Sharing data publicly adds value to research. The same data can potentially be used for new questions in different contexts. Furthermore, by making data publicly available, results become more transparent. This adds to the efficiency and quality of publicly funded research (Fecher, 2016). Individual researchers can also benefit from making their data available: Publications are reported to receive more citations if the underlying data are published (Piwowar et al, 2013).
Yet, public sharing of research data can be interpreted as social dilemma – despite the citation advantage, researchers are still reluctant to share their data. A recent survey among researchers showed that while 76% of respondents agree that research data should generally be published, only 13% of respondents have shared their data publicly before (Fecher et al, 2017).
If data sharing has so many benefits, why don’t all researchers share their data?
Besides missing incentives to share data, the aforementioned survey also concluded that knowledge regarding data management had a positive effect on researchers’ data sharing behavior. Lack of curation skills and missing awareness of suitable repositories deter researchers from sharing data (Fecher et al, 2015). This shows that knowledge regarding data management is an important factor in an individual researcher’s willingness to share their data.
The knowledge and skills associated with handling research data are subsumed under the concept of “data information literacy” (Carlson und Johnston, 2015). Promoting information literacy is traditionally considered an essential objective of German academic libraries (Piloiu, 2016). As libraries and other information service providers are expanding the definition of their competences to include research data management, they are starting to develop services aiming at promoting data information literacy (Helbig, 2017). A recent study showed that many libraries in western Europe already offer consultative research data services or plan to do so within a year. For example, 70% of the responding libraries consult or plan to consult researchers on data management plans (Tenopir, 2017).
In Germany, several universities have conducted surveys among students and researchers to assess the needs for research data services. The first analysis was conducted at the Humboldt University Berlin in 2013 (Simukovic et al, 2013). In the following years, other universities used and modified the questionnaire for their own purposes, eleven of which have published the results of their needs assessments. Of these eleven surveys, eight (Humboldt University Berlin, University Münster, University Kiel, University Marburg, University Kassel, University Hamburg, University Hannover and University Gießen) covered multiple disciplines and asked specifically for information services (Table 1).
A systematic comparison of these eight surveys shows that only a small percentage of respondents (6.8-16%) expressed no need for information services.
| Humboldt University Berlin | University Münster | University Kiel | University Marburg | University Kassel | University Hamburg | University Hannover | University Gießen | |
| published | 2013 | 2014 | 2015 | 2015 | 2015 | 2016 | 2016 | 2017 |
| n = | 499 | 667 | 218 | 427 | 207 | 90 | 247 | 266 |
| General | 30.66% | 39% | 41.28% | 47.2% | 56.2% | 31.2% | 51.4% | 42% |
| publishing, citation | 31.46% | 33% | 35.32% | 41.3% | 26% | 33.2% | 38.1% | 41% |
| technical | 47.47% | 48% | 59.17% | 66.1% | 69.8% | 56% | 64.4% | – |
| curation | – | 44.5% | – | |||||
| legal | 51.3% | 53% | 55.5% | 61.5% | 66.9% | 53.2% | 67.2% | 70% |
| long term preservation | – | – | – | – | – | – | – | 59% |
| no need | 9.61% | 16% | 9.17% | – | – | 6.8% | 9.7% | – |
Table 1: Results on data management surveys across German universities
Across all surveys, most respondents (51.3-70%) expressed the need for information services concerning legal issues, such as copyright law, licensing and protection of privacy. Between 47.47 and 69.8% were interested in information on technical/curation aspects, including metadata, standards and digital preservation. Respondents showed little less need for information on research data management in general (30.66-56.2%) and on publishing and citing data (26-41.3%).
The systematic comparison of needs assessments points to a high demand for information services in general. It also reveals a similar pattern across all surveys. This indicates that there are common information needs among German students and researchers, especially concerning legal and technical/curation issues.
As mentioned before, knowledge gaps prevent researchers from publishing their data. Since the demand for information services covering legal and technical/curation aspects are especially high, the lack of knowledge in these areas appears to be a significant barrier to data sharing.
Measures aiming at changing researchers’ data sharing behavior should take their information needs into account. Libraries and other information service providers can – and some already do – use their expertise in information literacy instruction and support researchers in overcoming these obstacles.
One should keep in mind that these surveys were sent to researchers who did not have any research data management experience and most of them are not publishing their data. Therefore, the answers represents what researchers think they need the most (i.e. what libraries should offer to attract researchers to follow any workshop), but it does not tell us anything about what they would actually need the most to go toward open data.
We would need a different type of survey given only to scientists doing open data to answer the latter question. But most research data professional would have pretty good guesses that data literacy, metadata creation and knowledge about repositories and “data life cycle” are key knowledge allowing to create open data in a smooth and efficient workflow. In short, one needs to plan the aquisition of the data with the one question in mind: how will I share this data and how can it then be re-used by my peers?
I agree that these surveys do not necessarily reflect the information needs of researchers with experience in sharing data – in all surveys, the group of respondents who shared their data publicly before was quite small.
However, I think the surveys are useful for indicating knowledge gaps that prevent researchers from sharing (or thinking about sharing) their data in the first place.