The Correlates of Self-Reported Delinquency: An Analysis of the National Longitudinal Survey of Children and Youth
There were 4,293 youth sampled during Cycle III of the NLSCY between the ages of 12 and 15 years. Using weighting procedures, this sample represents 1,659,105 Canadian youth. Table 1 provides basic demographic information on the weighted sample.  Most of the youth in the sample resided in Ontario and Quebec , were non-aboriginal, and lived within dual parent families. The median annual household income was $58,098 with a range of $6,122 to $555,000.
The prevalence of self-reported delinquency for the 12-month period prior to the survey was approximately 39% - this translates to more than 540,000 youth across Canada who admitted to at least one act of delinquency during the previous year (see Table 2). The SRDS scores, which take into consideration the severity and frequency of offending, ranged from zero to 306 with a mean score of 5.3 (SD=16.6). In order to present a clearer understanding of the SRDS, we grouped the youth into five categories from non-delinquents to very serious delinquents.  As Table 2 indicates, the vast majority of youth engaging in delinquent behavior (68%) would be considered 'minor' offenders.
The offence reported most frequently was 'stealing from parents' followed by 'stealing from a store/school' and 'damaging property' (see Table 3). 'Forcing someone into sex' was the least likely offence committed followed by 'carrying a gun to defend' and 'threatening for money/possessions'.
Delinquent youth were significantly more likely to use drugs than non-delinquent youth (Phi=.29, p<.0001). Approximately two-thirds of those who reported engaging in at least one delinquent act also reported illegal drug use, while only one-third of non-delinquents admitted to using illegal drugs.
A somewhat higher percentage of males than females (43% vs. 35%) reported engaging in delinquency. Given that the overwhelming majority of young offenders processed in youth court are male, this finding is somewhat surprising. However, a more detailed examination of offence patterns revealed significant differences in the severity of offences by gender. Using the SRDS, the mean score for males was 6.7; while for females it was 3.9 (t=5.27, p<.0001). Females were equally likely to be 'minor' offenders, but were less likely to be considered 'moderate', 'serious', or 'very serious' offenders. The major difference can be found in the violent and sexual offence categories with males being three times more likely to report engaging in these behaviors . Females were equally as likely, however, to engage in drug trafficking and almost as likely to report committing property offences.
There was a pattern of escalating criminal behavior by age. According to the SRDS, the frequency and severity of self-reported delinquency increased with age (r=0.11, p<.0001). In fact, the mean SRDS for 12 year-olds was 3.2, while the mean scores for 13 year-olds, 14 year-olds, and 15 year-olds were 4.4, 5.6, and 8.0, respectively.
Approximately 41% of Aboriginal youth reported engaging in delinquent behavior compared to 39% of non-Aboriginal youth. In terms of specific offence types, aboriginals were more likely than non-Aboriginals to report violent offending (24% vs. 14%) and drug trafficking (10% vs. 5%) but equally likely to report property offending (36% versus 34%).  Using the SRDS, Aboriginal youth had a mean score of 7.9 while non-Aboriginals had a mean score of 5.3 (t=1.21, p=.226). One of the other important differences found between Aboriginal and non-Aboriginal youth was drug use. Approximately 28% of Aboriginal youth reported using illegal drugs, while only 19% of non-Aboriginal youth reported illegal drug use.
The results of the overall regression analysis indicate that negative school behavior (i.e., truancy and suspensions) accounted for the largest variation in the SRDS, followed by association with negative peers, and the Victimization Score, which measures the frequency of being threatened or physically attacked/assaulted both outside and inside the home (see Table 4). Overall, the model explained one-quarter of the variance in the SRDS.
In terms of gender differences, male delinquency was positively correlated with peer drug use and a self-reported lack of motivation or effort (i.e., gives up easily). In addition, male delinquency was negatively correlated with the Parental Nurturance Score, which indicates a more positive and nurturing parental style (see Table 5).
The unique correlates of female delinquency were failing a grade at school, the Victimization Score, and parents threatening or actually hitting their children. In addition, female delinquency was positively correlated with destroying one's own belongings, inconsistent parenting (e.g., not following through with threats of punishment), and socio-economic status. A negative correlation was also found with positive academic aspirations (see Table 6).
Common correlates to both male and female delinquency included negative school behavior , negative peers, and indirect aggression. Both models explained approximately one-quarter of the variance in the SRDS.
In order to examine the correlates of sexual offending, we identified those youth who reported forcing someone into sex and/or touching someone's private parts without permission and compared them to all other youth. The model that emerged for sexual offending was relatively weak in that the independent variables only explained a very small proportion (5%) of the variance in sexual offending. The positive correlates in the model were the General Self Score, which measures positive self-image using statements such as "I like the way I am" or "I have a lot to be proud of", and the Hyperactivity/Inattention Score, which measures behaviors such as inability to concentrate, sit still and/or wait, and parental threats/use of violence. The Parental Monitoring Score, which asks questions concerning parental knowledge of the youths' activities, whereabouts, and companions, was negatively correlated with sexual offending.
A violent offence score, incorporating assault-based offences, robbery, and carrying a firearm, was developed to examine whether or not there were any unique or particularly strong correlates to violent offending. The resulting model explained approximately 18% of the variance in violent offending (see Table 8). A comparison between the general and violent offence regression models revealed some differences. Violent offending was positively correlated with the Hyperactivity/Inattention Score, witnessing violence in the home, and the General Self Score. Both the Parental Monitoring Score and the Parental Nurturance Score were negatively correlated with violent offending. These results indicate that youth with more nurturing parents who monitor their children's leisure time more closely were less likely to report engaging in violent behavior . Finally, males and older youth were more likely to report violent behavior than females and younger youth. Otherwise, the remaining correlates - negative school behavior , parents threaten/hit youth, negative peers, positive school aspirations, Indirect Aggression Score, Victimization Score, and extra-curricular involvement - were also found to correlate significantly with general delinquency.
As with violent and sexual offending, we identified those youth who reported engaging in property offences. The model explained approximately 25% of the variance in property offending (see Table 9). Unlike the other models, negative school behavior was the weakest correlate in the model, while indirect aggression and victimization were stronger. Property offending was also positively correlated with gender (males), peer drug use, socio-economic status, failing a grade in school, and destroying one's own belongings. Time spent with friends and the Hyperactivity/Inattention Score were also positively correlated. Property offending was negatively correlated with the School Attachment Score, which measures concepts such as school spirit and positive attitudes towards academic performance, and the number of close girl friends. Thus, positive school attachment and a high number of close girlfriends decreased the likelihood of engaging in property offences such as theft, vandalism, and break and enter.
The model that emerged from the regression analysis explained approximately 18% of the variance in drug trafficking. As table 10 illustrates, there was a positive correlation between drug trafficking and negative school behavior , destroying one's own belongings, peer drug use, negative peers, and parental monitoring. Parental monitoring, however, was unique in that a higher level of parental monitoring was correlated with trafficking, rather than a lower level of parental monitoring.
- Date modified: