Source: (2011) Dissertation submitted for the degree of Doctor of Philosophy in Criminal Justice. Michigan State University.

The purpose of this study is threefold: to develop an explanatory model that integrates social control theory, theory of the ecology of human development, and reintegrative shaming theory into a single paradigm that accounts for recidivism; to examine whether the processes leading to recidivism vary by demographic, background, and community characteristics; and to determine the efficacy of the restorative justice on recidivism. Using data collected on the restorative justice conferencing intervention for youthful offenders, a framework is established for testing the efficacy of restorative justice intervention compared to other court-ordered diversion programs. In order to examine efficacy, attempts to employ a longitudinal multilevel model to clarify methodological issues that would help researchers to conduct more sophisticated systematic research than has occurred in prior studies. The longitudinal multilevel model discovered several notable findings. First, sex and race appears to be most significantly related to initial level of recidivism at age 14. The consistently significant influence of sex on initial status implies that boys are more likely than girls to recidivate. Second, it is notable that a significant effect of school and family background on initial status on recidivism at age 14 is discovered in this study. However, As can be seen, the type of intervention and related characteristics show no significant effect in improving outcomes for youthful offenders. In addition, the youthful offenders in the restorative justice conference did not experience less risk of recidivism than their counterparts over time. It is important to examine the further questions of what happens over time to make this so. There are some concerns have to be taken into account. These findings have both theoretical and policy implications that contribute to the restorative justice interventions. (Excerpt).