188 Chapter 7 and training methods offered to prison staff. Based on our experiences, we consider the responsive approach in evaluations well suited for the need to have more ‘shared ownership’ of the methods used and to create fruitful interventions with all stakeholders. This helps adjust implementation styles and methods used along the way. Responsive evaluations do not necessarily always have to be done via academic research. The DCIA Educational Institute can corporate this approach in their own evaluations of training methods. However, research on the evaluation and impact of ESS of prison staff helps further develop evidence-based ethics trainings for prison staff. IDEAS FOR FUTURE RESEARCH Research design for more insights on practice improvements after MCD Our research design was not appropriate to research whether MCD created actual improvements in practice since we only had evaluation forms filled out directly after MCD sessions. Furthermore, our impact study on the series of sessions focused on the specific impact on MCS instead of the impact to practice. A study on the actual influence of MCD on practice would benefit a richer understanding of its value for daily prison work. Several options for such research are available. First, interviews with MCD participants or case presenters after several weeks of attending MCD sessions could provide information on how content-based insights were translated into practice. Secondly, research using the method of a ‘participatory action research’ (Baum et al., 2006; Kindon et al., 2007; McIntyre, 2008) could potentially fit the aims to improve practices and stimulate an active shared ownership with the involvement and representation of all stakeholders. Finally, in an impact study (such as in Ch.5), you can create a repeated and long-term measurement by the MCSQ after, e.g., one or two years. Ideally, there is a combination of short- and long-term measurements (with a ‘post-measurement’ and a ‘post-post measurement’). For example, an impact-study could have three sub-groups to see what is needed for MCD to have more or a lasting impact: a) the original intervention group who continued participating in MCD sessions several times a year, b) a part of the original intervention group who stopped participating in MCD after the first series, and c) the part of the control group who still have not started MCD sessions. Furthermore, it could be interesting to analyze differences in the impact to practice between team-based MCD and sessions with a multidisciplinary composition of participants.