The Combined Effects of Measurement Error and Omitting Confounders in the Single-Mediator Model

Matthew S. Fritz, David A. Kenny, David Mackinnon

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Mediation analysis requires a number of strong assumptions be met in order to make valid causal inferences. Failing to account for violations of these assumptions, such as not modeling measurement error or omitting a common cause of the effects in the model, can bias the parameter estimates of the mediated effect. When the independent variable is perfectly reliable, for example when participants are randomly assigned to levels of treatment, measurement error in the mediator tends to underestimate the mediated effect, while the omission of a confounding variable of the mediator-to-outcome relation tends to overestimate the mediated effect. Violations of these two assumptions often co-occur, however, in which case the mediated effect could be overestimated, underestimated, or even, in very rare circumstances, unbiased. To explore the combined effect of measurement error and omitted confounders in the same model, the effect of each violation on the single-mediator model is first examined individually. Then the combined effect of having measurement error and omitted confounders in the same model is discussed. Throughout, an empirical example is provided to illustrate the effect of violating these assumptions on the mediated effect.

Original languageEnglish (US)
Pages (from-to)681-697
Number of pages17
JournalMultivariate Behavioral Research
Volume51
Issue number5
DOIs
StatePublished - Sep 2 2016

Fingerprint

Confounding Factors (Epidemiology)
Mediator
Measurement Error
Model
Tend
Causal Inference
Mediation
Modeling Error
Confounding
Valid

Keywords

  • Confounding
  • measurement error
  • mediation
  • sensitivity analysis

ASJC Scopus subject areas

  • Statistics and Probability
  • Experimental and Cognitive Psychology
  • Arts and Humanities (miscellaneous)

Cite this

The Combined Effects of Measurement Error and Omitting Confounders in the Single-Mediator Model. / Fritz, Matthew S.; Kenny, David A.; Mackinnon, David.

In: Multivariate Behavioral Research, Vol. 51, No. 5, 02.09.2016, p. 681-697.

Research output: Contribution to journalArticle

@article{3f42504bc075432bad027bc282b74c0f,
title = "The Combined Effects of Measurement Error and Omitting Confounders in the Single-Mediator Model",
abstract = "Mediation analysis requires a number of strong assumptions be met in order to make valid causal inferences. Failing to account for violations of these assumptions, such as not modeling measurement error or omitting a common cause of the effects in the model, can bias the parameter estimates of the mediated effect. When the independent variable is perfectly reliable, for example when participants are randomly assigned to levels of treatment, measurement error in the mediator tends to underestimate the mediated effect, while the omission of a confounding variable of the mediator-to-outcome relation tends to overestimate the mediated effect. Violations of these two assumptions often co-occur, however, in which case the mediated effect could be overestimated, underestimated, or even, in very rare circumstances, unbiased. To explore the combined effect of measurement error and omitted confounders in the same model, the effect of each violation on the single-mediator model is first examined individually. Then the combined effect of having measurement error and omitted confounders in the same model is discussed. Throughout, an empirical example is provided to illustrate the effect of violating these assumptions on the mediated effect.",
keywords = "Confounding, measurement error, mediation, sensitivity analysis",
author = "Fritz, {Matthew S.} and Kenny, {David A.} and David Mackinnon",
year = "2016",
month = "9",
day = "2",
doi = "10.1080/00273171.2016.1224154",
language = "English (US)",
volume = "51",
pages = "681--697",
journal = "Multivariate Behavioral Research",
issn = "0027-3171",
publisher = "Psychology Press Ltd",
number = "5",

}

TY - JOUR

T1 - The Combined Effects of Measurement Error and Omitting Confounders in the Single-Mediator Model

AU - Fritz, Matthew S.

AU - Kenny, David A.

AU - Mackinnon, David

PY - 2016/9/2

Y1 - 2016/9/2

N2 - Mediation analysis requires a number of strong assumptions be met in order to make valid causal inferences. Failing to account for violations of these assumptions, such as not modeling measurement error or omitting a common cause of the effects in the model, can bias the parameter estimates of the mediated effect. When the independent variable is perfectly reliable, for example when participants are randomly assigned to levels of treatment, measurement error in the mediator tends to underestimate the mediated effect, while the omission of a confounding variable of the mediator-to-outcome relation tends to overestimate the mediated effect. Violations of these two assumptions often co-occur, however, in which case the mediated effect could be overestimated, underestimated, or even, in very rare circumstances, unbiased. To explore the combined effect of measurement error and omitted confounders in the same model, the effect of each violation on the single-mediator model is first examined individually. Then the combined effect of having measurement error and omitted confounders in the same model is discussed. Throughout, an empirical example is provided to illustrate the effect of violating these assumptions on the mediated effect.

AB - Mediation analysis requires a number of strong assumptions be met in order to make valid causal inferences. Failing to account for violations of these assumptions, such as not modeling measurement error or omitting a common cause of the effects in the model, can bias the parameter estimates of the mediated effect. When the independent variable is perfectly reliable, for example when participants are randomly assigned to levels of treatment, measurement error in the mediator tends to underestimate the mediated effect, while the omission of a confounding variable of the mediator-to-outcome relation tends to overestimate the mediated effect. Violations of these two assumptions often co-occur, however, in which case the mediated effect could be overestimated, underestimated, or even, in very rare circumstances, unbiased. To explore the combined effect of measurement error and omitted confounders in the same model, the effect of each violation on the single-mediator model is first examined individually. Then the combined effect of having measurement error and omitted confounders in the same model is discussed. Throughout, an empirical example is provided to illustrate the effect of violating these assumptions on the mediated effect.

KW - Confounding

KW - measurement error

KW - mediation

KW - sensitivity analysis

UR - http://www.scopus.com/inward/record.url?scp=84991696415&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84991696415&partnerID=8YFLogxK

U2 - 10.1080/00273171.2016.1224154

DO - 10.1080/00273171.2016.1224154

M3 - Article

VL - 51

SP - 681

EP - 697

JO - Multivariate Behavioral Research

JF - Multivariate Behavioral Research

SN - 0027-3171

IS - 5

ER -