Sharing data

Sharing raw data is still a new concept for many researchers who are more used to using it as part of a finished product such as a book chapter or journal article. As part of the wider Open Research movement, data sharing is becoming increasingly common and researchers often need help to adequately prepare and distribute their information. This is an ideal place for library staff to step in and help.

Why do researchers need to share their data?

There are many reasons why researchers share their data and the particular balance is likely to be unique to the individual researcher. These can often be split into the philosophical arguments for good research practices and the incentives which mean that researchers must share their data - a carrot and a stick approach.

Researchers will respond to different arguments in different ways and librarians will find that they need to fine tune their response according to who they are talking to. The 'stick' approach tends to have more of an impact as researchers must comply or risk losing their funding or even their academic reputation. 

Concerns around data sharing

Many researchers have expressed concerns around sharing their data. It is important that librarians understand and respond to these concerns in case researchers are misinformed when seeking help. 

Researchers often worry about the possibility of being scooped by a rival research group when sharing raw data. This is particularly true in disciplines where much of the final output is down to the interpretation of the researcher rather than undisputable results. In fact, sharing data can help to protect against theft and plagiarism as it automatically date stamps anything which has been shared. Sharing data is yet another thing that already time pressured researchers have to think about during their project and they often see it as a burden rather than a natural part of the process. For many researchers sharing their information conflicts with the issue of sensitive data - how can they be expected to share information when it could identify their participants? Research funders are aware of this issue and will make allowances for personally or commercially sensitive data, usually in the form of an embargo which allows the researcher to restrict public access.

How is data shared?

There is no one simple way of sharing data which covers all funders and disciplines and best practice is to follow the specific guidance provided by each funder. There are however some general principles which researchers should follow to make sure that they have adequately fulfilled their obligations. 

Data should be:

• Stored for at least ten years. This is sometimes ten years from the creation date and sometimes ten years from the date of the last access to the data. 
• Described using in-depth and accurate metadata to help make it meaningful to others who may want to use it in their own work. This also helps to offer longevity to the data. 
• Stored in a suitable repository. This could be an institutional repository, a format specific repository or a general repository which accepts all types of data. Uploading information to a repository helps to preserve the data and offers a stable link for sharing. 
• Linked to from the main publication in order to tie the two together and make the data more discoverable for those seeking it. 

The concept of sharing data is still scary for many researchers and they are likely to have many questions about both 'why' and 'how' they should do this. Library staff may find that they need to be patient when explaining this and cover the same topic in a range of ways in order to make the message stick.