184 Chapter 7 as seen in our data; in that case, it becomes even more important to translate moral reflections into concrete plans for a ‘follow-up’ after MCD. Ensuring such a follow-up on the conclusions of an MCD session is considered to be one of the quality characteristics of MCD (Hartman, 2020, p. 152). As Sennet (2009) stated, a craftsman wants to do good in practice for the sake of the work itself. Therefore, an organization needs to offer room for professionals to be able to actually do good or change their practice for the better. More attention for and explicit reflections about the level of influence of prison staff, and their responsibilities according to their role within the organization, would be beneficial. It could help if, more explicitly, staff reflect on following up MCD in terms of who is responsible for what part of the chosen plan. MCD facilitators should help MCD participants construct such ideas as concretely as possible before returning to the work floor. In Chapter 3, we saw how participants stated that the last step of MCD – in which concrete plans are discussed as conclusions of the sessions – often did not get enough time. Perhaps we need to explicitly add a reflection on the responsibilities regarding the last step of the dilemma method, which then also takes the existing protocols and hierarchical structures into the consideration. Some other conversation methods already add this more explicitly (such as the ‘Rotterdam Routekaart’, see e.g., De Beaufort et al., 2008, p. 29). Furthermore, it could help if insights fromMCD are discussed more during team meetings, to follow up on agreements on role and task responsibilities; this can help achieve goals for practice improvements. Such attention during team meetings is only advisable when the case presenter of the MCD is okay with sharing the topic more broadly. MCD sessions with a concrete follow-up can result in a more collaboratively supported strategy about what good action means and can thus lead to new visions or policies within the organization (De Bree & Veening, 2012, p. 13) A top-down and bottom-up approach Active participation of staff is essential, especially during the initial phase of a change in an organization or the start of an intervention (Kayzel, 2002, p. 94). A combination of a top-down (with structural training plans and an active stimulation of moral reflection in practice by team supervisors and managers) and a bottom-up approach (which bases plans for improvements on the needs of staff, adapts methods to the circumstances of teams and locations, and stays adaptive by having responsive evaluations) seems needed, as advised in the study of Hartman (2020, p. 157). Having only a top-down style during implementation can generate resistance, and it might even be experienced as ‘pedantic’ (Karssing, 2000, pp. 100–101). There is a strong need to stimulate the shared