Modeling relationships between two categorical variables when data are missing: Examining consequences of the missing data mechanism in an HIV data set

Shiela M. Strauss, David M. Rindskopf, Gregory P. Falkin

Research output: Contribution to journalArticle

Abstract

Analysts evaluating the strengths of relationships between variables in behavioral science research must often contend with the problem of missing data. Analyses are typically performed using data for cases that are either complete in all the variables, or assume that the data are missing at random. Often, these approaches yield biased results. Using empirical data, the current work explores the implications and consequences of using various statistical models to describe the association of two variables, one ordinal and one dichotomous, in which data are incomplete for the dichotomous variable. These models explicitly reflect the missing data mechanism; models that hypothesize nonignorable nonresponse are given particular attention. Both the statistical fit and substantive consequences of these models are examined. This new methodological approach to examining nonignorable nonresponse can be applied to many behavioral science data sets containing an ordinal variable.

Original languageEnglish (US)
Pages (from-to)471-500
Number of pages30
JournalMultivariate Behavioral Research
Volume36
Issue number4
StatePublished - 2001

Fingerprint

Missing Data Mechanism
Behavioral Sciences
Categorical variable
HIV
Behavioral Research
Statistical Models
Non-response
Modeling
response behavior
behavioral science
Ordinal Variables
Missing at Random
Missing Data
Statistical Model
Biased
Relationships
Datasets
Categorical
AIDS/HIV
Model

ASJC Scopus subject areas

  • Mathematics (miscellaneous)
  • Statistics and Probability
  • Psychology(all)
  • Experimental and Cognitive Psychology
  • Social Sciences (miscellaneous)

Cite this

Modeling relationships between two categorical variables when data are missing : Examining consequences of the missing data mechanism in an HIV data set. / Strauss, Shiela M.; Rindskopf, David M.; Falkin, Gregory P.

In: Multivariate Behavioral Research, Vol. 36, No. 4, 2001, p. 471-500.

Research output: Contribution to journalArticle

@article{6db55029596c4da4ba1217ff01b054d6,
title = "Modeling relationships between two categorical variables when data are missing: Examining consequences of the missing data mechanism in an HIV data set",
abstract = "Analysts evaluating the strengths of relationships between variables in behavioral science research must often contend with the problem of missing data. Analyses are typically performed using data for cases that are either complete in all the variables, or assume that the data are missing at random. Often, these approaches yield biased results. Using empirical data, the current work explores the implications and consequences of using various statistical models to describe the association of two variables, one ordinal and one dichotomous, in which data are incomplete for the dichotomous variable. These models explicitly reflect the missing data mechanism; models that hypothesize nonignorable nonresponse are given particular attention. Both the statistical fit and substantive consequences of these models are examined. This new methodological approach to examining nonignorable nonresponse can be applied to many behavioral science data sets containing an ordinal variable.",
author = "Strauss, {Shiela M.} and Rindskopf, {David M.} and Falkin, {Gregory P.}",
year = "2001",
language = "English (US)",
volume = "36",
pages = "471--500",
journal = "Multivariate Behavioral Research",
issn = "0027-3171",
publisher = "Psychology Press Ltd",
number = "4",

}

TY - JOUR

T1 - Modeling relationships between two categorical variables when data are missing

T2 - Examining consequences of the missing data mechanism in an HIV data set

AU - Strauss, Shiela M.

AU - Rindskopf, David M.

AU - Falkin, Gregory P.

PY - 2001

Y1 - 2001

N2 - Analysts evaluating the strengths of relationships between variables in behavioral science research must often contend with the problem of missing data. Analyses are typically performed using data for cases that are either complete in all the variables, or assume that the data are missing at random. Often, these approaches yield biased results. Using empirical data, the current work explores the implications and consequences of using various statistical models to describe the association of two variables, one ordinal and one dichotomous, in which data are incomplete for the dichotomous variable. These models explicitly reflect the missing data mechanism; models that hypothesize nonignorable nonresponse are given particular attention. Both the statistical fit and substantive consequences of these models are examined. This new methodological approach to examining nonignorable nonresponse can be applied to many behavioral science data sets containing an ordinal variable.

AB - Analysts evaluating the strengths of relationships between variables in behavioral science research must often contend with the problem of missing data. Analyses are typically performed using data for cases that are either complete in all the variables, or assume that the data are missing at random. Often, these approaches yield biased results. Using empirical data, the current work explores the implications and consequences of using various statistical models to describe the association of two variables, one ordinal and one dichotomous, in which data are incomplete for the dichotomous variable. These models explicitly reflect the missing data mechanism; models that hypothesize nonignorable nonresponse are given particular attention. Both the statistical fit and substantive consequences of these models are examined. This new methodological approach to examining nonignorable nonresponse can be applied to many behavioral science data sets containing an ordinal variable.

UR - http://www.scopus.com/inward/record.url?scp=0035528596&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035528596&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0035528596

VL - 36

SP - 471

EP - 500

JO - Multivariate Behavioral Research

JF - Multivariate Behavioral Research

SN - 0027-3171

IS - 4

ER -