scispace - formally typeset
Open AccessJournal ArticleDOI

Multilevel Latent Class Models

Jeroen K. Vermunt
- 01 Jan 2003 - 
- Vol. 33, Iss: 1, pp 213-239
Reads0
Chats0
TLDR
In this article, Parametric and non-parametric random-coefficient latent class models are proposed to modify the assumption that observations are independent, which can be used for the analysis of data collected with complex sampling designs, data with a multilevel structure, and multiple-group data for more than a few groups.
Abstract
The latent class (LC) models that have been developed so far assume that observations are independent. Parametric and non-parametric random-coefficient LC models are proposed here, which will make it possible to modify this assumption. For example, the models can be used for the analysis of data collected with complex sampling designs, data with a multilevel structure, and multiple-group data for more than a few groups. An adapted EM algorithm is presented that makes maximum-likelihood estimation feasible. The new model is illustrated with examples from organizational, educational, and cross-national comparative research.

read more

Content maybe subject to copyright    Report

Tilburg University
Multilevel Latent Class Models
Vermunt, J.K.
Published in:
Sociological Methodology
Publication date:
2003
Document Version
Peer reviewed version
Link to publication in Tilburg University Research Portal
Citation for published version (APA):
Vermunt, J. K. (2003). Multilevel Latent Class Models.
Sociological Methodology
,
33
(1), 213-239.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Download date: 10. aug.. 2022

Pxowlohyho Odwhqw Fodvv Prghov
Mhurhq N1 Yhupxqw
Ghsduwphqw ri Phwkrgrorj| dqg Vwdwlvwlfv/ Wloexuj Xqlyhuvlw|
Dgguhvv ri fruuhvsrqghqfh=
Mhurhq N1 Yhupxqw
Ghsduwphqw ri Phwkrgrorj| dqg Vwdwlvwlfv
Wloexuj Xqlyhuvlw|
S1R1Er{ <3486
8333 OH Wloexuj
Wkh Qhwkhuodqgv
H0pdlo= M1N1YhupxqwCxyw1qo
Dffhswhg e| Vrflrorjlfdo Phwkrgrorj| +Qryhpehu 7/ 5335,

Pxowlohyho Odwhqw Fodvv Prghov
Devwudfw
Odwhqw fodvv +OF, prghov ghyhorshg vr idu dvvxph wkdw revhuydwlrqv duh lqghshqghqw1 Sdud0
phwulf dqg qrqsdudphwulf udqgrp0frh!flhqw OF prghov duh sursrvhg wkdw pdnh lw srvvleoh wr
uhod{ wklv dvvxpswlrq1 Wkh prghov fdq/ iru h{dpsoh/ eh xvhg iru wkh dqdo|vlv ri gdwd froohfwhg
zlwk frpsoh{ vdpsolqj ghvljqv/ gdwd zlwk d pxowlohyho vwuxfwxuh/ dqg pxowlsoh0jurxs gdwd iru
pruh wkdq d ihz jurxsv1 Dq dgdswhg HP dojrulwkp lv suhvhqwhg wkdw pdnhv pd{lpxp olnh0
olkrrg hvwlpdwlrq ihdvleoh1 Wkh qhz prgho lv looxvwudwhg zlwk h{dpsohv iurp rujdql}dwlrqdo/
hgxfdwlrqdo/ dqg furvv0qdwlrqdo frpsdudwlyh uhvhdufk1
4

Pxowlohyho Odwhqw Fodvv Prghov
4 Lqwurgxfwlrq
Lq wkh sdvw ghfdgh/ odwhqw fodvv +OF, dqdo|vlv +Od}duvihog/ 4<83> Jrrgpdq/ 4<:7, kdv ehfrph
d pruh zlgho| xvhg whfkqltxh lq vrfldo vflhqfhv uhvhdufk1 Rqh ri lwv dssolfdwlrqv lv foxvwhulqj
ru frqvwuxfwlqj w|srorjlhv zlwk revhuyhg fdwhjrulfdo yduldeohv1 Dqrwkhu uhodwhg dssolfdwlrq lv
ghdolqj zlwk phdvxuhphqw huuru lq qrplqdo dqg ruglqdo lqglfdwruv1 Dq lpsruwdqw olplwdwlrq ri
wkh OF prghov ghyhorshg vr idu lv/ krzhyhu/ wkdw wkh| dvvxph wkdw revhuydwlrqv duh lqghshqghqw/
dq dvvxpswlrq wkdw lv riwhq ylrodwhg1 Wklv sdshu suhvhqwv d udqgrp0frh!flhqwv ru pxowlohyho
OF prgho wkdw pdnhv lw srvvleoh wr uhod{ wklv dvvxpswlrq1
Udqgrp0frh!flhqwv prghov fdq eh xvhg wr ghdo zlwk ydulrxv w|shv ri ghshqghqw revhuydwlrqv
+Djuhvwl hw do1/ 5333,1 Dq h{dpsoh lv ghshqghqw revhuydwlrqv lq gdwd vhwv froohfwhg e| wzr0vwdjh
foxvwhu vdpsolqj ru orqjlwxglqdo ghvljqv1 D pruh wkhru| edvhg xvh ri udqgrp0frh!flhqwv prghov
wkdw lv srsxodu lq hogv vxfk dv hgxfdwlrqdo dqg rujdql}dwlrqdo uhvhdufk lv riwhq uhihuuhg wr dv
pxowlohyho ru klhudufklfdo prgholqj/ d phwkrg wkdw lv lqwhqghg wr glvhqwdqjoh jurxs0ohyho iurp
lqglylgxdo0ohyho hhfwv +Eu|n dqg Udxghqexvk/ 4<<5> Jrogvwhlq/ 4<<8> Vqlmghuv dqg Ervnhu/
4<<<,1 Dqrwkhu lqwhuhvwlqj dssolfdwlrq ri wkhvh phwkrgv lv lq wkh frqwh{w ri pxowlsoh0jurxs
dqdo|vlv/ vxfk dv lq furvv0qdwlrqdo frpsdudwlyh uhvhdufk edvhg rq gdwd iurp d odujh qxpehu ri
frxqwulhv +vhh/ iru h{dpsoh/ Zrqj dqg Pdvrq/ 4<;8,1
Lq d vwdqgdug OF prgho/ lw lv dvvxphg wkdw wkh prgho sdudphwhuv duh wkh vdph iru doo
shuvrqv +ohyho04 xqlwv,1 Wkh edvlf lghd ri d pxowlohyho OF prgho lv wkdw vrph ri wkh prgho
sdudphwhuv duh doorzhg wr glhuv dfurvv jurxsv/ foxvwhuv/ ru ohyho05 xqlwv1 Iru h{dpsoh/ wkh
suredelolw| ri ehorqjlqj wr d fhuwdlq odwhqw fodvv pd| glhu dfurvv rujdql}dwlrqv ru frxqwulhv1
Vxfk glhuhqfhv fdq eh prghoohg e| lqfoxglqj jurxs gxpplhv lq wkh prgho/ dv lv grqh lq
5

pxowlsoh0jurxs OF dqdo|vlv +Forjj dqg Jrrgpdq/ 4<;7,/ zklfk dprxqwv wr xvlqj zkdw lv
fdoohg d {hg0hhfwv dssurdfk1 Dowhuqdwlyho|/ lq d udqgrp0hhfwv dssurdfk/ wkh jurxs0vshflf
frh!flhqwv duh dvvxphg wr frph iurp d sduwlfxodu glvwulexwlrq/ zkrvh sdudphwhuv vkrxog eh
hvwlpdwhg1 Ghshqglqj rq zkhwkhu wkh irup ri wkh pl{lqj glvwulexwlrq lv vshflhg ru qrw/ hlwkhu
d sdudphwulf ru d qrqsdudphwulf udqgrp0hhfwv dssurdfk lv rewdlqhg1
Wkh sursrvhg pxowlohyho OF prgho lv vlplodu wr d udqgrp0frh!flhqwv orjlvwlf uhjuhvvlrq
prgho +Zrqj dqg Pdvrq/ 4<;8> Khghnhu dqg Jleerqv/ 4<<9> Khghnhu/ 4<<<> Djuhvwl hw do1/
5333,1 D glhuhqfh lv wkdw wkh ghshqghqw yduldeoh lv qrw gluhfwo| revhuyhg/ exw d odwhqw yduldeoh
zlwk vhyhudo revhuyhg lqglfdwruv1 Wkh prgho fdq/ wkhuhiruh/ eh vhhq dv dq h{whqvlrq ri d udqgrp0
frh!flhqwv orjlvwlf uhjuhvvlrq prgho lq zklfk wkhuh lv phdvxuhphqw huuru lq wkh ghshqghqw
yduldeoh1 Lw lv zhoo0nqrzq wkdw OF prghov fdq eh xvhg wr frpelqh wkh lqirupdwlrq frqwdlqhg
lq pxowlsoh rxwfrph yduldeohv +Edqghhq0Urfkh hw do1/ 4<<:,1
Wkh prgho lv dovr vlplodu wr wkh pxowlohyho lwhp uhvsrqvh wkhru| +LUW, prgho wkdw zdv
uhfhqwo| sursrvhg e| Ir{ dqg Jodv +5334,1 D frqfhswxdo glhuhqfh lv/ krzhyhu/ wkdw lq LUW
prghov wkh xqghuo|lqj odwhqw yduldeohv duh dvvxphg wr eh frqwlqxrxv lqvwhdg ri glvfuhwh1 Eh0
fdxvh ri wkh vlplodulw| ehwzhhq OF dqg LUW prghov +vhh/ iru h{dpsoh/ Khlqhq/ 4<<9,/ lw lv
qrw vxusulvlqj wkdw uhvwulfwhg pxowlohyho OF prghov fdq eh xvhg wr dssur{lpdwh pxowlohyho LUW
prghov1
Wkh lghd ri lqwurgxflqj udqgrp hhfwv lq OF dqdo|vlv lv qrw qhz= vhh/ iru h{dpsoh/ Tx/ Wdq/
dqg Nxwqhu +4<<9,/ dqg Ohqn dqg GhVduer +5333,1 Wkh prghov sursrvhg e| wkhvh dxwkruv duh/
krzhyhu/ qrw pxowlohyho prghov/ dqg wkhuhiruh frqfhswxdoo| dqg pdwkhpdwlfdoo| yhu| glhuhqw
iurp wkh prghov ghvfulehg lq wklv sdshu1 Dowkrxjk wkhuh lv qr rswlrq iru frpelqlqj OF vwuxf0
wxuhv zlwk udqgrp hhfwv lq wkh fxuuhqw yhuvlrq ri wkh JOODPP surjudp ri Udeh0Khvnhwk/
Slfnohv/ dqg Vnurqgdo +5334,/ wkh prgho L sursrvh wv yhu| qdwxudoo| lqwr wkh jhqhudo pxowlohyho
6

Citations
More filters
Journal ArticleDOI

Latent Class Modeling with Covariates: Two Improved Three-Step Approaches

TL;DR: The correction method of Bolck, Croon, and Hagenaars is extended by showing that it involves maximizing a weighted log-likelihood function for clustered data, which makes it possible to apply the method not only with categorical but also with continuous explanatory variables.
Journal ArticleDOI

A Longitudinal Analysis of Customer Satisfaction and Share of Wallet: Investigating the Moderating Effect of Customer Characteristics

TL;DR: In this article, the authors investigated the relationship between customer satisfaction and share of wallet and found that the initial satisfaction level and the conditional percentile of change in satisfaction significantly correspond to changes in share of the wallet and the moderating effects of customer age, income, education, expertise and length of relationship.
Book ChapterDOI

Latent Class Models

TL;DR: A statistical model can be called a latent class (LC) or mixture model if it assumes that some of its parameters differ across unobserved subgroups, LCs, or mixture components as mentioned in this paper.
Journal ArticleDOI

Model-Based Clustering

TL;DR: A review of work to date in model-based clustering, from the famous paper by Wolfe in 1965 to work that is currently available only in preprint form, and a look ahead to the next decade or so.
References
More filters
Book

Multilevel Statistical Models

TL;DR: In this article, the authors present a general classification notation for multilevel models and a discussion of the general structure and maximum likelihood estimation for a multi-level model, as well as the adequacy of Ordinary Least Squares estimates.
Journal ArticleDOI

Marginal maximum likelihood estimation of item parameters: Application of an EM algorithm.

TL;DR: The Em procedure is shown to apply to general item-response models lacking simple sufficient statistics for ability, including models with more than one latent dimension, when computing procedures based on an EM algorithm are used.
Journal ArticleDOI

Multilevel Statistical Models

TL;DR: Applied Longitudinal Data Analysis provides readers with a solid, thorough, and relatively accurate understanding of concepts and procedures, and is not a comprehensive treatment of all methods for longitudinal data.
Journal ArticleDOI

Hidden Markov models for speech recognition

TL;DR: The role of statistical methods in this powerful technology as applied to speech recognition is addressed and a range of theoretical and practical issues that are as yet unsolved in terms of their importance and their effect on performance for different system implementations are discussed.
Frequently Asked Questions (1)
Q1. What are the contributions in this paper?

If you believe that this document breaches copyright please contact us providing details, and the authors will remove access to the work immediately and investigate your claim.