An Empirical Assessment of the Sensitivity of Mixture Models to Changes in Measurement

Veronica T. Cole, Daniel J. Bauer, Andrea M. Hussong, Michael L. Giordano

Research output: Contribution to journalArticle

Abstract

This study explored the extent to which variations in self-report measures across studies can produce differences in the results obtained from mixture models. Data (N = 854) come from a laboratory analogue study of methods for creating commensurate scores of alcohol- and substance-use-related constructs when items differ systematically across participants for any given measure. Items were manipulated according to 4 conditions, corresponding to increasing levels of alteration to item stems, response options, or both. In Study 1, results from latent class analyses (LCAs) of alcohol consequences were compared across the 4 conditions, revealing differences in class enumeration and configuration. In Study 2, results from factor mixture models (FMMs) of alcohol expectancies were compared across 2 of the conditions, revealing differences in patterns and magnitude of the factor loadings and thresholds. The results suggest that even subtle differences in measurement can have substantively meaningful effects on mixture model results.

LanguageEnglish (US)
Pages159-179
Number of pages21
JournalStructural Equation Modeling
Volume24
Issue number2
DOIs
StatePublished - Mar 4 2017

Fingerprint

Mixture Model
Alcohols
alcohol
Alcohol
Latent Class
Factor Models
Enumeration
Mixture model
Analogue
Configuration

Keywords

  • factor mixture models
  • latent class analysis
  • mixture models
  • self-report

ASJC Scopus subject areas

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

Cite this

An Empirical Assessment of the Sensitivity of Mixture Models to Changes in Measurement. / Cole, Veronica T.; Bauer, Daniel J.; Hussong, Andrea M.; Giordano, Michael L.

In: Structural Equation Modeling, Vol. 24, No. 2, 04.03.2017, p. 159-179.

Research output: Contribution to journalArticle

Cole, Veronica T. ; Bauer, Daniel J. ; Hussong, Andrea M. ; Giordano, Michael L./ An Empirical Assessment of the Sensitivity of Mixture Models to Changes in Measurement. In: Structural Equation Modeling. 2017 ; Vol. 24, No. 2. pp. 159-179
@article{4fe070d5d0244f16b3fafbb8287e9592,
title = "An Empirical Assessment of the Sensitivity of Mixture Models to Changes in Measurement",
abstract = "This study explored the extent to which variations in self-report measures across studies can produce differences in the results obtained from mixture models. Data (N = 854) come from a laboratory analogue study of methods for creating commensurate scores of alcohol- and substance-use-related constructs when items differ systematically across participants for any given measure. Items were manipulated according to 4 conditions, corresponding to increasing levels of alteration to item stems, response options, or both. In Study 1, results from latent class analyses (LCAs) of alcohol consequences were compared across the 4 conditions, revealing differences in class enumeration and configuration. In Study 2, results from factor mixture models (FMMs) of alcohol expectancies were compared across 2 of the conditions, revealing differences in patterns and magnitude of the factor loadings and thresholds. The results suggest that even subtle differences in measurement can have substantively meaningful effects on mixture model results.",
keywords = "factor mixture models, latent class analysis, mixture models, self-report",
author = "Cole, {Veronica T.} and Bauer, {Daniel J.} and Hussong, {Andrea M.} and Giordano, {Michael L.}",
year = "2017",
month = "3",
day = "4",
doi = "10.1080/10705511.2016.1257354",
language = "English (US)",
volume = "24",
pages = "159--179",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "Psychology Press Ltd",
number = "2",

}

TY - JOUR

T1 - An Empirical Assessment of the Sensitivity of Mixture Models to Changes in Measurement

AU - Cole,Veronica T.

AU - Bauer,Daniel J.

AU - Hussong,Andrea M.

AU - Giordano,Michael L.

PY - 2017/3/4

Y1 - 2017/3/4

N2 - This study explored the extent to which variations in self-report measures across studies can produce differences in the results obtained from mixture models. Data (N = 854) come from a laboratory analogue study of methods for creating commensurate scores of alcohol- and substance-use-related constructs when items differ systematically across participants for any given measure. Items were manipulated according to 4 conditions, corresponding to increasing levels of alteration to item stems, response options, or both. In Study 1, results from latent class analyses (LCAs) of alcohol consequences were compared across the 4 conditions, revealing differences in class enumeration and configuration. In Study 2, results from factor mixture models (FMMs) of alcohol expectancies were compared across 2 of the conditions, revealing differences in patterns and magnitude of the factor loadings and thresholds. The results suggest that even subtle differences in measurement can have substantively meaningful effects on mixture model results.

AB - This study explored the extent to which variations in self-report measures across studies can produce differences in the results obtained from mixture models. Data (N = 854) come from a laboratory analogue study of methods for creating commensurate scores of alcohol- and substance-use-related constructs when items differ systematically across participants for any given measure. Items were manipulated according to 4 conditions, corresponding to increasing levels of alteration to item stems, response options, or both. In Study 1, results from latent class analyses (LCAs) of alcohol consequences were compared across the 4 conditions, revealing differences in class enumeration and configuration. In Study 2, results from factor mixture models (FMMs) of alcohol expectancies were compared across 2 of the conditions, revealing differences in patterns and magnitude of the factor loadings and thresholds. The results suggest that even subtle differences in measurement can have substantively meaningful effects on mixture model results.

KW - factor mixture models

KW - latent class analysis

KW - mixture models

KW - self-report

UR - http://www.scopus.com/inward/record.url?scp=85010669537&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85010669537&partnerID=8YFLogxK

U2 - 10.1080/10705511.2016.1257354

DO - 10.1080/10705511.2016.1257354

M3 - Article

VL - 24

SP - 159

EP - 179

JO - Structural Equation Modeling

T2 - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 2

ER -