Recommendation 5: All researchers should make research data, methods, and other information integral to their publicly reported results publicly accessible in a timely manner to allow verification of published findings and to enable other researchers to build on published results, except in unusual cases in which there are compelling reasons for not releasing data. In these cases, researchers should explain in a publicly accessible manner why the data are being withheld from release.
This principle may seem to apply only to publicly funded research, but a strong case can be made that much data from privately funded research should be made publicly available as well. Making such data available can produce societal benefits while also preserving the commercial opportunities that motivated the research. As discussed earlier, differences in technological infrastructure, publication practices, data-sharing expectations, and other cultural practices have long existed between research fields. In some fields, aspects of this “data culture” act as barriers to access and sharing of data. With the growing importance of research results to certain areas of public policy, the rapid increase of interdisciplinary research that involves integration of data from different disciplines, and other trends, it is important for fields of research to examine their standards and practices regarding data and to make these explicit.
Data accessibility standards generally depend on the norms of scholarly communication within a field. In many fields these norms are now in a state of flux. In some fields, researchers may be expected to disseminate data and conclusions more rapidly than is possible through peer-reviewed publications. Digital technologies are providing new ways to disseminate research results—for example, by making it possible to post draft papers on archival sites or by employing software packages, databases, blogs, or other communications on personal or institutional Web sites. Data sharing is greatly facilitated when a field of research has standards and institutions in place that are designed to promote the accessibility of data.
02 October 2009
NAS on Release of Data
A reader (Thanks JK!) just point me to this recent NAS report, Ensuring the Integrity, Accessibility, and Stewardship of Research Data in the Digital Age, and in particular its comments on release of data underlying scientific studies. It clearly places the onus for justifying why data should not be released on the researcher, calling instances of non-release "unusual cases."
10 comments:
If a study cannot be audited or replicated, it should never, ever be used by policy makers. This is true, even if the study was not publicly funded. Actually, no study should be used in making policy until it has been replicated by others (who have no conflicts).
This may not be the way science is currently done, but it most definitely is the way public policy should be done.
What is your policy on releasing to a third party of data that has been given to you for your use by its owners.
Eli would suggest that if you tried it you would get cut off at the knees.
The Russians owned the data, not Briffa. It was theirs to release or not.
Now the NAS just needs to establish a police force to get climate scientists who behave badly to toe the line.
Any volunteers in the meanwhile?
I vote for Roger and Michael Tobis.
EliRabett,
The responsible policy on data which can not be released is this:
1. Don't publish anything based on it. Briffa should not have published his 2000 hockey stick paper (or any of the followups based on that data).
2. If you do publish, and your integrity is attacked, and you need to produce the missing data to protect your integrity do one of two things:
A. Get permission to release the data
B. Retire
C. Release the data without permission
3. You seem to suggest elsewhere that releasing the data without permission would end Briffa's career would end it.
I doubt this very much. Maybe, people who control restricted data will refuse to give it to Briffa. To this I say, Good!
As stated in point #1, Briffa shouldn't be publishing papers based on hidden data anyway. Let him role up his sleeves and collect cores that he and his coauthors can personally vouch for.
Throughout this episode I have repeatedly though about the amounts of these grants and wondered why such a small portion of the research dollars is being spent on data collection. Its time for that to change.
Climatologists should be spending their time collecting climate data, not inventing new statistical methods, which wouldn't even pass peer review in a statistics journal, or coming up with new inventive reasons why they can not disclose the data.
It is a waste of US and UK tax payer money.
In rare circumstances it may be necessary to publish papers based on "secret" data, but in that case the secrecy of the data should be declared up front. The rest of us can then decide how seriously we want to take it; indeed the journals can decide whether they want to touch it.
What is absolutely not acceptable is using secret data without making this completely clear.
There is a significant difference between secret and private. Why are you guys such data socialists?
EliRabett, I have no problem with you keeping data private; that is your right. But you can't reasonably expect me to take any actions based on your private data, whether that action is changing my beliefs on a scientific question, or changing my carbon footprint.
If you want to have a public effect then make your data and methods public.
Jonathan, this is the nonsense that the "Data Quality Act" was designed to push. You can basically nit pick anything. There is a huge amount of proxy data out there for you to look at. You don't need any one of them to see the general hockey stick shape. McIntyre was clearly wrong about Yamal being necessary.
What is now perfectly clear is that McIntyre asked people who didn't own the data to hand it over to him and got the back of their hand. He copped to it and continues to ignore the huge hint that Science gave him. Think who had and owned the "measurement data".
EliRabett, if there is really a huge amount of proxy data available then I suggest you stick to the fraction of it which is public and stop trying to fob me off with secret data.
I would remind you that most serious journals require that data be made available on request. It's no good suddenly announcing that you can't actually fulfil your obligations after the paper is published.
Oh, and how did "nullius in verba" suddenly move from being the motto of the Royal Society to a supposed sign of being a "denier"?
Post a Comment