Abstract
Background: Drinking problem among American adolescents is one of the major concerns to generate various issues, such as violent crime, sexual assault, family deprivation, and physical and mental health problems. Despite growing concerns about the alcohol consumption among American adolescents, not many studies have examined the correlates and predictors of their alcohol behaviors.
Purpose: The purpose of the current study was to explore the relationship between drinking behaviors, parental intervention, low self-control, and opportunity.
Methods: Based on the theoretical framework of Gottfreson and Hirschi’s (1990) theory, this study conducted Structural Equation Modeling (SEM) with a national American youth data (N = 5,030).
Results: Results in this study were consistent with previous empirical studies, supporting the general theory of crime. In addition, this study found the role of parental intervention in the theory was significant: particularly, decreases in low self-control, opportunity, and drinking behaviors. Further, the significant moderation between low self-control and opportunity on underage drinking was found in this study.
Conclusion: Those findings will provide a solution to reduce drinking problems among American adolescents.
Declaration of interest
The authors declare that they have no conflict of interest. The authors alone are responsible for the content and writing of the article.
Notes
1 Raykov (Citation1997) demonstrated that the composite reliability is analogous to coefficient alpha, addressing the consistency of the observed measures in respective latent variables (higher than .70).
2 Gottfredson and Hirschi (Citation1990) demonstrated low self-control with six characteristics: “impulsive, insensitive, physical (as opposed to verbal), risk-taking, short-sighted, and nonverbal” (p. 90).
3 Although researchers have commonly known that 3 or more items per factor use in confirmatory factor analysis (CFA), many studies showed evidences that two items could be used (e.g., Worthington & Whittaker, Citation2006; Yong & Pearce, Citation2013).
4 Kline (Citation2016) indicates that χ2 should be non-significant for the goodness-of-fit model. However, large sample sizes increase the level of χ2’s significance. Thus, other fit statistics (CFI, RMSEA, and SRMR) are required in the SEM process.