| This dissertation examined the relationship between four different measures of precocious opposite sex relationships, and the development of delinquency, in a sample of high-risk youth, from the city of Rochester, New York. The goal of the analysis was to determine whether early dating, opposite sex friendships, mixed gender peer groups, and sexual activity, could be linked to the expression of delinquent behavior one year later. Furthermore, this study investigated whether these "bad influences" varied developmentally, and across gender. Finally, an attempt was made to explain the theoretical nature of these relationships. Previous research had established that precocious transitions are associated with anti-social behavior, and that early dating may be criminogenic; however, it has not presented a coherent understanding of these phenomena.;The analysis confirmed that certain opposite sex relationships are important for understanding the development of delinquency in this group of adolescents. Significant differences were detected based on gender, and developmental period; however, it is difficult to summarize the pattern of effects. No sweeping generalizations such as "the effects are overwhelmingly stronger for girls" could be drawn. There was evidence that sexual activity may be the key risk factor for males, while mixed gender peer relationships are most salient for females. The theoretical model examined here was able to partially explain the "bad influences" effect; however, a substantial number of direct relationships remained unmediated. Opposite sex relationships seemed to impact delinquency most directly by influencing exposure to peer deviance, but the pattern of results was complex, and required in-depth discussion to be fully appreciated. Nonetheless, the author suggests that the results are sufficiently compelling to warrant further research in this area.;The data for this project were drawn from waves 3 through 9 of the Rochester Youth Development Study (RYDS). Three age cohorts were created: the early cohort represented children attending middle school, the middle cohort was composed of children entering high school, and the late cohort consisted of subjects who had been in high school for approximately one year. All of the research hypotheses were evaluated using negative binomial regression analysis. |
