Bias attenuation results for nondifferentially mismeasured ordinal and coarsened confounders

Elizabeth Leigh Ogburn, Tyler J. Vanderweele

Research output: Contribution to journalArticle

Abstract

Suppose we are interested in the effect of a binary treatment on an outcome where that relationship is confounded by an ordinal confounder. We assume that the true confounder is not observed but, rather, we observe a nondifferentially mismeasured version of it. We show that, under certain monotonicity assumptions about its effect on the treatment and on the outcome, an effect measure controlling for the mismeasured confounder will fall between the corresponding crude and true effect measures. We also present results for coarsened and, under further assumptions, multiple misclassified confounders.

Original languageEnglish (US)
Pages (from-to)241-248
Number of pages8
JournalBiometrika
Volume100
Issue number1
DOIs
StatePublished - Mar 2013
Externally publishedYes

Fingerprint

Attenuation
Outcome Assessment (Health Care)
Monotonicity
Binary

Keywords

  • Bias
  • Confounder
  • Measurement error
  • Misclassification

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Agricultural and Biological Sciences (miscellaneous)
  • Statistics and Probability
  • Mathematics(all)
  • Applied Mathematics
  • Statistics, Probability and Uncertainty

Cite this

Bias attenuation results for nondifferentially mismeasured ordinal and coarsened confounders. / Ogburn, Elizabeth Leigh; Vanderweele, Tyler J.

In: Biometrika, Vol. 100, No. 1, 03.2013, p. 241-248.

Research output: Contribution to journalArticle

@article{8e4b94b28bdd47f6bc9910df6ea0e65a,
title = "Bias attenuation results for nondifferentially mismeasured ordinal and coarsened confounders",
abstract = "Suppose we are interested in the effect of a binary treatment on an outcome where that relationship is confounded by an ordinal confounder. We assume that the true confounder is not observed but, rather, we observe a nondifferentially mismeasured version of it. We show that, under certain monotonicity assumptions about its effect on the treatment and on the outcome, an effect measure controlling for the mismeasured confounder will fall between the corresponding crude and true effect measures. We also present results for coarsened and, under further assumptions, multiple misclassified confounders.",
keywords = "Bias, Confounder, Measurement error, Misclassification",
author = "Ogburn, {Elizabeth Leigh} and Vanderweele, {Tyler J.}",
year = "2013",
month = "3",
doi = "10.1093/biomet/ass054",
language = "English (US)",
volume = "100",
pages = "241--248",
journal = "Biometrika",
issn = "0006-3444",
publisher = "Oxford University Press",
number = "1",

}

TY - JOUR

T1 - Bias attenuation results for nondifferentially mismeasured ordinal and coarsened confounders

AU - Ogburn, Elizabeth Leigh

AU - Vanderweele, Tyler J.

PY - 2013/3

Y1 - 2013/3

N2 - Suppose we are interested in the effect of a binary treatment on an outcome where that relationship is confounded by an ordinal confounder. We assume that the true confounder is not observed but, rather, we observe a nondifferentially mismeasured version of it. We show that, under certain monotonicity assumptions about its effect on the treatment and on the outcome, an effect measure controlling for the mismeasured confounder will fall between the corresponding crude and true effect measures. We also present results for coarsened and, under further assumptions, multiple misclassified confounders.

AB - Suppose we are interested in the effect of a binary treatment on an outcome where that relationship is confounded by an ordinal confounder. We assume that the true confounder is not observed but, rather, we observe a nondifferentially mismeasured version of it. We show that, under certain monotonicity assumptions about its effect on the treatment and on the outcome, an effect measure controlling for the mismeasured confounder will fall between the corresponding crude and true effect measures. We also present results for coarsened and, under further assumptions, multiple misclassified confounders.

KW - Bias

KW - Confounder

KW - Measurement error

KW - Misclassification

UR - http://www.scopus.com/inward/record.url?scp=84874772545&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84874772545&partnerID=8YFLogxK

U2 - 10.1093/biomet/ass054

DO - 10.1093/biomet/ass054

M3 - Article

C2 - 24014285

AN - SCOPUS:84874772545

VL - 100

SP - 241

EP - 248

JO - Biometrika

JF - Biometrika

SN - 0006-3444

IS - 1

ER -