Summary of important issues that could be addressed through policy changes and governance adaptations on different levels.

Despite not focused explicitly on research policy and research governance, the interviews conducted with stakeholders within the framework of ROSiE yield some tentative insights that might merit further consideration. In addition to developing performance assessment schemes that reward open science, which obviously has an important policy component but was already outlined in the previous sub-chapter, interviewees mentioned several issues that could be addressed through policy changes and governance adaptations on different levels. 

Some interviewees pointed out that their countries lack a national open science policy, which they consider a problem. In their view, this tends to make efforts to promote open science more difficult because the lack of explicit guidance on the national level tends to obfuscate the value and importance of open science to researchers. Thus, national open science policies could be helpful to signal to researchers that political commitments to support and promote open science are credible.

Moreover, interviewees agreed that funding open science practices is crucial, an issue closely related to aligning open science practices to incentive structures. However, opinions on current Research funding organization (RFO) performances varied to at least a certain extent. Some interviewees view current RFOs policies as exemplary, whereas others rather pointed to the heterogeneity of RFO approaches and policies. 

Yet in general the viewpoints of interviewees who explicitly elaborated on the role of RFOs seemed to converge on the argument that open science should be rewarded financially rather than (only) mandated by, for example, grant requirements. Consequently, interviewees consider RFOs key actors in the open science transition and overall expressed a preference of rewards over conditionality. 

Another issue two interviewees mentioned that potentially affects the effectiveness of governance arrangements is the training of reviewers. If open science should be considered in reviews, reviewers need to have the necessary competences to assess whether commitments of research teams and consortia are in line with standards of good practice. One touched upon this issue when discussing competences of grant reviewers, while the other interviewee referred to ethics advisers on the institutional level (somewhat similar to a Research ethics committee). 

When it comes to how institutional governance schemes could help to support researchers in implementing responsible open science practices, several interviewees stated that data stewards on the institute or faculty level could be effective advisers. Unlike advisers on the general institutional level, interviewees see data stewards as better positioned to offer specific advice because of their familiarity with disciplinary cultures and challenges. Furthermore, they are usually more accessible than advisers located on a higher and thus more distant level. The system implemented by the Technical University of Delft was highlighted as a good practice example in two interviews. More generally, various interviewees alluded to the importance of offering guidance on the appropriate level and emphasised the importance of meeting the needs of target groups. 

On a more general level, one interviewee pointed out that at least in her country (located in Central Europe) efforts to promote responsible open science are somewhat hampered by the fact that the research ethics and integrity community and the open science community are largely separated. In her view, creating and strengthening linkages between these communities could help to increase synergies between ethics, integrity and open science.

Finally, interviewees from all stakeholder groups perceive citizen science positively and believe that it can make an important contribution to the open science transition. Several interviewees mentioned citizen science either when asked about what they associate with open science or when asked about good practices they would recommend ROSiE to include among the tools the project will develop. As ROSiE is built around the assumption that citizen science and open science are closely related, this might be counted as tentative evidence that this view is shared in the research community.

---

This passage is part of D3.3: Report on interviews written by Tom Lindemann, Lisa Häberlein, Philipp Hövel.