Exploring barriers to Data Availability Statements
This is a guest post from Mel Oakley, Project Analyst for Driving good practice in data sharing and use of data availability statements to support funder requirements, an InFrame project run jointly between the University of Edinburgh, the University of Glasgow, and the University of St Andrews. Mel will join us for an additional guest post about data availability statements in our Ds of Data series, as well as a final post with a summary of project findings. Here she reflects on the first stage of the project’s research and where the project is headed. Stay tuned for an update later this summer!
Our Wellcome-funded project Driving good practice in data sharing and use of data availability statements to support funder requirements investigates reasons why researchers don’t share low-risk research data and include a meaningful DAS in articles. The aim is to develop resources specifically addressing these barriers. These will be shared with the wider open research community, used as a catalyst for raising awareness around open data sharing and used to evidence training and support provision.
The first stage of the project focussed on desk-based analysis of the existing literature and the three existing datasets we found on the topic . From this we have been able to identify a number of barriers:
- Time
- Fear of data misuse or misunderstanding of the data
- Funding
- Technology
- Researcher insecurities or competitive concerns with sharing data
- Lack of knowledge or confidence in data sharing
- Recruitment and promotions prioritising publications over open access
We have also observed that science-based subjects share data and include a DAS in their outputs more commonly, whilst the Humanities and Social Sciences are less likely to share their data and include DAS in their publications. Further to this, the longer someone has been in research, the more senior their role, or the more likely it is that they’re a lead researcher on a project, the more likely they are to share data. Finally, the sources evidenced that when it comes to data sharing, people will pick whatever is easiest for them in the moment.
The next stage of this process will investigate which of the barriers listed above have been overcome with recent work in the area. For those barriers still in existence, we will look at what will help remove these and encourage researchers to share their data and include quality DASs in their publications. The outputs of this project will take into consideration some easy options available to researchers to share data and include high quality DASs in publications.
Thanks to Mel and the team for this contribution and we look forward to hearing further findings from the project.