Difficult questions as register-based research becomes more open

Published: 2021-09-20

A domain specialist in their own words is a recurring series of interviews.Anders Brändström is a domain specialist in register-based data at SND. He has more than 40 years of experience from researching historical and present register data. In the A domain specialist in their own words series, he tells us more about a changing research field and tricky questions connected to sensitive personal data, official documents, and incentives for sharing data.

Anders Brändström is professor of historical demography at Umeå University. Over the years, he has researched topics such as aging, mortality, public health, and ethics in register-based research. Today he divides his time between being a domain specialist at SND, and researcher in a project about an aging population and one-person households, now and from a historical perspective. The types of data that the research project is based on are, as Anders Brändström puts it, “similar but very different.”

—Like all researchers in register-based research, I work with national public authority registers, and in this current project it’s mainly registers from Statistics Sweden and the National Board of Health and Welfare. But we also work quite a lot with historical data, as in data based on the information that priests used to collect in their parish registers. They may look like public registers, apart from the lack of personal identity numbers and that type of data. Then you link the information from the parish registers to birth registers, name, occupation, etc. So it’s two very different types of data, but in an analysis program, they will appear more or less the same, he says.

Openness soon became natural

Anders Brändström
 Anders Brändström is an SND domain specialist in register-based research.

As a domain specialist, Anders Brändström is used to working with and answering questions regarding open data. He tells us that there are very differing views in register-based research, depending on what types of data you use.

—Sharing data is traditionally something very rare for the historical sciences. Researchers often work alone and are sort of afraid that someone else will “steal” their hard-earned data before they have time to publish or process them for later articles or books. When you work with data from parish registers, it’s a different matter altogether. Those data are often collected by others. In our case, we work a lot with the Demographic Data Base in Umeå, which is an infrastructure with the specific purpose of registering data for the sake of research. And then, of course, different researchers can ask for the same data several times.

Anders gained a positive view to open data as early as his postgraduate years, when he had a tutor who had been denied to research on data that someone else had collected. 

—He taught me a great deal about openness: if others want to have a look at your material, then let them! It’s something good. You might say that I was raised into that way of thinking.

Many researchers ask questions about personal data in research data

Overall, Anders Brändström believes that the domain is developing towards greater openness. One example of this is the system MONA at Statistics Sweden. It is common to work in the system during the analysis phase of the research, and previously the data would have been destroyed when the research project ended. But in 2020 it was decided that the data will be stored and archived at the HEI after a finalized project. This means that the data become an official document, which the HEI can then release to other researchers and interested parties, after due assessment of the release request.

How the field is changing is also evident from the multitude of questions that Anders Brändström receives about openness and sensitive personal data.

—We have researchers in the social and historical sciences that work with interview surveys. There are cases where they do in-depth interviews with families or 10 to 20 individuals over a long time, and with some sensitive questions. As a researcher, you may have built a relationship based on deep trust with the informants. In those cases, they are terrified that their research will be impossible if they cannot promise absolute anonymity.

He says that legally, you cannot promise that the data will never be released, as that would be incompatible with the rule about public access to official documents, which all HEIs must follow.

—I think it’s reasonable that the data controller maintains a dialogue with the researcher, if they are still with the HEI, so that they can describe how sensitive the data are in an assessment of whether they can be released. But if someone wants to have a look at the material, has all the necessary permits, and will store the data in a secure way, it should be possible to release the data. A researcher cannot stop a data release that has been assessed by the data controller, and that rests on legal grounds. But I understand the concerns from those researchers.

The matter of academic merit needs to be solved

Anders Brändström highlights two major issues that need to be solved if we want open data to become the norm. One is to affect a cultural shift where we can see that openness leads to a scientific value, for yourself and others; that it can pave the way for collaborations and contacts across disciplines. The other concerns academic merit.

—I believe that if we are ever going to reach a situation where open research data are the norm, it must be meriting. It’s more than to receive a DOI [editor’s note: a type of persistent identifier] and that someone can cite the data, but something that can be used when you apply for or award academic positions and that has a clear value for your career. Today, it’s more accepted and valued in some scientific fields, but far from in all of them.

On 29 September, SND is hosting a digital Q&A, where Anders Brändström and other experts will be available to answer questions about how to manage and share historical and archaeological data.

Here you can read previous interviews in the A domain specialist in their own words series.