Handling Missing Data in the Modeling of Intensive Longitudinal Data

Linying Ji, Sy Miin Chow, Alice C. Schermerhorn, Nicholas C. Jacobson, E. Mark Cummings

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Myriad approaches for handling missing data exist in the literature. However, few studies have investigated the tenability and utility of these approaches when used with intensive longitudinal data. In this study, we compare and illustrate two multiple imputation (MI) approaches for coping with missingness in fitting multivariate time-series models under different missing data mechanisms. They include a full MI approach, in which all dependent variables and covariates are imputed simultaneously, and a partial MI approach, in which missing covariates are imputed with MI, whereas missingness in the dependent variables is handled via full information maximum likelihood estimation. We found that under correctly specified models, partial MI produces the best overall estimation results. We discuss the strengths and limitations of the two MI approaches, and demonstrate their use with an empirical data set in which children’s influences on parental conflicts are modeled as covariates over the course of 15 days (Schermerhorn, Chow, & Cummings, 2010).

Original languageEnglish (US)
Pages (from-to)715-736
Number of pages22
JournalStructural Equation Modeling
Volume25
Issue number5
DOIs
StatePublished - Sep 3 2018

Fingerprint

Multiple Imputation
Data handling
Longitudinal Data
Missing Data
Maximum likelihood estimation
Modeling
Time series
Covariates
Missing Data Mechanism
Missing Covariates
time series
Partial
coping
Multivariate Time Series
Dependent
Multivariate Models
Time Series Models
Maximum Likelihood Estimation
Longitudinal data
Multiple imputation

All Science Journal Classification (ASJC) codes

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

Cite this

Ji, Linying ; Chow, Sy Miin ; Schermerhorn, Alice C. ; Jacobson, Nicholas C. ; Cummings, E. Mark. / Handling Missing Data in the Modeling of Intensive Longitudinal Data. In: Structural Equation Modeling. 2018 ; Vol. 25, No. 5. pp. 715-736.
@article{f0926c84072c4fda93ad1ff010d840fc,
title = "Handling Missing Data in the Modeling of Intensive Longitudinal Data",
abstract = "Myriad approaches for handling missing data exist in the literature. However, few studies have investigated the tenability and utility of these approaches when used with intensive longitudinal data. In this study, we compare and illustrate two multiple imputation (MI) approaches for coping with missingness in fitting multivariate time-series models under different missing data mechanisms. They include a full MI approach, in which all dependent variables and covariates are imputed simultaneously, and a partial MI approach, in which missing covariates are imputed with MI, whereas missingness in the dependent variables is handled via full information maximum likelihood estimation. We found that under correctly specified models, partial MI produces the best overall estimation results. We discuss the strengths and limitations of the two MI approaches, and demonstrate their use with an empirical data set in which children’s influences on parental conflicts are modeled as covariates over the course of 15 days (Schermerhorn, Chow, & Cummings, 2010).",
author = "Linying Ji and Chow, {Sy Miin} and Schermerhorn, {Alice C.} and Jacobson, {Nicholas C.} and Cummings, {E. Mark}",
year = "2018",
month = "9",
day = "3",
doi = "10.1080/10705511.2017.1417046",
language = "English (US)",
volume = "25",
pages = "715--736",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "Psychology Press Ltd",
number = "5",

}

Handling Missing Data in the Modeling of Intensive Longitudinal Data. / Ji, Linying; Chow, Sy Miin; Schermerhorn, Alice C.; Jacobson, Nicholas C.; Cummings, E. Mark.

In: Structural Equation Modeling, Vol. 25, No. 5, 03.09.2018, p. 715-736.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Handling Missing Data in the Modeling of Intensive Longitudinal Data

AU - Ji, Linying

AU - Chow, Sy Miin

AU - Schermerhorn, Alice C.

AU - Jacobson, Nicholas C.

AU - Cummings, E. Mark

PY - 2018/9/3

Y1 - 2018/9/3

N2 - Myriad approaches for handling missing data exist in the literature. However, few studies have investigated the tenability and utility of these approaches when used with intensive longitudinal data. In this study, we compare and illustrate two multiple imputation (MI) approaches for coping with missingness in fitting multivariate time-series models under different missing data mechanisms. They include a full MI approach, in which all dependent variables and covariates are imputed simultaneously, and a partial MI approach, in which missing covariates are imputed with MI, whereas missingness in the dependent variables is handled via full information maximum likelihood estimation. We found that under correctly specified models, partial MI produces the best overall estimation results. We discuss the strengths and limitations of the two MI approaches, and demonstrate their use with an empirical data set in which children’s influences on parental conflicts are modeled as covariates over the course of 15 days (Schermerhorn, Chow, & Cummings, 2010).

AB - Myriad approaches for handling missing data exist in the literature. However, few studies have investigated the tenability and utility of these approaches when used with intensive longitudinal data. In this study, we compare and illustrate two multiple imputation (MI) approaches for coping with missingness in fitting multivariate time-series models under different missing data mechanisms. They include a full MI approach, in which all dependent variables and covariates are imputed simultaneously, and a partial MI approach, in which missing covariates are imputed with MI, whereas missingness in the dependent variables is handled via full information maximum likelihood estimation. We found that under correctly specified models, partial MI produces the best overall estimation results. We discuss the strengths and limitations of the two MI approaches, and demonstrate their use with an empirical data set in which children’s influences on parental conflicts are modeled as covariates over the course of 15 days (Schermerhorn, Chow, & Cummings, 2010).

UR - http://www.scopus.com/inward/record.url?scp=85041540173&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85041540173&partnerID=8YFLogxK

U2 - 10.1080/10705511.2017.1417046

DO - 10.1080/10705511.2017.1417046

M3 - Article

AN - SCOPUS:85041540173

VL - 25

SP - 715

EP - 736

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 5

ER -