Research That Matters (January 17 - 20, 2008)

Background and Purpose: In spite of recent declines in juvenile delinquency (Juvenile Offenders and Victims, 2006; National Report, 2006), involvement in delinquent activity remains a serious threat to youth outcomes (Indicators of School Crime and Safety, 2005). Agnew's General Strain Theory (GST) explains juvenile delinquency as resulting from “strain” or the uncomfortable emotions individuals feel in response to encountering negative stimuli, failing to achieve goals, and difficult transitions. GST also considers vicarious and anticipated strain. Vicarious strain refers to real-life strains of others that an individual hears about or witnesses. Anticipated strain refers to an individual's expectation that a strain will reoccur or develop. The purpose of this study was to assess patterns of potential vicarious and anticipated strain among middle school students and determine their relationship to outcomes negatively associated with delinquency.

Methods: A random subsample of 1000 students was drawn from a sample of 13,405 middle school students in a dataset containing self-report social environmental and outcome data. Data were collected using the School Success Profile. Of the subsample, 31%, 37%, and 32% were in 6th, 7th, and 8th grade, respectively. Just over half were boys (52%); 45% were African American, 8% were Latino, 40% were White, and 7% identified with other race/ethnic categories.

Latent Class Analysis with MPlus was used to identify patterns of strain based on three continuous vicarious strain variables (crime in the neighborhood and school, friend involvement in problem behavior) and two categorical anticipated strain variables (fear of being hurt or bothered in the neighborhood and at school). The class model was validated using three additional random samples of 1000 students and by examining the association of class membership with trouble avoidance, social support, school engagement, and academic performance.

Results: A 5-class solution emerged as the best model. The first class (64.3% of the sample) had low strain scores on all indicators (Low Strain). A second class (12.3%) was distinguished by high vicarious school strain and a high probability of school anticipation strain (School Strain). A third class (7.9%) was distinguished by its high neighborhood vicarious strain scores (Neighborhood Strain). The fourth class (11.4%) was characterized by high vicarious friend and school strain (2-Domain Strain). The fifth and smallest class (4.1%) experienced high vicarious strain in all three domains as well as a high probability of anticipation strain in the neighborhood (3-Domain Strain).

Performance in assessed outcomes was associated in expected directions with class membership in most cases. The Low Strain class had a pattern of the best outcome scores, followed by the School Strain class. In general, the worst patterns of performance were demonstrated by members of the 3-Domain strain and the Neighborhood strain classes.

Conclusion and Implications: Consistent with Agnew's GST, the results suggest that vicarious and anticipated strain in one or more domains is associated with outcomes commonly linked to delinquency. The findings suggest that school-based social workers must work with public safety officials, neighborhood groups, and school staffs to promote stability and safety in the environments of students.