation
  • Multilevel
  • Random coefficient regression

ASJC Scopus subject areas

  • Management of Technology and Innovation
  • Strategy and Management
  • Decision Sciences(all)

Cite this

Testing multilevel mediation using hierarchical linear models : Problems and solutions. / Zhang, Zhen; Zyphur, Michael J.; Preacher, Kristopher J.

In: Organizational Research Methods, Vol. 12, No. 4, 10.2009, p. 695-719.

Research output: Contribution to journalArticle

Zhang, Zhen ; Zyphur, Michael J. ; Preacher, Kristopher J. / Testing multilevel mediation using hierarchical linear models : Problems and solutions. In: Organizational Research Methods. 2009 ; Vol. 12, No. 4. pp. 695-719.
@article{7692484ed7ea4300a8946bd83b82ba2a,
title = "Testing multilevel mediation using hierarchical linear models: Problems and solutions",
abstract = "Testing multilevel mediation using hierarchical linear modeling (HLM) has gained tremendous popularity in recent years. However, potential confounding in multilevel mediation effect estimates can arise in these models when within-group effects differ from between-group effects. This study summarizes three types of HLM-based multilevel mediation models, and then explains that in two types of these models confounding can be produced and erroneous conclusions may be derived when using popularly recommended procedures. A Monte Carlo simulation study illustrates that these procedures can underestimate or overestimate true mediation effects. Recommendations are provided for appropriately testing multilevel mediation and for differentiating within-group versus between-group effects in multilevel settings.",
keywords = "Hierarchical linear models, Mediation, Multilevel, Random coefficient regression",
author = "Zhen Zhang and Zyphur, {Michael J.} and Preacher, {Kristopher J.}",
year = "2009",
month = "10",
doi = "10.1177/1094428108327450",
language = "English (US)",
volume = "12",
pages = "695--719",
journal = "Organizational Research Methods",
issn = "1094-4281",
publisher = "SAGE Publications Inc.",
number = "4",

}

TY - JOUR

T1 - Testing multilevel mediation using hierarchical linear models

T2 - Problems and solutions

AU - Zhang, Zhen

AU - Zyphur, Michael J.

AU - Preacher, Kristopher J.

PY - 2009/10

Y1 - 2009/10

N2 - Testing multilevel mediation using hierarchical linear modeling (HLM) has gained tremendous popularity in recent years. However, potential confounding in multilevel mediation effect estimates can arise in these models when within-group effects differ from between-group effects. This study summarizes three types of HLM-based multilevel mediation models, and then explains that in two types of these models confounding can be produced and erroneous conclusions may be derived when using popularly recommended procedures. A Monte Carlo simulation study illustrates that these procedures can underestimate or overestimate true mediation effects. Recommendations are provided for appropriately testing multilevel mediation and for differentiating within-group versus between-group effects in multilevel settings.

AB - Testing multilevel mediation using hierarchical linear modeling (HLM) has gained tremendous popularity in recent years. However, potential confounding in multilevel mediation effect estimates can arise in these models when within-group effects differ from between-group effects. This study summarizes three types of HLM-based multilevel mediation models, and then explains that in two types of these models confounding can be produced and erroneous conclusions may be derived when using popularly recommended procedures. A Monte Carlo simulation study illustrates that these procedures can underestimate or overestimate true mediation effects. Recommendations are provided for appropriately testing multilevel mediation and for differentiating within-group versus between-group effects in multilevel settings.

KW - Hierarchical linear models

KW - Mediation

KW - Multilevel

KW - Random coefficient regression

UR - http://www.scopus.com/inward/record.url?scp=70350076901&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70350076901&partnerID=8YFLogxK

U2 - 10.1177/1094428108327450

DO - 10.1177/1094428108327450

M3 - Article

VL - 12

SP - 695

EP - 719

JO - Organizational Research Methods

JF - Organizational Research Methods

SN - 1094-4281

IS - 4

ER -