Gambling, Substance Use, and Other Problem Behaviors Among Youth: A Test of General Deviance Models

• NCJ Number: 206577
• Journal: Journal of Criminal Justice Volume: 32 Issue: 4 Dated: July/August 2004 Pages: 297-306
• Author(s): John W. Welte; Grace M. Barnes; Joseph H. Hoffman
• Editor(s): Kent B. Joscelyn
• Date Published: July 2004
• Length: 10 pages
• Publication Format: Article
• Publication Type: Research (Applied/Empirical)
• Language: English
• Country: United States of America

Annotation

Using first-order and second-order factor models, this study tested the hypothesis that gambling, alcohol misuse, other drug use, and delinquency are manifestations of a single general deviance construct.

Abstract

Gambling, alcohol misuse, other substance use, and delinquency among youth are each distinct and significant social problems; however there is a significant amount of empirical research showing strong correlations between these various forms of adolescent problem behaviors. In order to advance theory and practice, it is important to develop a better theoretical understanding of the relationships between these behaviors. This study tested two hypotheses regarding the relationships between gambling, alcohol misuse, other drug use, and delinquency: (1) the second-order model with a single general deviance construct would fit the data as well or better than a first-order model with correlations between these four distinct problem behaviors and (2) the second-order model would be superior to the first-order model for females as well as males, controlling for race and age. Data were taken from two general population household samples of adolescents and young adults from the metropolitan area of Buffalo, NY: the Adolescent Study (Study 1) and Delinquency in Young Men (Study 2). The results for the first hypothesis showed that the first-order model, with correlations between four youthful problem behavior constructs, fit better than the second-order model with a single general deviance construct. The results of the second hypothesis, which involved the comparison of first- and second-order factor models for gender-specific subgroups, showed that the first-order model with four factors fit better than the second-order model in each group examined. In summation, while problem behaviors are related, there also must be uncorrelated antecedents predicting distinct types of youthful problem behaviors. References