Latent Variable Model for Joint Analysis of Multiple Repeated Measures and Bivariate Event Times

Wenzheng Huang, Scott Zeger, James C. Anthony, Elizabeth Garrett

Research output: Contribution to journalArticle

Abstract

This article presents a novel approach to analyzing a complex dataset from a prevention trial, where outcomes comprise multiple repeated mental health items and times to initiation of alcohol and tobacco use. The dataset has a nonnegligible portion of missing values and interval or left censored events. The substantive interest of the trial suggests a psychiatric distress latent variable that is reflected in the mental health items and potentially affects initiation of alcohol and tobacco use. We describe the data with a combination of three types of component model: a marginal model for the longitudinal latent process for psychiatric distress given study interventions and covariates: logistic regression models for the repeated mental health items given the latent process; and hazard models for times to initiation of alcohol and tobacco use given the latent process, study interventions, and covariates. To aid in fitting these models simultaneously, we use automatic differentiation to find the first two derivatives of the total log-likelihood function, thus speeding up convergence relative to the regular expectation-maximization algorithm with a direct calculation of valid variance estimates.

Original languageEnglish (US)
Pages (from-to)906-914
Number of pages9
JournalJournal of the American Statistical Association
Volume96
Issue number455
StatePublished - Sep 2001

Fingerprint

Latent Process
Latent Variable Models
Repeated Measures
Tobacco
Alcohol
Health
Covariates
Multiple Outcomes
Marginal Model
Automatic Differentiation
Hazard Models
Logistic Regression Model
Model Fitting
Missing Values
Expectation-maximization Algorithm
Latent Variables
Component Model
Likelihood Function
Valid
Derivative

Keywords

  • Automatic differentiation
  • Failure time analysis
  • Latent variable
  • Longitudinal data analysis
  • Missing data
  • Prevention research

ASJC Scopus subject areas

  • Mathematics(all)
  • Statistics and Probability

Cite this

Latent Variable Model for Joint Analysis of Multiple Repeated Measures and Bivariate Event Times. / Huang, Wenzheng; Zeger, Scott; Anthony, James C.; Garrett, Elizabeth.

In: Journal of the American Statistical Association, Vol. 96, No. 455, 09.2001, p. 906-914.

Research output: Contribution to journalArticle

@article{3b8439336bba4cd8ad05755f9f511bc0,
title = "Latent Variable Model for Joint Analysis of Multiple Repeated Measures and Bivariate Event Times",
abstract = "This article presents a novel approach to analyzing a complex dataset from a prevention trial, where outcomes comprise multiple repeated mental health items and times to initiation of alcohol and tobacco use. The dataset has a nonnegligible portion of missing values and interval or left censored events. The substantive interest of the trial suggests a psychiatric distress latent variable that is reflected in the mental health items and potentially affects initiation of alcohol and tobacco use. We describe the data with a combination of three types of component model: a marginal model for the longitudinal latent process for psychiatric distress given study interventions and covariates: logistic regression models for the repeated mental health items given the latent process; and hazard models for times to initiation of alcohol and tobacco use given the latent process, study interventions, and covariates. To aid in fitting these models simultaneously, we use automatic differentiation to find the first two derivatives of the total log-likelihood function, thus speeding up convergence relative to the regular expectation-maximization algorithm with a direct calculation of valid variance estimates.",
keywords = "Automatic differentiation, Failure time analysis, Latent variable, Longitudinal data analysis, Missing data, Prevention research",
author = "Wenzheng Huang and Scott Zeger and Anthony, {James C.} and Elizabeth Garrett",
year = "2001",
month = "9",
language = "English (US)",
volume = "96",
pages = "906--914",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "455",

}

TY - JOUR

T1 - Latent Variable Model for Joint Analysis of Multiple Repeated Measures and Bivariate Event Times

AU - Huang, Wenzheng

AU - Zeger, Scott

AU - Anthony, James C.

AU - Garrett, Elizabeth

PY - 2001/9

Y1 - 2001/9

N2 - This article presents a novel approach to analyzing a complex dataset from a prevention trial, where outcomes comprise multiple repeated mental health items and times to initiation of alcohol and tobacco use. The dataset has a nonnegligible portion of missing values and interval or left censored events. The substantive interest of the trial suggests a psychiatric distress latent variable that is reflected in the mental health items and potentially affects initiation of alcohol and tobacco use. We describe the data with a combination of three types of component model: a marginal model for the longitudinal latent process for psychiatric distress given study interventions and covariates: logistic regression models for the repeated mental health items given the latent process; and hazard models for times to initiation of alcohol and tobacco use given the latent process, study interventions, and covariates. To aid in fitting these models simultaneously, we use automatic differentiation to find the first two derivatives of the total log-likelihood function, thus speeding up convergence relative to the regular expectation-maximization algorithm with a direct calculation of valid variance estimates.

AB - This article presents a novel approach to analyzing a complex dataset from a prevention trial, where outcomes comprise multiple repeated mental health items and times to initiation of alcohol and tobacco use. The dataset has a nonnegligible portion of missing values and interval or left censored events. The substantive interest of the trial suggests a psychiatric distress latent variable that is reflected in the mental health items and potentially affects initiation of alcohol and tobacco use. We describe the data with a combination of three types of component model: a marginal model for the longitudinal latent process for psychiatric distress given study interventions and covariates: logistic regression models for the repeated mental health items given the latent process; and hazard models for times to initiation of alcohol and tobacco use given the latent process, study interventions, and covariates. To aid in fitting these models simultaneously, we use automatic differentiation to find the first two derivatives of the total log-likelihood function, thus speeding up convergence relative to the regular expectation-maximization algorithm with a direct calculation of valid variance estimates.

KW - Automatic differentiation

KW - Failure time analysis

KW - Latent variable

KW - Longitudinal data analysis

KW - Missing data

KW - Prevention research

UR - http://www.scopus.com/inward/record.url?scp=0442280678&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0442280678&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0442280678

VL - 96

SP - 906

EP - 914

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 455

ER -