The idea of open research data is part of the umbrella concept of open science. Open science involves making the research process as accessible and transparent as possible by sharing information and results on the internet.
“In one sense, open science is a development of the fundamental basis of science: the results of research must withstand scrutiny. In the 17th century, experiments were demonstrated before an audience, whilst today they are published on the internet”, says Christopher Kullenberg, researcher in science studies at the University of Gothenburg.
In the autumn’s research proposal, the Swedish government hopes that the transition to publishing Swedish research data and research publications online will be fully implemented within ten years.
This development is also supported by the EU’s research commissioner Carlos Moedas. In a speech in July, he talked about how open data is an important tool in combating ‘a post-factual era of democracy’.
Ten years of open science
Since 2006, free access to scientific publications – open access – has been on the agenda for Sweden’s Royal Library, higher education institutions and research financiers alike.
“It began with scientific publications, but open science also includes open data, civil research and open learning resources”, says Beate Eellend, supervisor at Sweden’s Royal Library.
Open research data involves using the internet to share the measurement data and the source code that lie behind the results presented in the scientific publication.
“Everything that answers the question ‘What makes this true?’”, clarifies Christopher Kullenberg.
Make research easier to review
One advantage is that it makes it easier to review research, both for researchers and for the public. Christopher Kullenberg thinks that this is a good idea as research is often produced under pressure, with the risk of mistakes.
“But it can be nerve-racking for researchers to show everything. Researchers make mistakes and choose the wrong methods but this is nothing to be ashamed of; we must move away from this idea of a genius conjuring up magic in a laboratory.”
At the same time, it places demands on journalists and the general public who are reading the results: they must scrutinise sources, be able to review the studies and assess the credibility of the researcher. There are some forums that are less serious than others.
“An undergrowth of false scientific journals has emerged with open access – predatory journals.”
One of the arguments that has fuelled the development of open data is that the results of publicly funded research must be made available to the public. Furthermore, this also encourages new research and innovation. Data can be re-used for new analyses and to develop new analytical tools.
Christopher Kullenberg refers to how the USA’s long tradition of open data from publicly funded research has laid the foundation for the majority of the major research databases used today, and it has also been of benefit to the technical development of various cloud-based services in Silicon Valley.
“Major platforms such as Google and Amazon earn money by sharing data. In this area, Europe is lagging behind the USA.”
The question is, who in the future will maintain and pay for the interface on the internet where European research data will be shared.
“The development of this infrastructure could be platform-driven – different cloud services from private companies. The disadvantage of private companies is that they want to lock-in users and earn money through advertising or by sharing user data”, says Christopher Kullenberg .
Alternatively, this sharing service could be created by means of national or European partnerships.
“The risk of the latter alternative is that, if the investment is insufficient, it can result in a technically poor solution.”
Klaus Lindgaard Høyer, professor at the Department of Public Health at the University of Copenhagen, also points out how the demands for open data and open access can distort competition. When there is an obligation for publicly funded research to be made freely available but the same rule does not apply to private companies, this will be disadvantageous for the publicly funded researchers.
The open access journal PLOS One, for example, requires that data be published at the same time as the article. This could lead to researchers delaying publication until they have finalised all their analyses.
“And then they run the risk of another group of researchers publishing before they do.”
Conflict with personal data protection
In the field of medical research, the crucial issue is the protection of people whose personal data is included in the research – for example, through interviews, DNA sequences and health data.
“Although there are many advantages to open data, there is also a fundamentally irreconcilable conflict between data that is fully open and data that is fully protected”, says Klaus Lindgaard Høyer, who also teaches ethics.
He also identifies a conflict in the relationship between doctor and patient where the purpose is to provide treatment for the patient.
“Your patients do not expect to see their gene sequences published online. Anonymising the data is no longer a guarantee.”
He points to the fact that data-mining robots are getting better and better at identifying people using information that is freely available online. Data mining uses powerful computer software to identify patterns and relationships within large sets of data.
“What was once anonymous, is anonymous no longer. Open data has many advantages, but there is a need for long-term planning concerning how the data is to be shared and protected.”
He cites the example of when researchers make good use of the Sequence Read Archive (SRA).
“They do not simply present everything, but they use the archive in order to show that they have the material for those who wish to collaborate.”
The rules regarding personal data protection in research are somewhat thorny, and they differ from country to country. In 1995, the EU issued a directive that concerns consent for participation in research studies but that also allows room for the different countries of the EU to make their own local exemptions. A new EU regulation will come into force in 2018.
The EU works to promote open science
On 27 May 2016, the member states of the EU decided to work to promote open science. Jean-Claude Burgelman, chair of the EU’s Open Science Task Force, does not, however, believe that there will be any new EU regulations concerning open data. Instead, the focus is on the realisation of initiatives such as the European Open Science Cloud and Copyright Reform.
Research projects financed by the EU’s Horizon 2020 research programme are required to make scientific articles available on the internet for free. The same requirements do not apply to research data. On the other hand, the ‘ORD pilot’ pilot scheme is being extended, where data is made available for certain projects. After 2017, this will cover all subject areas within Horizon 2020.
“When project applications are being assessed, however, participation in the pilot project does not constitute an advantage, and there is no disadvantage for non-participating applicants.”
Jean-Claude Burgelman also stresses that the option to decline to participate in the pilot project will always exist.
“The Commission understands that there may be very good reasons to refrain from making either all or part of the data freely available.”
At the same time, he would like to draw attention to the benefits of opening up the research process for transparency.
“With open science, we can benefit the quality of the research and innovation, as well as strengthening the links between researchers and the public. But, in order to begin this process, I think there is a need for radical changes to be made to researchers’ rewards and careers systems.”
Read also in Curie: