Eliminating Bias in Classify-Analyze Approaches for Latent Class Analysis

Bethany Cara Bray, Stephanie Trea Lanza, Xianming Tan

Research output: Contribution to journalArticle

72 Citations (Scopus)

Abstract

Despite recent methodological advances in latent class analysis (LCA) and a rapid increase in its application in behavioral research, complex research questions that include latent class variables often must be addressed by classifying individuals into latent classes and treating class membership as known in a subsequent analysis. Traditional approaches to classifying individuals based on posterior probabilities are known to produce attenuated estimates in the analytic model. We propose the use of a more inclusive LCA to generate posterior probabilities; this LCA includes additional variables present in the analytic model. A motivating empirical demonstration is presented, followed by a simulation study to assess the performance of the proposed strategy. Results show that with sufficient measurement quality or sample size, the proposed strategy reduces or eliminates bias.

Original languageEnglish (US)
Pages (from-to)1-11
Number of pages11
JournalStructural Equation Modeling
Volume22
Issue number1
DOIs
StatePublished - Jan 2 2015

Fingerprint

Latent Class Analysis
Latent Class
Behavioral research
Classify
Posterior Probability
trend
Demonstrations
class membership
behavioral research
Sample Size
Eliminate
Simulation Study
Sufficient
Model
Estimate
simulation
Latent class analysis
present
performance
Strategy

All Science Journal Classification (ASJC) codes

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

Cite this

@article{eee06082e30048d9a4229a68cc7c9fd3,
title = "Eliminating Bias in Classify-Analyze Approaches for Latent Class Analysis",
abstract = "Despite recent methodological advances in latent class analysis (LCA) and a rapid increase in its application in behavioral research, complex research questions that include latent class variables often must be addressed by classifying individuals into latent classes and treating class membership as known in a subsequent analysis. Traditional approaches to classifying individuals based on posterior probabilities are known to produce attenuated estimates in the analytic model. We propose the use of a more inclusive LCA to generate posterior probabilities; this LCA includes additional variables present in the analytic model. A motivating empirical demonstration is presented, followed by a simulation study to assess the performance of the proposed strategy. Results show that with sufficient measurement quality or sample size, the proposed strategy reduces or eliminates bias.",
author = "Bray, {Bethany Cara} and Lanza, {Stephanie Trea} and Xianming Tan",
year = "2015",
month = "1",
day = "2",
doi = "10.1080/10705511.2014.935265",
language = "English (US)",
volume = "22",
pages = "1--11",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "Psychology Press Ltd",
number = "1",

}

Eliminating Bias in Classify-Analyze Approaches for Latent Class Analysis. / Bray, Bethany Cara; Lanza, Stephanie Trea; Tan, Xianming.

In: Structural Equation Modeling, Vol. 22, No. 1, 02.01.2015, p. 1-11.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Eliminating Bias in Classify-Analyze Approaches for Latent Class Analysis

AU - Bray, Bethany Cara

AU - Lanza, Stephanie Trea

AU - Tan, Xianming

PY - 2015/1/2

Y1 - 2015/1/2

N2 - Despite recent methodological advances in latent class analysis (LCA) and a rapid increase in its application in behavioral research, complex research questions that include latent class variables often must be addressed by classifying individuals into latent classes and treating class membership as known in a subsequent analysis. Traditional approaches to classifying individuals based on posterior probabilities are known to produce attenuated estimates in the analytic model. We propose the use of a more inclusive LCA to generate posterior probabilities; this LCA includes additional variables present in the analytic model. A motivating empirical demonstration is presented, followed by a simulation study to assess the performance of the proposed strategy. Results show that with sufficient measurement quality or sample size, the proposed strategy reduces or eliminates bias.

AB - Despite recent methodological advances in latent class analysis (LCA) and a rapid increase in its application in behavioral research, complex research questions that include latent class variables often must be addressed by classifying individuals into latent classes and treating class membership as known in a subsequent analysis. Traditional approaches to classifying individuals based on posterior probabilities are known to produce attenuated estimates in the analytic model. We propose the use of a more inclusive LCA to generate posterior probabilities; this LCA includes additional variables present in the analytic model. A motivating empirical demonstration is presented, followed by a simulation study to assess the performance of the proposed strategy. Results show that with sufficient measurement quality or sample size, the proposed strategy reduces or eliminates bias.

UR - http://www.scopus.com/inward/record.url?scp=84917699284&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84917699284&partnerID=8YFLogxK

U2 - 10.1080/10705511.2014.935265

DO - 10.1080/10705511.2014.935265

M3 - Article

VL - 22

SP - 1

EP - 11

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 1

ER -