News blog

Opening up scientific data

Should scientific data be shared more openly – and how would that help science? London’s Royal Society has launched a study, Science as a public enterprise: opening up scientific information, which it says “will look at how scientific information should best be managed to improve the quality of research and build public trust”.

That statement makes it clear that two separate trends are driving the inquiry. First, issues such as the public policy debate over climate change shows that “in a democracy that has lost many of its habits of deference to authority, science should become a more public enterprise,” explains Geoffrey Boulton, a professor emeritus of geology at the University of Edinburgh, who is chairing the working group undertaking the study. “Citizens should have access to information hitherto locked behind laboratory doors,” he says.

Second, says Boulton, is the change that has taken place in the practice of science itself, where “increasingly we have come to regard the published paper as almost an advertisement for the underlying science – the data.” The more rapidly that data – and the methods used to develop it – are made publicly available, the more readily other researchers can challenge it and add to it, potentially improving the efficiency of science, he adds.


The issue raises thorny risk-benefit questions, as Boulton and other working group members explain in an editorial in The Lancet [pdf]. “Is the potential for misuse, misinterpretation, and the triggering of spurious finding from data a price worth paying for greater openness?” they ask. Other tricky questions include how to make information more accessible, and who should pay to do it; whether privately-funded scientists should be held to the same standards as publicly-funded ones; how to cope with the need for confidentiality, data security, intellectual property rights, and anonymisation; and whether any rules on data sharing could apply globally.

Boulton says the working group are hoping to complete their study within the year, and that it should be released in early 2012. The Royal Society is welcoming online submissions (here) and, for those in London, will be holding a public meeting to discuss the issues at the South Bank Centre on 8 June.

Comments

  1. Report this comment

    Egon Willighagen said:

    The study may have interesting results, but I do hope the study will take into account that more openly available scientific data (Open Data?) is not just for the public, but important to science itself too. Boulton worries about misuse and misinterpretation, which is interesting as it was one of the reasons why people called for Open Data in the cited climate change debate. In fact, if data is openly available, misuse and misinterpretation is much more easily identified. It seems to me this this ‘thorny’ risk-benefit question is neither thorny nor risky, and the Boulton may have to look for more serious potential risks.

    His second tricky question is not tricky at all either. Open Data combined perfectly well with confidential data, and issues like data security, IP rights, etc, are more technical problems than problems resulting from openly available data. In fact, Open Data greatly simplifies these issues, making some even just obsolete. Moreover, these technical problems are already solved, though the uptake is limited for various non-technical reasons.

    All in all, it seems to me that it will be of utmost importance to this study to carefully separate goal and implementation. This post does not do so, in my humble opinion.

    That said, the UK has a rich community around Open Data, issues involved, and I am very confident that Prof. Boulton will find the right people to talk to, and to come to an accurate analysis, and I am very much looking forward to the outcome.

    In fact, maybe Boulton can set an example, and perform this study in an Open Notebook Science approach, and take advantage of the collective knowledge the Open Science community has built up in the past 10 years.

  2. Report this comment

    Mat Todd said:

    On the most general level, it’s very good news that these two important issues are being addressed. Yes, they are distinct. The arguments are complex and it will be useful to have them assembled and interrogated (openly, one hopes…). However, at heart there is a very simple idea. If one is interested in solving a problem, or answering a question, or understanding a system, then one needs to view all the available data, and it is beneficial to have as many people examine the data as possible. If that is an acceptable statement, then science is presumably more efficient when data are openly available to scrutiny by everyone. To my mind this prize – an increased pace of discovery – outweighs the dangers mentioned by Professor Boulton. At least that will be my submission to this call for inputs. Perhaps it would be useful if we could think of examples of how open science could be used to trigger “a spurious finding” when that finding can be refuted by someone else with access to the same data. Disagreement about the meaning of data leads to argument and discussion, and (ideally) spurious findings should not last long in that environment, but I’d be interested to hear of an example where open data might lead to unfortunate consequences.

  3. Report this comment

    Stephan Matthiesen said:

    Just a simple question: If the Royal Society is serious about “opening up scientific information”, why are their own journals still not completely open access?

    Individual articles as well as subscriptions of Royal Society journals are so expensive that interested lay people or journalists have realistically no chance to read them.

    Of course, the Royal Society makes articles accessible after a year (and this is laudable), but this is far too late for journalists or members of the public who want to read up on the background of the most recent “breakthroughs” or just generally keep up-to-date on scientific research.

    The pricing of Royal Society Journals is particularly dubious as the Royal Society – unlike commercial publishers – receives a lot of public funding for their work, so it is difficult to justify why the public should have to pay again when they actually want to read scientific papers.

    Making data accessible is, of course, an important thing, but realistically, every scientist knows that actually analysing datasets is a huge effort which most people will not be able to do.

    If you are serious about making science more accessible, then we need a strict policy that scientific papers are open access immediately from publication date – or at least available at a reasonable price (1-2 Dollars, not 25-30).

Comments are closed.