How are school-related factors associated with delinquency?

View the most recent version.

The second model in Table 1 and Table 2 added four school-related variables to the demographic model in order to assess whether or not commitment to school and perceptions of the school environment are associated with self-reported youth delinquency net of the demographic characteristics.

In general, the results showed that variables measuring aspects of commitment to school were significantly associated with delinquency as reported by youth. In particular, youth who indicated that they aspired to attending university had lower chances of reporting involvement in both types of delinquency, while those who indicated that they had skipped school within the last 12 months had higher chances.

Results also showed that perceptions of school safety were associated with self-reported delinquency among youth. In particular, when youth perceived that their school environment was characterized by activities such as fighting, theft, drug use and vandalism, they were more likely to report delinquency: each one-point increase in the unsafe school environment scale corresponded to a roughly 20% increase in both property and violent delinquency. These results suggest that there may be a link between certain elements of the school environment and individual behaviours. Many studies have investigated the possible effects of local contexts such as schools or neighbourhoods on individual behaviour (Sampson, Morenoff and Gannon-Rowley, 2002); however, further research would be necessary to assess whether or not this type of "school effect" is at play in the sample of Toronto youth. There was no evidence that negative attitudes towards school were associated with property or violent delinquency.

The addition of school-related variables reduced the strength of association between the demographic characteristics and the likelihood of reporting property and violent delinquency. This was particularly the case for violent delinquency (Table 2). Although the relationship between sex and delinquency was not strongly influenced by school-related factors, the effects of age and family composition were rendered statistically non-significant.

In the case of property delinquency (Table 1), school-related variables served to reduce, but not fully account for, the effects of the demographic variables. There was a reduction in the effect of living in a single-parent family of about 22% (i.e., the odds ratio was reduced from 1.99, p ≤ 0.001 to 1.56, p ≤ 0.01). Similarly, the effects of living in a step/blended family (21%), age (15%) and sex (6%) became weaker, and the compensatory effect of being a recent immigrant was reduced by about 15% (0.40 p ≤ 0.001 to 0.46 p ≤ 0.01). Thus, even given similar levels of school commitment, recent immigrant youth were still less likely to report property-related delinquency than native-born youth, while boys, older youth and youth who did not come from intact families remained more likely to report property-related delinquency.