Latent Class Analysis for Multiple Discrete Latent Variables: A Study on the Association Between Violent Behavior...

Area of Science:

• Social Sciences
• Criminology
• Public Health

Background:

• Adolescent behavioral patterns, including violence and substance use, are complex and often co-occur.
• Understanding these joint patterns is crucial for effective prevention and intervention strategies.
• Existing statistical methods may not fully capture the intricate relationships between multiple behavioral variables.

Purpose of the Study:

• To introduce and validate a novel statistical approach, joint latent class analysis (JLCA), for analyzing multiple discrete latent variables.
• To systematically identify subsets of joint patterns among adolescent violent behavior and drug-using behaviors.
• To investigate the co-occurrence of specific violent and substance use behaviors in adolescent males.

Main Methods:

• Development of joint latent class analysis (JLCA) principles for identifying joint patterns of multiple discrete latent variables.
• Application of a hybrid Expectation-Maximization (EM) and Newton-Raphson algorithm for parameter inference.
• Utilizing data from the Youth Risk Behavior Surveillance System (YRBSS) 2015, including 4,957 male high-school students.

Main Results:

• The JLCA approach successfully identified distinct joint patterns across four latent variables: violent behavior, alcohol consumption, tobacco cigarette smoking, and other drug use.
• Four common patterns of violent behavior and three representative patterns for each of alcohol, tobacco, and other drug use were uncovered.
• Three common joint classes were identified, illustrating the most probable simultaneous patterns of violent and drug-using behaviors among adolescent males.

Conclusions:

• Joint latent class analysis (JLCA) is an effective method for uncovering complex behavioral patterns in adolescents.
• The findings highlight specific co-occurring patterns of violence and substance use, providing valuable insights for targeted interventions.
• JLCA offers a robust framework for future research on the interplay of multiple risk behaviors in adolescent populations.