A model-based cluster analysis approach to adolescent problem behaviors and young adult outcomes

Eun-Young Mun, Michael Windle, Lisa M. Schainker

Research output: Contribution to journalArticleResearchpeer-review

45 Citations (Scopus)

Abstract

Data from a community-based sample of 1,126 10th- and 11th-grade adolescents were analyzed using a model-based cluster analysis approach to empirically identify heterogeneous adolescent subpopulations from the person-oriented and pattern-oriented perspectives. The model-based cluster analysis is a new clustering procedure to investigate population heterogeneity utilizing finite mixture multivariate normal densities and accordingly to classify subpopulations using more rigorous statistical procedures for the comparison of alternative models. Four cluster groups were identified and labeled multiproblem high-risk, smoking high-risk, normative, and low-risk groups. The multiproblem high risk exhibited a constellation of high levels of problem behaviors, including delinquent and sexual behaviors, multiple illicit substance use, and depressive symptoms at age 16. They had risky temperamental attributes and lower academic functioning and educational expectations at age 15.5 and, subsequently, at age 24 completed fewer years of education, and reported lower levels of physical health and higher levels of continued involvement in substance use and abuse. The smoking high-risk group was also found to be at risk for poorer functioning in young adulthood, compared to the low-risk group. The normative and the low risk groups were, by and large, similar in their adolescent and young adult functioning. The continuity and comorbidity path from middle adolescence to young adulthood may be aided and abetted by chronic as well as episodic substance use by adolescents.

Original languageEnglish
Pages (from-to)291-318
Number of pages28
JournalDevelopment and Psychopathology
Volume20
Issue number1
DOIs
StatePublished - 1 Dec 2008

Fingerprint

Adolescent Behavior
Cluster Analysis
Young Adult
Smoking
Problem Behavior
Population Characteristics
Sexual Behavior
Health Status
Substance-Related Disorders
Comorbidity
Depression
Education

Cite this

@article{909c17440ad34efaa82de8a08e821f57,
title = "A model-based cluster analysis approach to adolescent problem behaviors and young adult outcomes",
abstract = "Data from a community-based sample of 1,126 10th- and 11th-grade adolescents were analyzed using a model-based cluster analysis approach to empirically identify heterogeneous adolescent subpopulations from the person-oriented and pattern-oriented perspectives. The model-based cluster analysis is a new clustering procedure to investigate population heterogeneity utilizing finite mixture multivariate normal densities and accordingly to classify subpopulations using more rigorous statistical procedures for the comparison of alternative models. Four cluster groups were identified and labeled multiproblem high-risk, smoking high-risk, normative, and low-risk groups. The multiproblem high risk exhibited a constellation of high levels of problem behaviors, including delinquent and sexual behaviors, multiple illicit substance use, and depressive symptoms at age 16. They had risky temperamental attributes and lower academic functioning and educational expectations at age 15.5 and, subsequently, at age 24 completed fewer years of education, and reported lower levels of physical health and higher levels of continued involvement in substance use and abuse. The smoking high-risk group was also found to be at risk for poorer functioning in young adulthood, compared to the low-risk group. The normative and the low risk groups were, by and large, similar in their adolescent and young adult functioning. The continuity and comorbidity path from middle adolescence to young adulthood may be aided and abetted by chronic as well as episodic substance use by adolescents.",
author = "Eun-Young Mun and Michael Windle and Schainker, {Lisa M.}",
year = "2008",
month = "12",
day = "1",
doi = "10.1017/S095457940800014X",
language = "English",
volume = "20",
pages = "291--318",
journal = "Development and Psychopathology",
issn = "0954-5794",
publisher = "Cambridge University Press",
number = "1",

}

A model-based cluster analysis approach to adolescent problem behaviors and young adult outcomes. / Mun, Eun-Young; Windle, Michael; Schainker, Lisa M.

In: Development and Psychopathology, Vol. 20, No. 1, 01.12.2008, p. 291-318.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - A model-based cluster analysis approach to adolescent problem behaviors and young adult outcomes

AU - Mun, Eun-Young

AU - Windle, Michael

AU - Schainker, Lisa M.

PY - 2008/12/1

Y1 - 2008/12/1

N2 - Data from a community-based sample of 1,126 10th- and 11th-grade adolescents were analyzed using a model-based cluster analysis approach to empirically identify heterogeneous adolescent subpopulations from the person-oriented and pattern-oriented perspectives. The model-based cluster analysis is a new clustering procedure to investigate population heterogeneity utilizing finite mixture multivariate normal densities and accordingly to classify subpopulations using more rigorous statistical procedures for the comparison of alternative models. Four cluster groups were identified and labeled multiproblem high-risk, smoking high-risk, normative, and low-risk groups. The multiproblem high risk exhibited a constellation of high levels of problem behaviors, including delinquent and sexual behaviors, multiple illicit substance use, and depressive symptoms at age 16. They had risky temperamental attributes and lower academic functioning and educational expectations at age 15.5 and, subsequently, at age 24 completed fewer years of education, and reported lower levels of physical health and higher levels of continued involvement in substance use and abuse. The smoking high-risk group was also found to be at risk for poorer functioning in young adulthood, compared to the low-risk group. The normative and the low risk groups were, by and large, similar in their adolescent and young adult functioning. The continuity and comorbidity path from middle adolescence to young adulthood may be aided and abetted by chronic as well as episodic substance use by adolescents.

AB - Data from a community-based sample of 1,126 10th- and 11th-grade adolescents were analyzed using a model-based cluster analysis approach to empirically identify heterogeneous adolescent subpopulations from the person-oriented and pattern-oriented perspectives. The model-based cluster analysis is a new clustering procedure to investigate population heterogeneity utilizing finite mixture multivariate normal densities and accordingly to classify subpopulations using more rigorous statistical procedures for the comparison of alternative models. Four cluster groups were identified and labeled multiproblem high-risk, smoking high-risk, normative, and low-risk groups. The multiproblem high risk exhibited a constellation of high levels of problem behaviors, including delinquent and sexual behaviors, multiple illicit substance use, and depressive symptoms at age 16. They had risky temperamental attributes and lower academic functioning and educational expectations at age 15.5 and, subsequently, at age 24 completed fewer years of education, and reported lower levels of physical health and higher levels of continued involvement in substance use and abuse. The smoking high-risk group was also found to be at risk for poorer functioning in young adulthood, compared to the low-risk group. The normative and the low risk groups were, by and large, similar in their adolescent and young adult functioning. The continuity and comorbidity path from middle adolescence to young adulthood may be aided and abetted by chronic as well as episodic substance use by adolescents.

UR - http://www.scopus.com/inward/record.url?scp=38649131997&partnerID=8YFLogxK

U2 - 10.1017/S095457940800014X

DO - 10.1017/S095457940800014X

M3 - Article

VL - 20

SP - 291

EP - 318

JO - Development and Psychopathology

JF - Development and Psychopathology

SN - 0954-5794

IS - 1

ER -