119 Contribution of MCD to moral craftsmanship (Principal Component Analysis with varimax rotation), we checked whether the 70 items formed separate clusters. As this was not the case, we continued further analyses on the individual item level. To check whether the results on the 70 ordinal level items could be analysed as continuous variables, i.e., allowing multivariate analyses, we compared the P-values of the Chi-squared test and the independent sample t-test, between the pre- and postmeasurement in both the control group and the intervention group. As these results were fairly similar, more advanced analyses were possible. ANOVA tests were performed to compare the mean scores of the items on differences between a) the pre- and post-measurement within the control group, and within the intervention group, b) the control- and the intervention group at the pre-measurement, and c) the control- and intervention group at the post-measurement. In subgroup analyses, we examined whether certain professional disciplines showed larger or smaller changes between the pre- and post-measurement on specific items. Furthermore, ANOVA tests were conducted to see whether the number of MCD sessions and the evaluation of these MCD sessions affected the impact of the intervention. In multivariate linear regression analyses, the effect of the intervention of MCD was adjusted for differences between the control- and intervention group in the various professional disciplines involved, having or not having contact with prisoners, sex, age, education level and duration of employment. We added a multi-level component to take into account that some of the participants completed both the pre- and postmeasurement. Next, a multiple regression was conducted to check whether the closing of the prisons of Zwaag and Almere had an effect on the results. Finally, we performed an extra analysis in the subgroup of participants who completed both a pre- and post-measurement, and whose scores could be compared using a coded link. We compared the impact of MCD on MCS in this linked file to the results of the main analysis. To this end, the same analyses as described above were performed. Research ethics The Institutional Review Board of Amsterdam UMC (location VUmc) waived the Dutch Medical Research Involving Human Subjects Act (WMO). The DCIA of the Ministry of Justice and Security in the Netherlands gave permission and all involved researchers 5