Data without Boundaries offered support for transnational access as a unique opportunity for EU researchers to work with European microdata. There were eight calls for research proposals in 2012–2014. 40 out of over 60 proposals received support.
In the 2nd European Data Access Forum Karen Dennison had a presentation about the transnational access. Her presentation »Research Data Centre network for transnational access - four years of experiences by seven European RDCs can be found from the 2nd European Data Access Forum »web page under the Day 1, Session 2.
Three of the selected researchers gave us their reflections about the support. They were asked to give their views on good sides and drawbacks alike. Also two Research Data Centre representatives gave their feedback of the experiment.
[ The experiences of the participating researchers]
[ The experiences of the Research Data Centres]
[ Hans Koster, VU University Amsterdam]
[Carolina V. Zuccotti, European University Institute, University of Brighton]
[Thijs Bol, University of Amsterdam]
[Andrea Lengerer, German Microdata Lab at GESIS, Leibniz Institute for the Social Sciences]
[Karen Dennison, The UK Data Service]
Hans Koster, Ph.D. Assistant Professor
Department of Spatial Economics
VU University Amsterdam
» More information on Koster's own website
- Job Mobility and Productivity: Are Job-Hoppers More Productive? Vu University Amsterdam, Netherlands
- Land Use Planning, Agglomeration Economies and the Office Market: Evidence From Blitz Bombings, Vu University Amsterdam, Netherlands
Koster has been involved in two projects that have received DwB support. In both cases, he has used UK Data Archive Secure Lab data. He has been working in two universities, the LSE and Oxford.
“It took three months in both projects to actually get to access the data after the application was submitted”, estimates Koster. “My co-authors are experienced, which saves time, but there’s always some hassle. A lot of forms has to be filled in, and then something is missing. – It has to do with local bureaucracy more than the data research centre per se”, says Koster.
The first project focused on what determines how often people change jobs and how changes in infrastructure affect the labour market. Unfortunately, according to Koster, they did not yet find much in the data they accessed.
The second project is about effects of land use planning in London on firm location choices. “This one is ongoing. Preliminary findings seem quite promising, though.”
It is unfortunate that we did not find much in the first project. In the second, there are no publications yet either, as we just started last year. It takes at least two to three years to publish.
Koster has stayed two two-week periods in Oxford. In the first project he stayed four weeks at the London School of Economics.
Koster had some earlier experience in use of microdata on firms in the Netherlands, but according to him it was not as secure as the data in the UK. “To be able to use this data I had to take a training day in the UK to gain access.”
Koster's project team had the research problem in mind and they were looking for funding, when they learned about the DwB support and applied.
There is one major problem that Koster mentions, when asked for criticism. “The funding is only for a very short period and the projects take long time. Two to four weeks is short time. In that time you maybe just get acquainted with the data and are able to see what the possibilities are. It is not enough to finish anything.” Another more minor criticism, Koster has, deals with the reimbursement of the costs. It is a little bit inefficient as it involves filling in forms that are send back and forth by mail.
The best thing about the support provided by the Data without Boundaries project was, according to Koster, simply the funding for travel and that it enabled him to work on the microdata he wanted.
“Without the DwB support, I would not say it would have been impossible, but it makes it a lot easier.”
For any continuation of support for data access, Koster has one recommendation for improvement.
“There should be a possibility for a bit longer visit to a research centre, three months would likely be enough”, estimates Koster. Another improvement could be that the local statistical offices provide access to data in other countries. Less financial support for travel and accommodation costs would be needed then.
Carolina V. Zuccotti, PhD candidate, Department of Political and Social Sciences, European University Institute
Research Fellow, Brighton Business School, University of Brighton
- Carolina V. Zuccotti Moving Up, Moving Away? Ethicity, Neighborhoods and Social Mobility in the UK (1991–2011)
Carolina Zuccotti visited the Virtual Microdata Laboratory (VML) of the UK Office for National Statistics (ONS) in London. She happened to live in London during the time and got her meals and transportation costs paid by the DwB on the days she visited the Virtual Microdata Laboratory.
Because of the special situation, she was able to stretch her access period from the normal two – three weeks at a time to over 12 months. During that time she visited the laboratory three to seven times each month.
Zuccotti was also lucky to get involved in a project, which was testing the ONS longitudinal studies data. Professor Lucinda Platt made her part of the LS Beta Test project.
“This is how I accessed the data, and especially, data that was still not released to the public. I think I accessed the data more or less at the same time I got the grant”, Zuccotti explains.
Zuccotti is sociologist from the University of Buenos Aires originally. She has carried out two masters in the field of urban studies at the University of Urbino and the University of Amsterdam. Now she has submitted her Ph.D. thesis (Shaping Ethnic Inequalities: The Production and Reproduction of Social and Spatial Inequalities among Ethnic Minorities in England and Wales) at the European University Institute, and since September 2014 is research fellow at the University of Brighton.
“My main research areas are social inequality and social mobility, migration and ethnicity, and all the processes that connect individuals to the space, such as spatial segregation, neighbourhood effects and residential mobility. Using quantitative methods, I have carried out research on these topics in my own country, Argentina, and in various European countries. In particular, my Ph.D. thesis – most of it based on VML data – and my current work at the University of Brighton deal with these topics in the UK”, tells Zuccotti.
Zuccotti did not have any experience of this kind of microdata before. She did have the research problem in mind before learning about the DwB support. She, however, tells that when she started actually working in Virtual Microdata Laboratory she learned new things about the data. “In particular, although I had a vague idea that I could link neighbourhood data to individuals, only when I started working in VML I realized the potential of this link. This led me to new research questions and a new chapter!”
Zuccotti’s experiences in the Virtual Microdata Laboratory were positive overall. “I can say that people in VML were extremely helpful, qualified and always available to answer questions. Regarding the money, it was very useful for the daily expenses.” As she was part of the LS Beta Test project, she got support from them in using the data.
In Zuccotti’s case, though, the financial support was not of major importance, as she had decided to move to London in any case. However, she says, the Virtual Microdata Laboratory support was crucial for being able to conduct her research.
When asked for critique, Zuccotti names four issues:
1) The actual physical space to work is not very nice [in VML]. It's a small office with many computers in it. 2) Also, the program to analyse data (I used STATA) is old and extremely slow. In fact, this had an actual impact on the models I used. If I had had a newer and faster computer and programs, I would have used different models (even if the conclusions would have been the same). 3) Sometimes the VML was not working. So if someone goes for a certain period of time, this might bring problems. 4) Finally, some data restrictions are overstated, I believe.
Thijs Bol, PhD Assistant professor in Sociology
University of Amsterdam
» More information on Bol's own website
- Patterns of labour market entry in closed occupations. Occupational closure and the school-to-work transition in Germany and the UK.
Thijs Bol visited the GESIS Microdata lab in Mannheim, Germany. He had a three-day visit in December 2012.
Bol does not remember exactly how long it took from application to getting access to the data. He says it did not take very long, since he already had experience in working with the microdata. “I stalled going to Germany for a while because of my dissertation.”
Bol had already worked on the German Microcensus data for his PhD and visited GESIS in 2011. During his first visit to GESIS he also thought about the research problem he research on DwB support. After that visit he applied to DwB.
The best thing about data access support by DwB was that it took away financial barriers to work with foreign data. "Instead of trying to get money from your own department, you are able to work with data that you think is interesting.”
“I think the next step forward should really be to get remote access. Instead of investing all the money in travel costs for scholars, to me it makes much more sense to set up remote access workplaces at universities”, say Bol when asked for critique. “This is also a more durable solution to the problem with data access. The analyses could then be done on the remote server in, in my case, Germany, but I would be able to work on that remote server from Amsterdam. The technology to do this ensures security of personal data.”
When asked if he would have been able to conduct this kind of research without the DwB support, Bol thinks it unlikely.
“I would’ve needed money from my department, which, as a postdoc (at that time), would be quite hard.”
“I think it’s a very good initiative, that takes away the unnecessary barriers for scientific research”, concludes Bol.
Andrea Lengerer, Senior Scientist
German Microdata Lab at GESIS, Leibniz Institute for the Social Sciences
The experience of the Research Data Centres is similar to the participating researchers, to an extent. Andrea Lengerer from GESIS says they expected higher demand for transnational access. In their experience the financial support that was available for travel and subsistence was not a concluding factor.
“We believe that transnational access is a relatively complicated and inconvenient way of data access for researchers. A few days or even weeks of data access is a very limited amount of time for analysing the data. Research is always a process, a process of exploring a question, the data, ways of analysing the data and so on. It is very seldom the case that researchers have a definite research question with a finalised strategy of data analysis, before they have even seen the data. It’s rather a process of looking into the data, of refining the research question, of analysing and of reanalysing the data. A long-term data access – especially with Scientific Use Files – is much more suitable for that”, Lengerer summarises.
In GESIS’ experience, the language barrier is high and documentation in English is essential. Thus, GESIS is translating their data documentation into English, including the variable and value labels.
There is a relatively high dropout rate among the approved projects. There are cases where everything has been organised by the hosting Research Data Centre, and the researcher cancels the stay. A lot of work for the host has gone for nothing, which causes frustration.
Karen Dennison, Collections Development Manager
The UK Data Service
"Being part of the project has been a very positive experience and I feel that the project has enabled real research that might not have otherwise happened and has made more researchers aware of the Research Data Centres and the possibility to use their data", says Karen Dennison from the UKDA.
The main difference regarding expectations was that there were fewer applicants than estimated originally but with more access days per project. As in the case of GESIS, Dennison also says that it was frustrating that so many of the selected projects got to a certain stage of the process and then dropped out for whatever reason.
Dennison also points out some lessons learned that are in line with the GESIS and participants’ experiences. “Whilst funding for travel was useful a proper research grant would be better that would cover more access days. In future, applications should not be restricted to those resident in the EU – we had to turn away applicants”, she says.
"For these sorts of transnational projects, we need a centralised application that fulfils each of the national accreditation processes. We know that remote access is preferable and need to work towards this being possible [towards European Remote Access Network].”
The UK Data Archive is considering continuing to allow researchers outside the UK to apply for access to their Secure Lab data but the successful applicants would still need to physically go to the UK to use most of the data to be subject to UK law.