This paper presents an integrated theoretical model of the etiology of delin quent behavior. The proposed model expands and synthesizes strain, social- learning, and social control perspectives into a single explanatory paradigm which avoids the class bias inherent in traditional perspectives and takes into account multiple causal paths to sustained patterns of delinquent behavior. Major variables in the model include two types of social bonds (integration and commitment), attenuation experiences (such as failure in conventional social settings), and delinquent learning and performance structures. Two dominant etiological paths to delinquency are postulated by the paradigm. The first path is characterized by weak integration into and commitment to the social order and by involvement in delinquent peer groups. Characteristics of the second major path are initially strong bonds to the social order, subse quent attenuation, or weakening, of those bonds, and exposure and commit ment to delinquent peer groups.

This article examines sex differences in family bonds as a possible explanation of sex differences in self‐reported delinquent behavior (SRD) among a national probability sample of 1725 adolescents. It was hypothesized that girls would report significantly stronger family bonds and significantly lower delinquency than boys. The findings provide limited support for these hypotheses. The expected sex differences in SRD were observed. males report significantly greater involvement in all forms of delinquent behavior. However, the expected sex differences in family bonds were not observed. Family bonds are modestly correlated with delinquency, but at comparable rates for males and females, and with evidence of their greater association among males in many cases. Theoretical and methodological explanations are suggested for the divergence of the observed findings from existing research.

This study describes the distribution, nature, pattern, and magnitude of sex differences in self-report delinquency. A national probability sample of 1725 youths, ages 11-17, was interviewed using a comprehensive measure of self-report delinquency including serious and violent crimes. Major findings include (1) small but consistent sex differences in numerous delinquent behaviors and in most types of delinquent behavior due to higher offense frequencies of males and greater numbers of male offenders, (2) generally similar patterns of male and female delinquency (r = .82).(3) sex differences in the proportion of high-frequency offenders, (4) few interaction effects of sex with other demographic variables, and (5) stable sex differences in certain delinquent acts across a decade.The debate over the nature, magnitude, and distribution of sex differences in delinquency continues, despite a proliferation of studies. While there is general agreement between official and self-report studies that males are involved in more delinquency than females, there is widespread disagreement over the magnitude and nature of the sex difference. Indeed, the controversy extends to the very method (official or selfreport measures) that is best suited to resolve the debate.Studies using official and self-report measures of delinquency agree that males are involved in substantially more delinquent behavior than females, but disagree as to the magnitude of this difference. In official statistics malejfemale ratios generally fall within the range of 3:l to 7:l (Cockburn and Maclay, 1965;Edwards, 1973;Wattenberg and Saunders, 1954;Wolfgang, 1979). The 1978 Uniform Crime Reports indicate a maldfemale ratio of 3.6:l for youth under 18 (US. Department of Justice, 1979). Malejfemale ratios observed in self-report delinquency AUTHORS NOTE: The author i s indebted to Drs. D. Wayne Osgood, David Huizinga, and Hart Weichselbaum for their valuable comments on earlier drafts of this article.

A theoretical analysis and a review of the uses of role-playing and deception methods support the position that role-playing methods are based on a more comprehensive and inclusive conceptualization of human behavior than are deception methods. It is shown further that role-playing assumptions are better able to account for the empirical findings in research on the behavior of subjects in laboratory experiments. It follows that role-playing studies should be the standard against which deception studies are compared rather than vice versa as is the current practice. Several recent developments in role-enactment methods are discussed. These innovations are categorized according to whether their major functions are for exploratory, descriptive, or parametric analyses.

Studied the relationships between personal locus of control (Rotter's Internal-External Control Scale), teaching discipline, student performance, and satisfaction. 54 undergraduates were randomly assigned to 2 teaching conditions. In the high discipline condition, Ss were exposed to a 2-hr "mini-course" on computer programming that adhered to 5 empirically derived dimensions of perceived high discipline. In the low discipline condition, Ss were handed identical computer programming materials and told to proceed at their own pace with no rules. Results show that Ss high on internal locus of control performed better under low discipline conditions, while high external control Ss performed better under high teacher discipline conditions. (24 ref)