75% of researchers make their data accessible
Do Swiss researchers share their data with other researchers and with the public? And if not, why? Which data repositories and other channels do they use for data sharing? A large-scale survey by the SNSF and swissuniversities offers some answers.
More than 2000 researchers returned the questionnaire. Approximately 1500, or three quarters, provide access to their data in some form. However, only 44 per cent of these researchers use public repositories or data archives for this purpose. A further 38 per cent publish the data in scientific journals and/or make them available on their own websites. But these channels are not optimal for data sharing, and they do not guarantee that datasets are easy to locate and access, as well as to cite and reuse. The remaining 18 per cent of researchers share their data only when personally requested to.
SNSF takes reservations into account
The main reason why data are not accessible, or at least not accessible in repositories: scientists want to publish their research work first. Doubts as to rights of use or confidentiality issues are also significant reasons. Moreover, researchers often point to time constraints, low data relevance and a lack of knowledge about repositories.
The SNSF accepts these reservations. It only asks researchers to disclose data that underlie published findings - by no later than the publication date. And the data are only to be made accessible to other researchers if there are no obstructing legal, ethical, copyright, confidentiality or other clauses.
High importance of discipline-specific repositories
When data are made available in repositories, this happens in general repositories and on a multitude of smaller discipline-specific platforms in equal measure. However, for reusing other researchers' data, the majority of those interviewed turned to discipline-specific repositories. This is the case even in disciplines in which accessibility as such is largely given via general platforms.
All in all, the survey respondents named 150 different repositories. Three quarters thereof - including most of the general platforms - are located abroad.
Comparable with other countries
The results show that, overall, the behaviour of the Swiss research community and its reservations with regard to data sharing in public repositories are similar to other research communities around the world. What is more, the results of this survey confirm older studies with regard to the amount of shared data. Though this hardly comes as a surprise, it is useful knowledge nonetheless. International initiatives in the area of open science are therefore applicable to Swiss research, and Switzerland does not necessarily need to look for its own national solutions.
Funded largely by the public sector
The second part of the survey collected information from more than 200 repositories in Switzerland and abroad. Approximately 80 per cent are financed by public money. Only a few are based on other business models. The costs of setting up and maintaining the repositories vary greatly, depending on the requirements that need to be met in specific disciplines.
Developing guidelines
Open data is a basic principle of good academic practice. The SNSF strives to support researchers in dealing with this requirement. And it is fully aware of the challenges involved in the disclosure of data. Based on the findings of the new report, it will continue to develop its guidelines on open research data.
The survey was conducted on behalf of the SNSF and swissuniversities in 2018, and its final report is now available. The report is freely accessible, including all of the underlying survey data (see link).