Institutional Framework for the Implementation of Open Science
Difficulties in the Implementation of Open Science
When it comes to the question of where politics and university management can provide support in order to further advance the implementation of open science, it makes sense to examine which scientists see difficulties in the implementation in the first place. Only where difficulties are perceived and a need for support is identified do targeted support measures which are not based on incentives but on cooperation with the researchers concerned have a chance of success. In the Berlin Science Survey, researchers were also asked for each individual open science practice whether and to what extent they perceive difficulties in its implementation.
This showed that a comparatively large number of scientists had problems evaluating the difficulties in implementing individual open science practices (see Figure 22). This is especially true for citizen science (41.9%), but also for open peer review (24.4%) and code / material sharing (22%).
Furthermore, even the most established open science practice, open access publishing, still shows hurdles: nearly 25% of respondents say they see great or very great difficulties in implementing OA. This figure exceeds the proportion of respondents who see great or very great difficulties with open peer review (18.9%) (see Figure 22). However, far fewer respondents can assess the difficulties with open peer review (OPR). Looking only at respondents who can provide an estimate, 25% indicate great to very great difficulty for OPR and 26.2% for OA (not shown).
The situation is different for data sharing, where there appear to be the most difficulties. A good 40% of respondents report great or very great difficulties with implementation. If again only those respondents who can give an assessment of the difficulties are considered, the figure is as high as 47.9% (not shown).
Figure 22 Difficulties in implementing Open Science practices
A comparison of the status groups reveals surprisingly few differences in the assessment of implementation difficulties (Figure 23). Only in the case of open peer review do the professors see significantly more difficulties and in the case of citizen science somewhat fewer than the other status groups.
Figure 23 Difficulties in implementing Open Science practices, by status groups
The extent of the problems in implementing open science practices is assessed significantly differently depending on subject groups (see Figure 24). Humanities scientists see great or very great difficulties in all five open science practices more often than average. It is also striking that the natural scientists are significantly less likely than the other subject groups to report difficulties in implementing both data and code /material sharing. Thus, they form the contrast profile to the social scientists, who see difficulties in sharing practices more often than average. The type of data plays a major role here. If data is personal or even sensitive, qualitative or difficult to anonymize, then this sometimes presents major hurdles for provision.
Figure 24 Difficulties in implementing open science practices, by subject groups
Need for Support from the Institutions
Since, overall, the hurdles to implementing open science are still considerable, the question arises as to what extent institutions can provide support in eliminating difficulties. The respondents were therefore also asked about the need for support in the implementation and expansion of open science practices by their own institution. On average, 40.5% indicated such a need for support (see Figure 25). Differentiation by status groups shows that professors are the least likely to indicate such a need, and non-doctoral scientists the most likely (see Figure 25). A comparison of subject groups reveals an above-average need for support among engineering scientists and social scientists (see Figure 26).
Figure 25 Need for support with Open Science, by status groups
Overall, it is clear that there are still many difficulties in implementing open science practices. Even in the case of already well-established open access publishing, almost a quarter of the respondents still see difficulties. Accordingly, the expressed need for support is also relatively high, but this also varies between status groups and research contexts.
Figure 26 Need for support with Open Science, by subject groups
Assessment of the Berlin Research Area with regard to Open Science
In order to gain a comprehensive insight into the research realities of the scientists, they were finally also asked to assess the Berlin Research Area with regard to various aspects (Lüdtke and Ambrasat 2022a). The implementation of open science was rated as less good compared to other goals. Only a share of 39.7% of the respondents assess this aspect of the Berlin research area as "somewhat good" or "very good". Although 32.8% of respondents to the survey on the topic of open science say that they are unable to assess it for the Berlin research area (see Figure 27), even if these respondents are excluded from the calculation, the result for open science is only a slightly positive assessment of 59.1% (Figure 28).
Figure 27 Assessment of the Berlin Research Area
Figure 28 Assessment of the Berlin Research Area, without "I cannot judge"