This study examined the characteristics typically seen before the start of delinquency. Questionnaire surveys were administered twice to 1 524 junior high school students, in the first and again in the second school term. Students were classified into three groups according to the time when they started engaging in delinquent behavior: "experienced" (having engaged in delinquency before the first survey: N = 304), "started" (reporting their start of delinquency after the first survey: N = 157), and "no experience" (having no experience of delinquency: N = 1 063). Comparisons of the three groups showed that the "started" group shared the same characteristic factors with the "experienced" group. Compared with the "no experience" group, both delinquency groups reported a less intimate relationship with their parents, more experience of domestic abuse, less parental control, and more conforming behavior with their friends. Also, the "started" group showed a sharp change in their degree of enjoyment of school life. Although the "started" group reported the same level of school life before starting delinquency as the "no experience" group, they reported a lower degree of enjoyment, nearly equal with that of the "experienced" group, after starting delinquency.

This research investigates repetitive problem behavior (in-school problems and mild delinquency) in Japanese junior high school students from their first to third year (12-15 years old) and examines the risk factors for repetitive problem behavior among parent-child relationships, school, individual, and lifestyle factors. An analysis of the longitudinal data of 470 students (Data Set 1) found that repeated problem behavior in school was affected by low self-control, poor relationships with teachers, and lack of parental monitoring. Repeated mild delinquent behavior was affected by parental violence, poor relationships with mothers, less sleep duration, and not belonging to club activities. Some students exhibited both mild delinquency and in-school problems. Students who experienced problem behavior 10 times or more in their second year had often experienced some problem behavior (1-9 times) in their first year. Analysis of the longitudinal data of 368 students (Data Set 2) showed that in many cases, students who exhibited repeated delinquent behavior in their second year continued to exhibit problem behavior after they became third-year students. Though the number of students with repetitive mild delinquency that was totally remedied in their third year was small, a low incidence of parental violence was seen as an antecedent factor for students who did not commit mild delinquency in their third year in spite of a high frequency of delinquency in their second year.

Today, adolescents can easily make contact with strangers online. The present study examines the relation of online communication with strangers (OCS) and mild delinquency among junior high school students and explores which risk factors are associated with both types of behavior, taking gender differences into account. A total of 1873 Japanese adolescents (aged 12–15) with mobile phones completed online questionnaires assessing internet usage (duration, variability), parental behavior (violence, monitoring), and adolescent characteristics (depression, self-control) as potential predictors of OCS and mild delinquency. Sequential model testing revealed that the duration of internet usage was associated with both dependent variables, whereas its variability was associated with OCS alone. Self-control and parental violence predicted both outcome variables in girls, but only mild delinquency in boys. Depression was related with both behaviors in boys. Parental monitoring was negatively associated with OCS among girls and with mild delinquency among boys. Overall, OCS appeared to have a similar function for girls as mild delinquency did for boys. Implications for future studies and preventive programs will be discussed.