PLoS ONE has a stringent and tough peer-review (although you are correct that “sexiness” and media-worthiness of manuscripts is not evaluated by reviewers).

Shortening embargos on publicly funded data is a good thing, but one has to be careful about picking the right length of time. In the astronomy community it is generally a year – which works out fine. In some areas, genetically modified crops – there is no openness at all and that has to change. Many scientists believe that practice, by definition, means it isn’t real science.

We have what starts out as essentially public good, going into a more traditional research setting which has been mostly rewarded through “publish or perish” reward systems.

How do we provide the optimum environment or context for the research, but then free the data and research results to address that desired public good, especially when most of the funding is in essense, public? What has to change?

Some of the issues that I’ve heard put on the table include:

• Better modes of data attribution so that scientists feel more comfortable releasing data and results earlier, more often and not letting data, especially raw data that they will likely never be able to analyse, out to others to use.
• Revise reward systems to reward not only peer reviewed publishing, but quality of data streams, uptake of data streams, etc (which can be significantly aided by social media and other technologies. I think the term social media may be significantly inadequate here.) I think the trust issue is very much part of this. But that also plays into the institutional cultures and the overall environment of international development which is very competitive because of funding constraints.
• Put time limits on the length of time raw data that is publicly funded can be withheld from wider distribution (i.e. after 2 years, all data funded by project X will b released, regardless of the state of the research publication, etc.)
• Continue to improve data sharing standards and practices to simply make it EASIER (which requires moving past a lot of “it’s not good enough if we didn’t design it” which has both factual basis, but also bias.)
• Measure and reward UPTAKE of research results. This puts the whole monitoring and evaluation practice on its head – as well as funding priorities. It may not ever wash.

What is reasonable? What might work? What might dramatically improve the landscape for international development based research? Any scientific research? Should there be a difference?

The sharing of information and eduction has changed dramatically. Citizen science is alive and growing!

PLoS is wonderful for the domains it covers. Much of this is their review process (which doesn’t have much to do with social media, but is well ahead of the conventional journals). Some of their journals are very accessible to the public.

Physics has arXiv.org, which has been running for over 20 years and is wonderful for document access.

Physics tends not to be accessible to the public as a fair amount of background is required to understand the papers. There is quite a bit of noise and confusion generated by people who don’t understand what is going on and/or have agendas (this is also true in the public discussions of climate change). It is an interesting problem and a big time waster for most scientists who get involved in these arguments. But it highlights other issues that need addressing – such as the poor level of science education in the US.

There are debates about how to make science accessible – some argue that putting it in the publics face – the gee whiz – is the way to go and others argue that teaching what is involved and how the scientific method works is more important. I’m more interested in the later. There is a lot of the former in social media, as well as old style broadcast media, now, but the potential for doing the later exists and needs to be exploited.

On open data and the real time open lab – it might work in some domains, but understanding failure in context with those you are working with is an important part of the learning process. Ultimately documenting that may be important if it isn’t a big time synch, but making the data and early playgrounds constantly available in an informative fashion strikes me as counterproductive. It would be like doing the same with art or learning an instrument or writing. Some bits are important, but there is a lot of noise if you don’t have the local context.