Sensitivity Analysis of Multiple Informant Models When Data Are Not Missing at Random

Shelley A. Blozis, Xiaojia Ge, Violet Shu Xu, Misaki N. Natsuaki, Daniel S. Shaw, Jenae M. Neiderhiser, Laura V. Scaramella, Leslie D. Leve, David Reiss

Research output: Contribution to journalArticle

Abstract

Missing data are common in studies that rely on multiple informant data to evaluate relationships among variables for distinguishable individuals clustered within groups. Estimation of structural equation models using raw data allows for incomplete data, and so all groups can be retained for analysis even if only 1 member of a group contributes data. Statistical inference is based on the assumption that data are missing completely at random or missing at random. Importantly, whether or not data are missing is assumed to be independent of the missing data. A saturated correlates model that incorporates correlates of the missingness or the missing data into an analysis and multiple imputation that might also use such correlates offer advantages over the standard implementation of SEM when data are not missing at random because these approaches could result in a data analysis problem for which the missingness is ignorable. This article considers these approaches in an analysis of family data to assess the sensitivity of parameter estimates and statistical inferences to assumptions about missing data, a strategy that could be easily implemented using SEM software.

Original languageEnglish (US)
Pages (from-to)283-298
Number of pages16
JournalStructural Equation Modeling
Volume20
Issue number2
DOIs
StatePublished - Apr 1 2013

Fingerprint

Missing at Random
Multiple Models
Sensitivity analysis
Sensitivity Analysis
Scanning electron microscopy
Missing Data
Correlate
Statistical Inference
Missing Completely at Random
Multiple Imputation
Structural Equation Model
Incomplete Data
Missing data
Data analysis
Correlates
Group
Software
structural model
Evaluate
Statistical inference

Keywords

  • auxiliary variables
  • missing data
  • missing not at random
  • multiple imputation
  • multiple informant data

ASJC Scopus subject areas

  • Modeling and Simulation
  • Decision Sciences(all)
  • Economics, Econometrics and Finance(all)
  • Sociology and Political Science

Cite this

Blozis, S. A., Ge, X., Xu, V. S., Natsuaki, M. N., Shaw, D. S., Neiderhiser, J. M., ... Reiss, D. (2013). Sensitivity Analysis of Multiple Informant Models When Data Are Not Missing at Random. Structural Equation Modeling, 20(2), 283-298. https://doi.org/10.1080/10705511.2013.769393

Sensitivity Analysis of Multiple Informant Models When Data Are Not Missing at Random. / Blozis, Shelley A.; Ge, Xiaojia; Xu, Violet Shu; Natsuaki, Misaki N.; Shaw, Daniel S.; Neiderhiser, Jenae M.; Scaramella, Laura V.; Leve, Leslie D.; Reiss, David.

In: Structural Equation Modeling, Vol. 20, No. 2, 01.04.2013, p. 283-298.

Research output: Contribution to journalArticle

Blozis, SA, Ge, X, Xu, VS, Natsuaki, MN, Shaw, DS, Neiderhiser, JM, Scaramella, LV, Leve, LD & Reiss, D 2013, 'Sensitivity Analysis of Multiple Informant Models When Data Are Not Missing at Random', Structural Equation Modeling, vol. 20, no. 2, pp. 283-298. https://doi.org/10.1080/10705511.2013.769393
Blozis, Shelley A. ; Ge, Xiaojia ; Xu, Violet Shu ; Natsuaki, Misaki N. ; Shaw, Daniel S. ; Neiderhiser, Jenae M. ; Scaramella, Laura V. ; Leve, Leslie D. ; Reiss, David. / Sensitivity Analysis of Multiple Informant Models When Data Are Not Missing at Random. In: Structural Equation Modeling. 2013 ; Vol. 20, No. 2. pp. 283-298.
@article{4902f14379f44a198a0321c629e632ed,
title = "Sensitivity Analysis of Multiple Informant Models When Data Are Not Missing at Random",
abstract = "Missing data are common in studies that rely on multiple informant data to evaluate relationships among variables for distinguishable individuals clustered within groups. Estimation of structural equation models using raw data allows for incomplete data, and so all groups can be retained for analysis even if only 1 member of a group contributes data. Statistical inference is based on the assumption that data are missing completely at random or missing at random. Importantly, whether or not data are missing is assumed to be independent of the missing data. A saturated correlates model that incorporates correlates of the missingness or the missing data into an analysis and multiple imputation that might also use such correlates offer advantages over the standard implementation of SEM when data are not missing at random because these approaches could result in a data analysis problem for which the missingness is ignorable. This article considers these approaches in an analysis of family data to assess the sensitivity of parameter estimates and statistical inferences to assumptions about missing data, a strategy that could be easily implemented using SEM software.",
keywords = "auxiliary variables, missing data, missing not at random, multiple imputation, multiple informant data",
author = "Blozis, {Shelley A.} and Xiaojia Ge and Xu, {Violet Shu} and Natsuaki, {Misaki N.} and Shaw, {Daniel S.} and Neiderhiser, {Jenae M.} and Scaramella, {Laura V.} and Leve, {Leslie D.} and David Reiss",
year = "2013",
month = "4",
day = "1",
doi = "10.1080/10705511.2013.769393",
language = "English (US)",
volume = "20",
pages = "283--298",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "Psychology Press Ltd",
number = "2",

}

TY - JOUR

T1 - Sensitivity Analysis of Multiple Informant Models When Data Are Not Missing at Random

AU - Blozis, Shelley A.

AU - Ge, Xiaojia

AU - Xu, Violet Shu

AU - Natsuaki, Misaki N.

AU - Shaw, Daniel S.

AU - Neiderhiser, Jenae M.

AU - Scaramella, Laura V.

AU - Leve, Leslie D.

AU - Reiss, David

PY - 2013/4/1

Y1 - 2013/4/1

N2 - Missing data are common in studies that rely on multiple informant data to evaluate relationships among variables for distinguishable individuals clustered within groups. Estimation of structural equation models using raw data allows for incomplete data, and so all groups can be retained for analysis even if only 1 member of a group contributes data. Statistical inference is based on the assumption that data are missing completely at random or missing at random. Importantly, whether or not data are missing is assumed to be independent of the missing data. A saturated correlates model that incorporates correlates of the missingness or the missing data into an analysis and multiple imputation that might also use such correlates offer advantages over the standard implementation of SEM when data are not missing at random because these approaches could result in a data analysis problem for which the missingness is ignorable. This article considers these approaches in an analysis of family data to assess the sensitivity of parameter estimates and statistical inferences to assumptions about missing data, a strategy that could be easily implemented using SEM software.

AB - Missing data are common in studies that rely on multiple informant data to evaluate relationships among variables for distinguishable individuals clustered within groups. Estimation of structural equation models using raw data allows for incomplete data, and so all groups can be retained for analysis even if only 1 member of a group contributes data. Statistical inference is based on the assumption that data are missing completely at random or missing at random. Importantly, whether or not data are missing is assumed to be independent of the missing data. A saturated correlates model that incorporates correlates of the missingness or the missing data into an analysis and multiple imputation that might also use such correlates offer advantages over the standard implementation of SEM when data are not missing at random because these approaches could result in a data analysis problem for which the missingness is ignorable. This article considers these approaches in an analysis of family data to assess the sensitivity of parameter estimates and statistical inferences to assumptions about missing data, a strategy that could be easily implemented using SEM software.

KW - auxiliary variables

KW - missing data

KW - missing not at random

KW - multiple imputation

KW - multiple informant data

UR - http://www.scopus.com/inward/record.url?scp=84876270832&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84876270832&partnerID=8YFLogxK

U2 - 10.1080/10705511.2013.769393

DO - 10.1080/10705511.2013.769393

M3 - Article

VL - 20

SP - 283

EP - 298

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 2

ER -