Modeling predictors of latent classes in regression mixture models

M. Kim, J.K. Vermunt, Z. Bakk, T. Jaki, M.L. Van Horn

Research output: Contribution to journalArticleScientificpeer-review

Abstract

The purpose of this study is to provide guidance on a process for including latent class predictors in regression mixture models. We first examine the performance of current practice for using the 1-step and 3-step approaches where the direct covariate effect on the outcome is omitted. None of the approaches show adequate estimates of model parameters. Given that Step 1 of the 3-step approach shows adequate results in class enumeration, we suggest using an alternative approach: (a) decide the number of latent classes without predictors of latent classes, and (b) bring the latent class predictors into the model with the inclusion of hypothesized direct covariate effects. Our simulations show that this approach leads to good estimates for all model parameters. The proposed approach is demonstrated by using empirical data to examine the differential effects of family resources on students’ academic achievement outcome. Implications of the study are discussed.
Keywords: finite mixture model, including covariates, latent class predictor, regression mixture model
Original languageEnglish
Pages (from-to)601-614
JournalStructural Equation Modeling
Volume23
Issue number4
DOIs
Publication statusPublished - 2016

Fingerprint

Latent Class
Mixture Model
Predictors
Regression Model
regression
Covariates
Modeling
Finite Mixture Models
Estimate
Enumeration
Guidance
academic achievement
Inclusion
Latent class
Mixture model
Model
inclusion
Resources
Students
Alternatives

Cite this

Kim, M. ; Vermunt, J.K. ; Bakk, Z. ; Jaki, T. ; Van Horn, M.L. / Modeling predictors of latent classes in regression mixture models. In: Structural Equation Modeling. 2016 ; Vol. 23, No. 4. pp. 601-614.
@article{98cace970f504bd5bbee550899eb4547,
title = "Modeling predictors of latent classes in regression mixture models",
abstract = "The purpose of this study is to provide guidance on a process for including latent class predictors in regression mixture models. We first examine the performance of current practice for using the 1-step and 3-step approaches where the direct covariate effect on the outcome is omitted. None of the approaches show adequate estimates of model parameters. Given that Step 1 of the 3-step approach shows adequate results in class enumeration, we suggest using an alternative approach: (a) decide the number of latent classes without predictors of latent classes, and (b) bring the latent class predictors into the model with the inclusion of hypothesized direct covariate effects. Our simulations show that this approach leads to good estimates for all model parameters. The proposed approach is demonstrated by using empirical data to examine the differential effects of family resources on students’ academic achievement outcome. Implications of the study are discussed.Keywords: finite mixture model, including covariates, latent class predictor, regression mixture model",
author = "M. Kim and J.K. Vermunt and Z. Bakk and T. Jaki and {Van Horn}, M.L.",
year = "2016",
doi = "10.1080/10705511.2016.1158655",
language = "English",
volume = "23",
pages = "601--614",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "ROUTLEDGE JOURNALS, TAYLOR & FRANCIS LTD",
number = "4",

}

Modeling predictors of latent classes in regression mixture models. / Kim, M.; Vermunt, J.K.; Bakk, Z.; Jaki, T.; Van Horn, M.L.

In: Structural Equation Modeling, Vol. 23, No. 4, 2016, p. 601-614.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Modeling predictors of latent classes in regression mixture models

AU - Kim, M.

AU - Vermunt, J.K.

AU - Bakk, Z.

AU - Jaki, T.

AU - Van Horn, M.L.

PY - 2016

Y1 - 2016

N2 - The purpose of this study is to provide guidance on a process for including latent class predictors in regression mixture models. We first examine the performance of current practice for using the 1-step and 3-step approaches where the direct covariate effect on the outcome is omitted. None of the approaches show adequate estimates of model parameters. Given that Step 1 of the 3-step approach shows adequate results in class enumeration, we suggest using an alternative approach: (a) decide the number of latent classes without predictors of latent classes, and (b) bring the latent class predictors into the model with the inclusion of hypothesized direct covariate effects. Our simulations show that this approach leads to good estimates for all model parameters. The proposed approach is demonstrated by using empirical data to examine the differential effects of family resources on students’ academic achievement outcome. Implications of the study are discussed.Keywords: finite mixture model, including covariates, latent class predictor, regression mixture model

AB - The purpose of this study is to provide guidance on a process for including latent class predictors in regression mixture models. We first examine the performance of current practice for using the 1-step and 3-step approaches where the direct covariate effect on the outcome is omitted. None of the approaches show adequate estimates of model parameters. Given that Step 1 of the 3-step approach shows adequate results in class enumeration, we suggest using an alternative approach: (a) decide the number of latent classes without predictors of latent classes, and (b) bring the latent class predictors into the model with the inclusion of hypothesized direct covariate effects. Our simulations show that this approach leads to good estimates for all model parameters. The proposed approach is demonstrated by using empirical data to examine the differential effects of family resources on students’ academic achievement outcome. Implications of the study are discussed.Keywords: finite mixture model, including covariates, latent class predictor, regression mixture model

U2 - 10.1080/10705511.2016.1158655

DO - 10.1080/10705511.2016.1158655

M3 - Article

VL - 23

SP - 601

EP - 614

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 4

ER -