RT Article T1 Self-Control Theory and Nonlinear Effects on Offending JF Journal of quantitative criminology VO 29 IS 3 SP 447 OP 476 A1 Mears, Daniel P. 1966- A2 Cochran, Joshua C. A2 Beaver, Kevin M. 1977- LA English YR 2013 UL https://krimdok.uni-tuebingen.de/Record/1764279816 AB Objectives This paper examines Gottfredson and Hirschi’s (A general theory of crime. Stanford University Press, Stanford, 1990 ) self-control theory and develops theoretical arguments for why self-control may have a differential effect on offending depending on the level of self-control. Methods We test the argument that the association between self-control and violent offending (n = 5,681) and non-violent offending (5,672) is nonlinear by using generalized propensity score analyses of data from the National Longitudinal Study of Adolescent Health. Results The results indicate that self-control and offending are nonlinearly related in a manner that involves two thresholds. Specifically, among individuals at the high end of the self-control spectrum, there was little evidence of an association between variation in self-control and offending. However, among individuals in the middle part of the self-control spectrum, a positive association obtained—that is, the greater the level of low self-control, the greater the likelihood of offending. Finally, among individuals at the low end of the self-control spectrum, there was, once again, little evidence of an association. Conclusions A nonlinear association between self-control and offending may exist and have implications for self-control theory and tests of it. Studies are needed to investigate further the possibility of a nonlinear association and to test empirically the mechanisms that give rise to it. K1 Nonlinear effects K1 Generalized Propensity Score K1 Offending K1 Self-control DO 10.1007/s10940-012-9187-5