Lower Level Mediation in Multilevel Models

David A. Kenny, Josephine D Korchmaros, Niall Bolger

Research output: Contribution to journalArticle

382 Citations (Scopus)

Abstract

Multilevel models are increasingly used to estimate models for hierarchical and repeated measures data. The authors discuss a model in which there is mediation at the lower level and the mediational links vary randomly across upper level units. One repeated measures example is a case in which a person's daily stressors affect his or her coping efforts, which affect his or her mood, and both links vary randomly across persons. Where there is mediation at the lower level and the mediational links vary randomly across upper level units, the formulas for the indirect effect and its standard error must be modified to include the covariance between the random effects. Because no standard method can estimate such a model, the authors developed an ad hoc method that is illustrated with real and simulated data. Limitations of this method and characteristics of an ideal method are discussed.

Original languageEnglish (US)
Pages (from-to)115-128
Number of pages14
JournalPsychological Methods
Volume8
Issue number2
DOIs
StatePublished - Jun 2003
Externally publishedYes

ASJC Scopus subject areas

  • Psychology(all)

Cite this

Lower Level Mediation in Multilevel Models. / Kenny, David A.; Korchmaros, Josephine D; Bolger, Niall.

In: Psychological Methods, Vol. 8, No. 2, 06.2003, p. 115-128.

Research output: Contribution to journalArticle

Kenny, David A. ; Korchmaros, Josephine D ; Bolger, Niall. / Lower Level Mediation in Multilevel Models. In: Psychological Methods. 2003 ; Vol. 8, No. 2. pp. 115-128.
@article{77e18e5abced438b921c12c3eafeb243,
title = "Lower Level Mediation in Multilevel Models",
abstract = "Multilevel models are increasingly used to estimate models for hierarchical and repeated measures data. The authors discuss a model in which there is mediation at the lower level and the mediational links vary randomly across upper level units. One repeated measures example is a case in which a person's daily stressors affect his or her coping efforts, which affect his or her mood, and both links vary randomly across persons. Where there is mediation at the lower level and the mediational links vary randomly across upper level units, the formulas for the indirect effect and its standard error must be modified to include the covariance between the random effects. Because no standard method can estimate such a model, the authors developed an ad hoc method that is illustrated with real and simulated data. Limitations of this method and characteristics of an ideal method are discussed.",
author = "Kenny, {David A.} and Korchmaros, {Josephine D} and Niall Bolger",
year = "2003",
month = "6",
doi = "10.1037/1082-989X.8.2.115",
language = "English (US)",
volume = "8",
pages = "115--128",
journal = "Psychological Methods",
issn = "1082-989X",
publisher = "American Psychological Association Inc.",
number = "2",

}

TY - JOUR

T1 - Lower Level Mediation in Multilevel Models

AU - Kenny, David A.

AU - Korchmaros, Josephine D

AU - Bolger, Niall

PY - 2003/6

Y1 - 2003/6

N2 - Multilevel models are increasingly used to estimate models for hierarchical and repeated measures data. The authors discuss a model in which there is mediation at the lower level and the mediational links vary randomly across upper level units. One repeated measures example is a case in which a person's daily stressors affect his or her coping efforts, which affect his or her mood, and both links vary randomly across persons. Where there is mediation at the lower level and the mediational links vary randomly across upper level units, the formulas for the indirect effect and its standard error must be modified to include the covariance between the random effects. Because no standard method can estimate such a model, the authors developed an ad hoc method that is illustrated with real and simulated data. Limitations of this method and characteristics of an ideal method are discussed.

AB - Multilevel models are increasingly used to estimate models for hierarchical and repeated measures data. The authors discuss a model in which there is mediation at the lower level and the mediational links vary randomly across upper level units. One repeated measures example is a case in which a person's daily stressors affect his or her coping efforts, which affect his or her mood, and both links vary randomly across persons. Where there is mediation at the lower level and the mediational links vary randomly across upper level units, the formulas for the indirect effect and its standard error must be modified to include the covariance between the random effects. Because no standard method can estimate such a model, the authors developed an ad hoc method that is illustrated with real and simulated data. Limitations of this method and characteristics of an ideal method are discussed.

UR - http://www.scopus.com/inward/record.url?scp=0141752952&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0141752952&partnerID=8YFLogxK

U2 - 10.1037/1082-989X.8.2.115

DO - 10.1037/1082-989X.8.2.115

M3 - Article

C2 - 12924810

AN - SCOPUS:0141752952

VL - 8

SP - 115

EP - 128

JO - Psychological Methods

JF - Psychological Methods

SN - 1082-989X

IS - 2

ER -