This study of 300 women and 300 men graduates of a boot camp finds that there are noteworthy gender differences in predictors of tenure in the community without criminal recidivism in a 5-year follow-up. The Cox proportional hazards models show that urban residence, childhood and recent abuses, living with a criminal partner, selling drugs, stress, depression, fearfulness, and suicidal thoughts are stronger positive predictors of recidivism for women than for men. Men are more likely to return to prison because of criminal peer associations, carrying weapons, alcohol abuse, and aggressive feelings. Job satisfaction and education lengthen time in the community more for men than women, whereas the number of children and relationships are more important to tenure in the community for women. The implications for the findings for theory are discussed.

No abstract

No abstract

This study of a statewide, stratified random sample of 3,335 public high school students is designed to examine the robustness of self-control in predicting alcohol consumption, other drug use, crimes against persons, and property offenses. Comparisons in prediction of these forms of delinquency are made between two measures of self-control, social characteristics, and elements of other theories. An examination also is made of whether self-control accounts for commensurate amounts of variance within categories of social characteristics. A behavioral measure of self-control generally is a stronger predictor of delinquency than a cognitive measure. Inconsistent findings are observed in analyses of whether self-control accounts for commensurate variation in delinquency between and within categories of these social characteristics. Both measures are significant predictors of delinquency when they are simultaneously analyzed with many factors extrapolated from other theories. Implications of the findings are discussed in terms of future theory and research.

This is a study of 1,093 adolescents who attended public high schools in rural Arkansas; Little Rock, Arkansas; or Baltimore, Maryland. The results do not support the hypothesis that antiascetic behaviors are more affected by religiosity than are criminal behaviors. Nor does the study support the theoretical hypothesis that religiosity is an antecedent factor that has effects that are fully mediated through other more proximate elements of social control. It also finds few real differences in the effects of religiosity on various forms of delinquency between Baltimore, Maryland; Little Rock, Arkansas; and rural areas of Arkansas. Implications of the study are discussed.