129 Contribution of MCD to moral craftsmanship significant differences on 4 of the 70 items. Prison staff who evaluate MCD positively indicated more frequently that they ‘are given tools to deal with moral dilemmas at work’ and ‘as a team ask themselves whether they are doing the right thing during their work’. A possible explanation is that participants who evaluate MCD positively are more aware of the tools they are given at work, and they consider reflecting on the question whether they are doing the right thing more valuable. In addition, participants who evaluate MCD positively indicated less frequently that ‘their work consists mostly of routine actions’ and more often that ‘it is allowed to make mistakes in their work’. Correction for other factors (multivariate regression analyses) The multivariate linear regression analysis showed that other variables had little or no influence on the impact of MCD on MCS; the items that showed significant differences are the same as those in the main analyses. Finally, an analysis was performed excluding the prisons of Almere and Zwaag. This analysis showed that the previously described differences sometimes increased or decreased on certain items, but there were no consistent changes in a particular direction. DISCUSSION In this study, we investigated the impact of a series of MCD sessions on elements of MCS for Dutch prison staff through a pre- and post-measurement questionnaire. This research is innovative in at least two ways. First, the concept of MCS is relatively unexplored and to date no validated questionnaire is available to measure impact on MCS. We therefore developed a questionnaire with several items based on an exploration of the concept of MCS from various literature sources. Second, ethics support in general, and MCD in particular, has not been offered, implemented and researched in prisons. Dutch prison staff is not used to specific methodologies for reflecting upon their moral challenges or dilemmas. Most research on outcomes of MCD in general, not specifically for impact onMCS, has been conducted in health care settings, where ESS and MCD are offered and implemented on a regular basis (Clark & Taxis, 2003; Gallagher, 2006). In quantitative research in the health care context no positive effect of MCD was found on reduced moral distress (Kälvemark Sporrong et al., 2007) or improved work climate (Forsgärde, 2000). Both studies indicate that it is, among other things, difficult to measure the effects of a complex intervention 5