Attitudes and Assessments towards Open Science
Previous surveys on Open Science among scientists show a clear picture: the vast majority have a positive attitude toward open science (Fecher et al. 2017; Ambrasat and Heger 2020; Tenopir et al. 2020; Christensen et al. 2020). This was also confirmed for the Berlin research area: the basic evaluation of the Berlin Science Survey already showed that open science is highly valued by Berlin scientists (Lüdtke and Ambrasat 2022a). Specifially, 78.8% of respondents say open science should be an overriding or even one of the highest goals in science (ibid, p. 4 and adopted here as Figure 8). This goal thus ranks behind the goals of originality and methodological rigor and good teaching, which are intrinsic to research, but well ahead of the goals of interdisciplinarity and social usability of research results, which are more externally applied to science.
Figure 8 Importance of different goals in the science system - normative assessments of the respondents
The Berlin Science Survey also differentiated various aspects of attitudes toward open science. In addition to the general assessment of the importance of open science, the individual benefits for the scientists were surveyed, as well as assessments of the possible effects of open science on the science system.
Assessment of the Importance of Open Science for Science
Over 83% of respondents view the expansion of open science as very important or fairly important to science (see Figure 9).
Figure 9 Importance of Open Science for Science
A comparison of the status groups, however, shows that the approval ratings for open science decrease significantly with higher status (see Figure 10). While 91.4 % of junior researchers consider open science to be 'quite' or 'very important', the value drops to 83 % for postdocs and to 75 % for professors, which are nevertheless high approval ratings.
Figure 10 Importance of Open Science for science, by status groups
A comparison of subject groups, on the other hand, reveals almost no differences in the assessment of the importance of open science for science (see Figure 11). In all subject groups, more than four fifths of the scientists find the expansion of open science practices 'fairly important' or even 'very important' for science. The variance in the assessment of importance does not seem to depend on research context or research culture, but plays out more at the level of individual opinion formation, not at the collective level of research contexts.
Figure 11 Importance of Open Science for science, by subject groups
Discourses on Open Science
Individual attitudes and opinions on the importance of open science can also be shaped by public debate and discourse that scientists perceive in different ways or in which they position themselves differently. In order to be able to map discourse positions, various statements were presented to the respondents on which they were asked to take a position. The nine different items represent statements that are heard in the debate about open science either as reasons for or concerns about open science. By having respondents take a position on these statements in the BSS, it is possible to capture the positions that scientists (would) take in the debate. In order to keep the questionnaire as short as possible, these statements were only presented to a randomized subsample of 50% of the respondents, while the other half of the respondents were asked more in-depth questions on the topic of collaborations.
The overall assessment clearly shows that a large majority expects positive effects from an expansion of open science, while only a minority of about one third of respondents also sees dangers and voices concerns. For example, 92.4% of respondents agree that open science increases the transparency of research (see Figure 12) and 74.3% agree that open science increases public acceptance of scientific work and reduces global inequality in the science system. About two-thirds of researchers believe that expanding open science will improve the quality of scientific output, and a slight majority believe that open science will increase research productivity (see Figure 12).
The three items describing dangers and risks are each agreed with by about one-third of respondents. 40.6% of respondents believe that an expansion of open science will lead to increased commercialization in science (e.g., through data tracking). 35.5 % fear an increased risk of idea theft and just under 33% expect "intensified competition in the scientific community" (see Figure 12). Although the positive assessments clearly predominate, at least one in three researchers also has concerns and shares expectations about possible dangers.
Overall, the predominantly positive assessments of the effects support the previously presented positive attitudes toward open science.
Figure 12 Agreement with statements on possible effects of Open Science
The congruence of discourse positions and attitudes is also reflected in the comparison of status groups. Predocs see the effects of open science significantly more positively than the other groups. Their expectations of an increase in productivity (68%) and improvement in quality (79.4%) are significantly higher than those of the other groups (see Figure 13). On the other hand, predocs expect potential dangers and risks less often. Among the professors, skepticism is more prominent: expectations of productivity gains (37%) and quality (56%) are significantly lower, while concerns are emphasized more strongly. In particular, the majority of professors (59%) expect increased commercialization. The postdocs' opinions are located between those of the predocs and the professors. While the "younger" scientists have a more positive attitude overall, the older scientists, and especially the professors, are clearly more skeptical in their assessment of the opportunities and risks of an expansion of open science.
Figure 13 Agreement with statements on possible effects of Open Science, by status groups
A comparison of subject groups also reveals relatively distinct discourse profiles (see Figure 14). There seems to be widespread agreement on the 'social' effects of open science in a broader sense: around three quarters of respondents from all subject groups expect the expansion of open science to reduce global inequality in the science system and to increase public acceptance of scientific research and its results. In contrast, the assessments of the possible positive effects on scientific work itself and the fears regarding negative developments in the course of open science differ between subject groups, in some cases considerably.
Here, the humanities are the most critical. A majority of 56.5% of the respondents from this discipline expect an increased commercialization of science as a result of the expansion of open science, while only 42.6% and 47.5% respectively expect improvements in the productivity and quality of scientific research. In contrast, the engineering sciences are much less concerned: only about a quarter of respondents from this field see serious risks. In contrast, two-thirds of the engineering scientists expect productivity and quality to increase (see Fig. 14).
Figure 14 Agreement with statements on possible effects of Open Science, by subject groups
The comparison of subject fields is particularly interesting with regard to the two effects that are expected to directly benefit science - the increase in productivity and improvement in the quality of research. While engineering scientists expect both effects to happen to the same extent, other subject groups show clear differences in the assessment of these two effects. In the life sciences, in particular, there is currently less expectation of productivity gains, but all the more of quality gains. Thus, the life scientists' rather pessimistic assessments of productivity gains are more similar to those of the humanities, while their expectations of quality gains are even more optimistic than those of the engineering sciences.
Such differences between subject groups indicate that the research context is of great importance for the assessment of opportunities and risks of open science. If in a particular context, open science is associated with a significant extra effort on the part of researchers, they may be more likely to doubt that productivity will go up in general, although the opportunity for quality improvement may nonetheless be seen. Conversely, productivity gains would presumably be expected in such research fields where the provision of data is relatively uncomplicated, while the reuse potential is seen as very high.
Assessment of Personal Benefits from Open Science
The BSS asked scientists how important they considered various scientific goals to be, what pressure they felt to fulfill them and how they prioritized the implementation of these goals in their own research practice (Lüdtke and Ambrasat 2022a). The largest gap between the assessment of the importance and the actual prioritization in one's own work is in open science (cf. Figure 15). At the same time, compared to other tasks and goals in science, open science is given lower priority by the scientists surveyed (Lüdtke and Ambrasat 2022a; cf. Figure 15).
Figure 15 Scientific goals, pressure of expectations and own research practice
One possible explanation for the gap between attitude and practice is the individual benefit that scientists derive from open science. The BSS took this into account by asking the respondents to what extent they personally benefit from the expansion of open science (see Figure 16).
The answers here are somewhat more reserved compared to the question about the benefits of open science for the science system as a whole (see Figure 9). While over 83% of respondents rate the expansion of open science as fairly or even very important (see Figure 15), only 40.1% say they personally benefit to a greater extent from open science (responses for "quite a lot" and "a lot"; see Figure 16). A relative majority of 37.2% sees a personal benefit from open science, but this is apparently judged to be rather moderate. Almost a quarter of the scientists state that they benefit “not very much” or “not at all” (see Figure 16).
Figure 16 Personal benefit through Open Science
How scientists assess their personal benefit from open science depends strongly on their status and the associated position in the knowledge production process (see Figure 17). Compared to the other status groups, professors are significantly less likely to state that they personally benefit from an expansion of open science. Only about a quarter of respondents in this status group say they benefit "quite a lot" or "a lot". In contrast, almost half of the young scientists surveyed see a personal benefit in the expansion of open science (see Figure 17). The very clear status group differences here point to the various positions in the science system, which are associated with different accessibility to resources. Younger scientists in particular hope that opening up science will improve their research situation.
Figure 17 Personal benefit through Open Science, by status groups
There are also clear differences between subject groups with regard to the assessment of their own benefit from the expansion of open science. At 61%, it is primarily engineering scientists who state that they personally benefit from open science. However, natural and humanities scientists also state with above-average frequency that they derive personal benefit from open science (see Figure 18).
This is a very interesting result, especially in light of the fact that there were no discipline-dependent differences in the assessment of the general importance of open science for the science system (see Figure 11). Although all subject groups equally consider open science to be important for science as a whole, the personal benefits of open science are apparently assessed very differently in different research contexts.
These results are consistent with the findings on discourse positions already presented above (cf. Figure 14), which show a very positive assessment, especially of the scientific effects in the narrower sense (increased productivity and quality) of open science in the engineering sciences. The research contexts and conditions in this subject group seem to be particularly suitable for tapping the positive potential of open science and making it usable for the individual scientists.
Figure 18 Personal benefit through Open Science, by subject groups
Relationship between Open Science research practices and attitudes toward Open Science
Now, the question to what extent individuals’ attitudes toward open science are related to the extent of their open science practices is a legitimate one. To explore and illustrate this, the statements about the potential impact from expanding open science for each of two groups that are involved in open science practices to different degrees were compared.
Figure 19 shows the comparison of the groups that publish a lot of open access with those that publish little or no open access. Figure 20 compares the group that regularly shares data with the group that shares data irregularly or not at all.
Looking at the differences in expectations between these respective groups, it appears that those who do more open science also have, by and large, more positive attitudes toward open science and expect its potentially negative sides to be less impactful than those who are less involved with open science.
For example, individuals who regularly engage in data sharing are more likely to agree with statements that open science increases the transparency and quality of research and reduces global inequality in science compared to individuals who do not share data at all or rarely share data.
On the other hand, people who regularly share data are also less skeptical about the potential problematic effects of open science. Here, for example, 33% suspect increased commercialization and 27% suspect an increased risk of idea stealing. In the group that does less or no data sharing, the approval ratings for the same statements are 44.6% and 41%, respectively.
Figure 19 Agreement with statements on Open Science, by Open Access practice
Figure 20 Agreement with statements on Open Science, by data sharing practice
This pattern is also evident in the other attitudes toward open science surveyed in the Berlin Science Survey (see Figure 21): The groups that are more involved in open science practices also consider the expansion of open science to be more important for science as a whole than the groups that are less or irregularly involved in open science activities. In addition, the more involved groups also rate their personal benefits from open science as greater.
These bivariate correlations can be interpreted in both directions. On the one hand, it may be assumed that individuals with more positive attitudes are also more likely to turn to open science practices. On the other hand, it is equally reasonable to assume that scientists in principle also affirm the practices in which they are involved anyway or for completely different reasons. We assume that both mechanisms are effective.
Figure 21 Attitudes toward Open Science by regularity of practicing different Open Science practices