Middle School Influences on Young Adult Socioeconomic and Criminal Justice Outcomes: A Latent Class Analyses of School Victimization and Problem Behaviors

Background: The transition to middle school marks an important change in adolescents’ social context. Student experiences and behaviors during this developmental phase can have marked implications for their life trajectories. Although it is known that victimization and problem behaviors during middle school affect outcomes in high school, far less is understood about the co-occurrence of victimization and problem behaviors, and how such patterns relate to young adult outcomes.

Purpose: This study seeks to (1) identify sex-specific profiles of middle school victimization and problem behaviors, and (2) to assess if profiles have significant differences in young adult socioeconomic and criminal justice outcomes.

Methods: Data was drawn from a panel of middle school students (n=1,023) from the National Longitudinal Survey of Youth 1997. Latent class analysis was conducted using four indicator variables from baseline (1997): school-based victimization, attendance problems, fighting, and suspension/expulsion. Identification of profiles was based on multiple fit indices, including Akaike Information and Bayesian Information criteria, entropy, and Lo-Mendell-Rubin adjusted likelihood ratio test (LMT-LRT). After the identification of profiles, posterior probability-based imputations were used to test for differences in three socioeconomic outcomes (educational attainment, years of employment, and income:poverty ratio) and three criminal justice outcomes (ever committed assault, ever arrested, and ever incarcerated) fourteen years later.

Results: Latent class analysis suggested two classes for females (entropy = .692, LMR-LRT = 113.487): one with moderate levels of victimization and problem behaviors (16.5%), and a second with low levels of victimization and problem behaviors (83.7%). The two patterns among girls (i.e., low vs. high-risk) were associated socioeconomic and criminal justice outcomes as hypothesized.

For boys, however, three profiles were identified (entropy = .850, LMR-LRT = 34.377): a high-risk group with high levels of victimization and other problem behaviors (16.0%); a victimized group with high levels of victimization but low levels of problem behaviors (32.5%); and a low-risk group who reported low levels of both victimization and problem behaviors (51.5%).

Nuanced differences were found across the three groups of boys. Among high-risk boys, only 9.6% of them obtained a four-year college degree, 55.2% of reported committing a physical assault, 68.8% had been arrested, and 22.9% had been incarcerated. Of boys in the victimized group (high victimization, low problem behaviors), a higher percentage obtained a college degree (25.5%) and a lower percentage were in poverty than the high-risk group.

Interestingly, the victimized group’s rates of committing assault (42.6%) were not significantly less than the high-risk group, but they were less likely to have been arrested (39.7%) or incarcerated (9.6%). While victimized boys may have comparable socioeconomic outcomes to low-risk youth, the increased rate of assault raises questions about psychosocial adjustment.

Implications: Findings suggest that the middle school context may be highly influential for long-term outcomes. While the results reaffirm the importance school victimization and problem behaviors independently, they also suggest that combined patterns of victimization and problem behaviors should be identified and further studied. Likewise, the results emphasize the urgency with which schools and agencies need to address middle school victimization and problem behaviors.