Open Book: Kick-starting open data support

Szabi Steiner describes efforts to encourage data sharing at City, University of London 

There has been rapid growth in interest in the potential of open data in recent years, and at City, University of London, we quickly recognised that if the university was to become more research intensive, we had to provide a more proactive research data service to our researchers. 

One of the steps that we took to encourage data sharing was to create an institutional data repository for non-traditional publications, using Figshare for Institutions (https://city.figshare.com). Like many other research institutions, City doesn’t have a long history of providing data support and for political and historical reasons City’s data repository sits outside the library in the research & enterprise office. I’m very much an ‘accidental data manager’, rather than a data specialist. I knew nothing about research data management two years ago, and it is only one of many responsibilities I have at the university.

Unfortunately uptake of the institutional repository was lower than anticipated. We found that researchers often have little experience of sharing data, and unlike traditional forms of publication (such as research articles and book chapters) data is not typically a priority for researchers, as it doesn’t count in the UK’s Research Excellence Framework (REF).

It’s not easy to convince someone to do something that they haven’t done before, especially when there is no immediate benefit to them. They ask “OK, what’s in it for me?”, and I’ll say “Well, you’re going to get a DOI and the data will be more visible”, and they just say “Oh I don’t care”. 

The low uptake was expected to continue without specialist data staff to offer more support to the researchers in their data submissions. Even when researchers see the importance of making their data discoverable, there are still significant practical challenges for them to overcome. They must identify an appropriate repository, organise and describe the data in a useful way, and then ensure adherence to copyright and licensing restrictions. All this takes time and knowledge that they often don’t have, and have little incentive to acquire. 

I wanted to hire someone to provide specialist data support, but I couldn’t get approval because I couldn’t demonstrate that there was a need for a person that does it in-house. I couldn’t predict the workload, and without a higher uptake of the data repository, it was difficult to demonstrate the need for specialist data staff. To help get around this impasse, we decided to partner with Springer Nature to provide all City researchers and doctoral students with access to its research data support services. 

Working with Springer Nature means that all I am required to do is carry out a preliminary check on the suitability of the data before assigning it to the data specialists in the Springer Nature research data team. This is done with just the click of a button from within the Figshare repository. Responsibility for helping the researchers is then passed to Springer Nature, who liaise with the researcher directly to provide one-to-one specialist research data support.  

The data specialists help with the creation of detailed metadata records, and ensure the accessibility and discoverability of the data, and its adherence to funder policy requirements. Where more information is required, they will request it from the researcher; where the data is out of the scope of the repository, because it contains mandated specialist data, they will provide guidance about specialist repositories to the researcher; where the submission contains sensitive data the research data team will either guide the researcher on anonymisation, or initiate controlled access. As well as the metadata record, the researcher receives a feedback report identifying enhancements that have been made to the data record, and the areas that could still be improved.

The process is interactive; data specialists work with the researcher to describe and present their data. By the end of the process, researchers can be assured that the data and metadata are published (or stored privately) with adherence to best practices and standards, and that they are as discoverable and accessible as possible. 

Outsourcing research data support has enabled us to provide our researchers with specialist data services that we just don’t have in-house. There’s a lot of knowledge and expertise required to do data curation properly, and without a science background it’s difficult for me to tell whether the data is reusable for someone else, how to write the metadata or even what different metadata standards there are. 

City now has the potential to scale-up its data support in a way that wasn’t possible while relying solely on our in-house resources. Interest in open research, open access, and open data is only going to increase, and it is not inconceivable that we will find by the next REF that open data is as necessary as open access is today. I have many other responsibilities at the university, and outsourcing data management support has allowed me to focus on these other responsibilities, while giving me the confidence that City’s researchers are provided with the specialist skills, knowledge and expertise they require.  

It’s too early to predict how great the impact of specialist data support will be on uptake of our data services, but it is already clear that the metadata at the end of the support process is often very different from the metadata initially submitted by the researchers. One dataset submitted recently was just a single Excel file with metadata stating ‘this is the data’. 

By the end of the support process it had been transformed with metadata, keywords, categories, and even the title of the final record, all being added by the research data team. The hope is that easing the submission of data will encourage researchers to use the repository in the near future, while improving the quality of the metadata will encourage use in the longer term, as the improved impact of the data begins to be seen. 

Szabi Steiner is research and enterprise operations and programme manager at City, University of London

• Do Szabi’s experiences chime with you? Would you like to share your thoughts as a librarian with the readers of Research Information? 
  Email: tim.gillett@europascience.com