The Effect of Model Misspecification on Growth Mixture Model Class Enumeration

Daniel McNeish, Jeffrey R. Harring

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Multiple criteria have been proposed to aid in deciding how many latent classes to extract in growth mixture models; however, studies are just beginning to investigate the performance of these criteria under non-ideal conditions. We review these previous studies and conduct a simulation study to address the performance of fit criteria under two previously uninvestigated assumption violations: (1) linearity of covariates and (2) proper specification of the growth factor covariance matrix. Results show that, provided that estimation is carried out with a large number of random starts and final stage optimizations, BIC and the bootstrap likelihood ratio test perform exceedingly well at identifying whether the data are homogenous or whether latent classes may be present, even with misspecifications present. Results were far less favorable when software default estimation choices were selected. We discuss implications to empirical studies and speculate on the relation between estimation choices and fit criteria perform.

Original languageEnglish (US)
Pages (from-to)223-248
Number of pages26
JournalJournal of Classification
Volume34
Issue number2
DOIs
StatePublished - Jul 1 2017
Externally publishedYes

Fingerprint

imidazole mustard
Model Misspecification
Growth Model
Mixture Model
Enumeration
Latent Class
Intercellular Signaling Peptides and Proteins
Software
Growth
Multiple Criteria
Misspecification
Growth Factors
Likelihood Ratio Test
Linearity
Bootstrap
Empirical Study
Covariance matrix
Covariates
Simulation Study
Specification

Keywords

  • Bootstrapped likelihood ratio test
  • Class enumeration
  • Growth mixture model
  • Latent class
  • Misspecification

ASJC Scopus subject areas

  • Mathematics (miscellaneous)
  • Psychology (miscellaneous)
  • Statistics, Probability and Uncertainty
  • Library and Information Sciences

Cite this

The Effect of Model Misspecification on Growth Mixture Model Class Enumeration. / McNeish, Daniel; Harring, Jeffrey R.

In: Journal of Classification, Vol. 34, No. 2, 01.07.2017, p. 223-248.

Research output: Contribution to journalArticle

@article{c0b5889e219f4cadb5e87a8550ba9eff,
title = "The Effect of Model Misspecification on Growth Mixture Model Class Enumeration",
abstract = "Multiple criteria have been proposed to aid in deciding how many latent classes to extract in growth mixture models; however, studies are just beginning to investigate the performance of these criteria under non-ideal conditions. We review these previous studies and conduct a simulation study to address the performance of fit criteria under two previously uninvestigated assumption violations: (1) linearity of covariates and (2) proper specification of the growth factor covariance matrix. Results show that, provided that estimation is carried out with a large number of random starts and final stage optimizations, BIC and the bootstrap likelihood ratio test perform exceedingly well at identifying whether the data are homogenous or whether latent classes may be present, even with misspecifications present. Results were far less favorable when software default estimation choices were selected. We discuss implications to empirical studies and speculate on the relation between estimation choices and fit criteria perform.",
keywords = "Bootstrapped likelihood ratio test, Class enumeration, Growth mixture model, Latent class, Misspecification",
author = "Daniel McNeish and Harring, {Jeffrey R.}",
year = "2017",
month = "7",
day = "1",
doi = "10.1007/s00357-017-9233-y",
language = "English (US)",
volume = "34",
pages = "223--248",
journal = "Journal of Classification",
issn = "0176-4268",
publisher = "Springer New York",
number = "2",

}

TY - JOUR

T1 - The Effect of Model Misspecification on Growth Mixture Model Class Enumeration

AU - McNeish, Daniel

AU - Harring, Jeffrey R.

PY - 2017/7/1

Y1 - 2017/7/1

N2 - Multiple criteria have been proposed to aid in deciding how many latent classes to extract in growth mixture models; however, studies are just beginning to investigate the performance of these criteria under non-ideal conditions. We review these previous studies and conduct a simulation study to address the performance of fit criteria under two previously uninvestigated assumption violations: (1) linearity of covariates and (2) proper specification of the growth factor covariance matrix. Results show that, provided that estimation is carried out with a large number of random starts and final stage optimizations, BIC and the bootstrap likelihood ratio test perform exceedingly well at identifying whether the data are homogenous or whether latent classes may be present, even with misspecifications present. Results were far less favorable when software default estimation choices were selected. We discuss implications to empirical studies and speculate on the relation between estimation choices and fit criteria perform.

AB - Multiple criteria have been proposed to aid in deciding how many latent classes to extract in growth mixture models; however, studies are just beginning to investigate the performance of these criteria under non-ideal conditions. We review these previous studies and conduct a simulation study to address the performance of fit criteria under two previously uninvestigated assumption violations: (1) linearity of covariates and (2) proper specification of the growth factor covariance matrix. Results show that, provided that estimation is carried out with a large number of random starts and final stage optimizations, BIC and the bootstrap likelihood ratio test perform exceedingly well at identifying whether the data are homogenous or whether latent classes may be present, even with misspecifications present. Results were far less favorable when software default estimation choices were selected. We discuss implications to empirical studies and speculate on the relation between estimation choices and fit criteria perform.

KW - Bootstrapped likelihood ratio test

KW - Class enumeration

KW - Growth mixture model

KW - Latent class

KW - Misspecification

UR - http://www.scopus.com/inward/record.url?scp=85020648152&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020648152&partnerID=8YFLogxK

U2 - 10.1007/s00357-017-9233-y

DO - 10.1007/s00357-017-9233-y

M3 - Article

AN - SCOPUS:85020648152

VL - 34

SP - 223

EP - 248

JO - Journal of Classification

JF - Journal of Classification

SN - 0176-4268

IS - 2

ER -