Power and type I error of local fit statistics in multilevel latent class analysis

Research output: Contribution to journalArticleScientificpeer-review

12 Downloads (Pure)

Abstract

In the social and behavioral sciences, variables are often categorical and people are often nested in groups. Models for such data, such as multilevel logistic regression or the multilevel latent class model, should account for not only the categorical nature of the variables, but also the nested structure of the persons. To assess whether the model accomplishes this goal adequately, local fit measures for multilevel categorical data were recently introduced by Nagelkerke, Oberski, and Vermunt (2015). The BVR-group evaluates the variable–group fit, and the BVR-pair evaluates the person–person fit within groups. In this article, we evaluate the performance of these 2 measures for the multilevel latent class model (Vermunt, 2003). An extensive simulation study indicates that whenever multilevel latent class modeling itself is viable, Type I error is controlled and power is adequate for both fit statistics. Thus, the BVR-group and BVR-pair are useful measures to locate important sources of misfit in multilevel latent class analysis.
Original languageEnglish
Pages (from-to)216-229
JournalStructural Equation Modeling
Volume24
Issue number2
DOIs
Publication statusPublished - 2017

Fingerprint

Latent Class Analysis
Multilevel Analysis
Type I error
statistics
Statistics
Latent Class Model
Multilevel Models
Categorical
Evaluate
Group
Latent Class
Nominal or categorical data
behavioral science
Logistic Regression
Logistics
Person
social science
logistics
Simulation Study
regression

Keywords

  • bivariate residual
  • latent class analysis
  • local fit
  • goodness-of-fit
  • multilevel

Cite this

@article{cfe7c1d0786146b29e2f276118e71e90,
title = "Power and type I error of local fit statistics in multilevel latent class analysis",
abstract = "In the social and behavioral sciences, variables are often categorical and people are often nested in groups. Models for such data, such as multilevel logistic regression or the multilevel latent class model, should account for not only the categorical nature of the variables, but also the nested structure of the persons. To assess whether the model accomplishes this goal adequately, local fit measures for multilevel categorical data were recently introduced by Nagelkerke, Oberski, and Vermunt (2015). The BVR-group evaluates the variable–group fit, and the BVR-pair evaluates the person–person fit within groups. In this article, we evaluate the performance of these 2 measures for the multilevel latent class model (Vermunt, 2003). An extensive simulation study indicates that whenever multilevel latent class modeling itself is viable, Type I error is controlled and power is adequate for both fit statistics. Thus, the BVR-group and BVR-pair are useful measures to locate important sources of misfit in multilevel latent class analysis.",
keywords = "bivariate residual, latent class analysis, local fit, goodness-of-fit, multilevel",
author = "E. Nagelkerke and D.L. Oberski and J.K. Vermunt",
year = "2017",
doi = "10.1080/10705511.2016.1250639",
language = "English",
volume = "24",
pages = "216--229",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "ROUTLEDGE JOURNALS, TAYLOR & FRANCIS LTD",
number = "2",

}

Power and type I error of local fit statistics in multilevel latent class analysis. / Nagelkerke, E. ; Oberski, D.L.; Vermunt, J.K.

In: Structural Equation Modeling, Vol. 24, No. 2, 2017, p. 216-229.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Power and type I error of local fit statistics in multilevel latent class analysis

AU - Nagelkerke, E.

AU - Oberski, D.L.

AU - Vermunt, J.K.

PY - 2017

Y1 - 2017

N2 - In the social and behavioral sciences, variables are often categorical and people are often nested in groups. Models for such data, such as multilevel logistic regression or the multilevel latent class model, should account for not only the categorical nature of the variables, but also the nested structure of the persons. To assess whether the model accomplishes this goal adequately, local fit measures for multilevel categorical data were recently introduced by Nagelkerke, Oberski, and Vermunt (2015). The BVR-group evaluates the variable–group fit, and the BVR-pair evaluates the person–person fit within groups. In this article, we evaluate the performance of these 2 measures for the multilevel latent class model (Vermunt, 2003). An extensive simulation study indicates that whenever multilevel latent class modeling itself is viable, Type I error is controlled and power is adequate for both fit statistics. Thus, the BVR-group and BVR-pair are useful measures to locate important sources of misfit in multilevel latent class analysis.

AB - In the social and behavioral sciences, variables are often categorical and people are often nested in groups. Models for such data, such as multilevel logistic regression or the multilevel latent class model, should account for not only the categorical nature of the variables, but also the nested structure of the persons. To assess whether the model accomplishes this goal adequately, local fit measures for multilevel categorical data were recently introduced by Nagelkerke, Oberski, and Vermunt (2015). The BVR-group evaluates the variable–group fit, and the BVR-pair evaluates the person–person fit within groups. In this article, we evaluate the performance of these 2 measures for the multilevel latent class model (Vermunt, 2003). An extensive simulation study indicates that whenever multilevel latent class modeling itself is viable, Type I error is controlled and power is adequate for both fit statistics. Thus, the BVR-group and BVR-pair are useful measures to locate important sources of misfit in multilevel latent class analysis.

KW - bivariate residual

KW - latent class analysis

KW - local fit

KW - goodness-of-fit

KW - multilevel

U2 - 10.1080/10705511.2016.1250639

DO - 10.1080/10705511.2016.1250639

M3 - Article

VL - 24

SP - 216

EP - 229

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 2

ER -