Self-Control: Association to Juvenile Offending and Victimization

For the most part, theories of victimization have developed independently of theories of offending. However, studies have demonstrated that victims of crime share many of the characteristics of offenders, including being male, young, and African-American. Therefore, it is possible that well-established predictors of offending are also predictive of victimization. The proposed study draws on the general theory of crime (Gottfredson & Hirschi, 1990), to examine whether self-control is a consistent correlate of both juvenile offending and victimization, even after controlling for other risk factors. A key strength of this study and addition to the literature is the predominantly ethnic minority sample. It was hypothesized that poorer self-control would be associated with higher levels of juvenile offending and victimization each.

Cross-sectional data from an ethnically diverse sample of 11th-Graders from a large Southern metropolitan region were used for this study. 233 high school students (39.5% male, 60.5% female; 33% African-American, 54.9% Hispanic, 12% Other) from six public high schools took part in the study. Both active parent consent and student assent were required.

Juvenile offending was measured with 11 items from the National Longitudinal Survey of Youth (2006) and the Rochester Youth Study (Thornberry et al., 2003), including set fire, sold marijuana, physically hurt someone, and purposely damaged property. Victimization was assessed with two indicators, namely exposure to violence (3 items from Youth Risk Behavior Surveillance Survey; e.g., seen someone else get hit, kicked, punched, or beaten up) and violent victimization (3 items from Youth Risk Behavior Surveillance Survey; e.g., been hit, kicked, punched, or beaten up). Levels of self-control were measured with two subscales from the Grasmick et al (1993) self-control scale, namely, risk-seeking (4 items; e.g., I take risks just for the fun of it) and temper or irritability (4 items; e.g., when I am really angry, other people better stay away from me), and scores from the Symbol Digits Modalities Test. Control variables included age, gender, economically disadvantaged, race/ethnicity, low parental education, average school grades in past year (from school records data), parents been arrested in past year (2 items), deviant peers (8 items), and neighborhood safety (11 items).

Potential clustering in the data for students who attended the same high school was addressed by including dummy variables for school membership in the regression models. Negative binomial and linear regression models, respectively, indicated that the pattern of correlates differed among juvenile offending and victimization. As expected, poorer self-control was significantly related to higher levels of juvenile offending and victimizations, but these associations were less consistent after controlling for other risk factors. Implications of the study findings for theory and prevention efforts will be discussed.