With or Without Them: Improving Self-Control in Juvenile Offenders

ABSTRACT

Increasing self-control is a key-factor in the rehabilitation process of young criminals. Based on two well-established theories of self-control, we tested a short mindfulness training and the motivational effect of group identification on a standard self-control measurement in an sample of 57 incarcerated male youth from a German youth detention center. In accordance with our hypothesis, both treatments led to higher self-control than a control group. These findings indicate that mindfulness and group identification foster self-control even in an untrained sample of incarcerated young men and contribute therefore to resource-oriented successful rehabilitation programs.

Acknowledgments

We thank Joachim Güttler, superintendent of JVA Iserlohn, and Gerd Asselborn, director of psychological services at JVA Iserlohn, for supporting the conduction of our study. All authors contributed to the study concept and design. Testing and data collection were performed by I. M. Sroka. I. M. Sroka and S. D. Isemann performed the data analysis and interpretation under the supervision of E. Walther. All authors contributed equally to the draft and following revisions. All authors approved the final version of the manuscript for submission.

Notes

Beside Cohen’s d, two additional effect sizes were computed building on the tripartite assumption. According to Trafimow (2017), the tripartite assumption assumes that there are three sources of systematic variance in the dependent variable : systematic variance associated with the manipulation (), with other (unknown) variables (, and randomness (). The author argues that Cohen’s d denotes in the dependent variable and therefore confounds and . To eliminate this confound, the square root of the total variance could be replaced with the square root of either or . In the case that is used, it clears the denominator from the independent variable and from randomness (see also Trafimow, 2014). In the case that is used, it clears randomness from the independent variable and from other (unknown) variables. Accordingly, we define effect size other (ES_O) as the difference between d2-R CP means divided by the square root of and effect size random (ES_R) as the difference between d2-R CP means divided by the square root of . For the necessary equations please see Trafimow (2017). As reliability measure of the dependent variable ( we used the internal consistency (Cronbach’s alpha) of .92 as reported by Brickenkamp et al. (2010) for d2-R CP for 17- to 19- and 20- to 39-year-old participants.

ES_O = 0.12, ES_R = 0.41.

ES_O = 0.45, ES_R = 1.50.

ES_O = 1.16, ES_R = 3.50.

ES_O = 1.54, ES_R = 4.35.

We chose the pretreatment SD of the groups as denominator for Glass’s Δ concerning the within-subjects comparisons.

ES_O = 3.70, ES_R = 3.12.

ES_O = 4.47, ES_R = 3.55.

ES_O = 0.19, ES_R = 0.63.