Digital repository audit method based on risk assessment

Transcription

There are significant risks with not managing research data effectively. Obviously there is the risk of data loss. When we're dealing with observational measurements of the weather we actually can't repeat those observations, when that information is lost, it's lost forever and that's clearly it's a loss to the individual project and it's a loss to society as a whole. There are also institutional reputational risks if data isn't available for verification, there is at least the suspicion or the one can't prove the integrity of the research that has been conducted. Complex and difficult experiments might have to be repeated because we can't make sense of the results we got from the last ones because we didn't document them effectively. The research enterprise as a whole produces huge amounts of research data. We're seeing in medical sciences 0:01:04.780,0:01:08.620 two publications, research publications being published uh... every minute of the day so there're huge amounts of information in the research community. Providing access to all of this information is absolutely critical if we want to accelerate the rate of progress for research and science. The British Antartic Survey uh... collects all types of environmental data and obviously the work in the climates and uh... sea level rise and things like that make it very important that we look after our research integrity and our data integrity. The term research data management has come up quite a lot recently and people tend to look at it as something, that's a new onus on researchers to consider and is a funding body mandate and I think it's important that researchers go back and realize that this is just part of good research practice. Data management is just about being able to verify your results, more importantly being able to make sure that those results can be shared with the right people at the right time and that people know how to cite it and how they can reuse it. Increasingly however there is an institutional responsibility to provide a supporting infrastructure and service to underpin good research and good research data management, and so institutions also run the risk of losing reputation and potentially even funding, if they don't provide um... that infrastructure to an acceptable standard. The problems you would have from researchers' point of view of not managing your data properly is that you would first of all find it very hard to find your data and combine it with other peoples' data. Two good reasons why researchers should manage their data effectively, one is so that they can find it themselves two years later down the line and the second reason is so that their colleagues around the world can find and use their data for new research to develop new discoveries. The sociocultural issues I think are the real problems right now. A lot of the technical issues are fairly easy to solve but changing sight of the culture is actually quite difficult. For me as a researcher it was that I didn't know what I could do with my data I didn't know... a lot of the people I know think that you can't do things when quite the opposite is true. Research Councils are encouraging users to share their data and people are thinking that they have to hold on to it and hide it and only publish it in normal traditional formats. Certainly one of the real challenges we face is the tension between the pressure to make more data more open earlier on and the real fear that many researchers have that if they do that others will reap the benefits from the hard work that they've done uh... that's a real fear we've got to acknowledge it, while also demonstrating the real value that we can get out of making data as open as possible, as early as possible. Researchers need to think seriously about how they're going to manage their data. It's not just that funders are going to require them to manage their data more carefully and make it reuseable but the researchers themselves will get more impact, their data will be cited their papers will be cited more often if they make it easier for other researchers to work with their data. Researchers need to be making all of their research objects available. By making their research data available on platforms such as Figshare, which have licensing rules which mean everything on Figshare is under CC Creative Commons Licensing this means that you have to attribute the researcher if you use their data. As opposed to being a loss of credit you can find yourself getting a lot more credit for your research, not only in citations in traditional publishing methods but also in other alt-metrics and what have you. The Digital Curation Centre has been a very important source nationally of expertise. Its international knowledge it's uh... developments in research disciplines, developments in research funders, developments at different institutions uh... and a good focus for that activity has been the Digital Curation Center as a place to go to find the information and the expertise. The DCC is a relatively small organization but despite that it can make a significant difference in the area of digital curation and research data management and indeed it has so. We've been running the DCC roadshows for one year and we visited a lot of places around the UK and people are now much more aware of the issues of data management and I think we can move forwards now and start to deliver more sophisticated, advanced workshops that take the DCC activities to a new space and particularly working with individual institutions on more specific strategic priorities. Given that research is very much international in scope, researchers are collaborating across international boundaries, there's a strong incentive for us to work with these other services so that we don't duplicate each others' work, so that we can share uh... and exploit what each other do, and also to encourage funders in multiple countries to work together in similar ways so that we don't need to develop 10 or 20 different tools to suit different national perspectives. That will save us all time it will save us all effort and it will make the process of actually doing the research far more straightforward We are a small organization, we cannot reach all the hundred thousand researches in the UK but if we can get more then we will have this domino effect, more of them will come to realize what's going on the more they will pass those standards and issues on. The DCC has also worked very effectively with UK universities there's the new uh... Data Management roadshows which have reached institutions which have been hitherto less engaged with these issues and that's been an important form of outreach, and the Digital Curation Centre's publications the 'How to' guides and other sources of of high quality material on these subjects have allowed the Curation Centre to reach an audience which might not have been expected for an organization of this size. The our scientist is really quite confused about what are the best ways to do things so if we could promote good standards and a fewer standards uh... simple approaches and then tools that make their jobs easier to do their work, then that would be a huge contribution to the data management movement. The Research Liaison Manager, the first place they might go to is the Digital Curation Centre to find what's already been done elsewhere and to try and uh... avoid reinventing wheels. My request would be for funders, for institutions to get the researchers early, get them on the first day of their PhD and tell them this is what you can do with your data, as well this is what you can't do but at least let them know so they can make their own decisions about what they want to do with it and uh... help encourage open data in that sense. The DCC can help researchers in a number of ways both directly by offering them tools and support that they can use while their research projects are going on and can also help them indirectly by supporting their funders and their institutions and the data centres to play their role in the management of research data. There are lots of things the DCC does to help institutions and individual researchers with research data management. We monitor research funders' data policies so we help raise awareness of specific requirements, we provide guidance on tools such as the DMP online, and in terms of actually supporting institutions to develop their own strategies in infrastructure for research data management, we are actually undertaking a programme of engagements at the moment where we're supporting specific institutions and providing tailored support. The DCC has a bundle of four main tools which help researchers and other data management stakeholders in the research data management process. The one which is probably the most useful for the researcher would be DMP online. DMP online helps researchers by asking them questions and making them think right at the start when they are applying for funding what they need to do to make their data accessible and usable in the long term, not just for them but for other people as well. The other three tools that the DCC offers that are of particular interest to research support staff and to data managers uh... in uh... in order in the order that you would probably use them, the Data Asset Famework or DAF helps you to identify the size and shape of your data related holdings. DRAMBORA is the digital repository audit method based on risk assessment, it helps you to identify and to communicate the risks which pertain to those assets which you've identified using DAF. Cardio helps to produce uh... maturity index uh... which uh... is based upon communication between all the different stakeholders in the research data management process, and what it does is it identifies areas of disconnect between, for example how the researchers may feel they manage things and how the research managers may feel that the researchers manage things and it helps to draw out uh... draw out these shared assumptions and help to communicate and and ideally find some middle ground. We are starting to see more awareness that data can take a number of different shapes and forms and uh... the drivers from funding bodies in the last few years has raised awareness but again it is coming back to this notion that data management is just part of good research practice, so I think there is an awareness starting to to filter through all the disciplines and that this is just the normal research practice activity. In future I think the Digital Curation Centre will have a very important role in coordinating expertise not just nationally but internationally in terms of data standards that people can use to make their data reusable uh... so I would see them as a kind of a focus point, a coordination point for international as well as national expertise in how you manage research data. We've seen a lot of researchers whether they're undergraduates or postgraduates or other careers researchers looking for training in data management uh... In the past few years because of the funding body mandates and demands for data management plans to be accompanied with bids, I think there is also a need to train peer reviewers so if we are making data available we have to make sure the people who review bids understand how to review data as well. There's been a change in the balance of the DCC developing new tools and doing its own research towards encouraging others and helping others to make use of the tools and the expertise that it's developed.