|
Please use this page to leave feedback, suggestions or comments pertaining to the talk. Please use a new section for each thread of discussion and sign your contributions. Thank you.
- Show of hands:
- Who has ever edited any Wikipedia article?
- %
- Who has not used any Wikipedia this week?
- %
One's scientific reputation is determined in substantial part by one's publication history, and while the bigger funding sources may make efforts to account for that, the fact remains that labs with large published output are going to be more sustainable in funding terms than labs with smaller output.
Some time ago, I was part of a small lab that was preparing to publish, and we got scooped, by a larger, older, and better funded lab working in the same scientific area. It meant we had to re-imagine what our endpoint was and try and include the work we had already done, because it was a small/unique enough area that a strategy like finding a lower-tier journal to publish results that essentially recapitulated this higher-profile publication was not a viable option. It required a great deal more work and time, and funding became much tighter as a result.
So while I think the idea that we can micropublish our data as we produce it, while attractive in many ways, is inconsistent with this structure we have already built for the financing of science. That's not to say that it can't be changed, or that one is somehow morally better or worse than the other. I don't get the sense that it is catching on, and i think that it is because of these structural barriers.
Furthermore, I think there is a risk that in adopting this micropublishing/open science paradigm at this point, i.e. without effective structures to enable this communication-method-set, some innovative researchers will get burned (in terms of both delayed or reduced accrual of reputation, and consequently funding), who might otherwise have been able to contribute more substantially. I think this is of particular concern in hard- and wet-science fields (though i may have some bias, having worked in biology myself). It takes a lot of cash to do rodent experiments with enough statistical power to be useful, to say nothing of clinical trials.
Even if there was significant adoption of this Open paradigm, I fear that because there are insufficient feedback mechanisms right now, the tendency of funding or capital to flow to existing large concentrations of it would be accelerated in scientific funding. What keeps that tendency under control, to the extent that it can be said to be so, right now is that there are authorities making the explicit decision to fund researcher xyz with, e.g. a young investigator grant. This Open paradigm in some ways decentralizes the scientific endeavor, but if it remains in the context of more or less centralized funding decision making, there is a heightened risk of tragedy of the commons.
So, it seems like in order for this to succeed (I think to the extent that it will, it will be as an adjunct to the existing system, rather than a replacement), we need, as seems so often to be the case, some norm setting and feedback mechanisms. I havent the faintest idea what those will look like, but i want us (speaking broadly) to have the discussion. UseTheCommandLine (talk) 18:48, 16 July 2012 (UTC)[reply]
- It is very difficult for me to imagine a funding model in which commercially-appointed research could be open and remain financially viable. I am interested in that special case, and if there is a good proposal for how that would work, then I would like to read it.
- I like to think first of public research. One proposal that I would make is that all research funded by any government's money and intended for the public ought to be open. Another proposal that I would make is that large philanthropic foundations which fund research only disburse funds to researchers who would make their work open. I think that there are enough researchers who would be willing to receive funds for open projects that doing this would only result in good.
- I may also support government regulation to encourage the promotion of open science from large commercial corporations, but I do not know what proposals have been made for this.
- You talk about the fate of the small commercial lab - I think I might say that such organizations are my lowest priority for persuading to change and if they want to adopt a closed research model so that they can sustain themselves then I would not blame them and only thank them for doing science. I think that smaller labs would get more benefit from open science than larger labs, and if this is not the case, perhaps an end goal of the open science movement ought to be that new ideas from new entrants ought to be more encouraged than large labs which could somehow exploit all good open ideas for profit. Blue Rasberry (talk) 20:05, 16 July 2012 (UTC)[reply]
- I was speaking entirely of publicly-funded research, such as that which is funded by the NIH. Commercial drugmakers as a sector may spend more on R&D, but I was not including them in the discussion. Commercial funding sources are something i have no experience with. I should have made that clearer. Although these concerns may hold for commercial labs too, I can't speak to that. These same funding and limited-secrecy models that cause substantial problems in e.g. the patent system also cause problems in scientific research that is eventually made public.UseTheCommandLine (talk) 20:14, 16 July 2012 (UTC)[reply]
I watched this presentation with interest and I was struck by several things.
I do have years of experience with several different kinds of research and its funding; commercial R and D, private nonprofit R and D, government R and D (both open and proprietary), and academic R and D. Until the structure of science (and engineering and mathematics and related fields) is drastically changed, this sort of proposal for an open research model will be difficult to implement for just about any kind of research, in my opinion.
First, obviously, for commercial R and D, competitors can easily scoop a company and take advantage of openness to get to a viable product faster. After all, where did the patent system come from? It was to try to encourage openness but still allow the innovators to reap some rewards, although this often does not work out in practice. In fact, it is well known that corporations spend literally billions of dollars on corporate espionage, including bugging of buildings, airplanes, hotel rooms and bars to obtain just this sort of information. Any openness at all will just be exploited by others. If this model were imposed or enforced without some sort of guarantees, corporations would cease to do research.
In the case of government-funded R and D, often it is performed with the hope or promise of economic stimulation. This is particularly true of countries with smaller economies that can ill afford to waste resources. Again, it is well known that many governments are involved in economic espionage to try to funnel early research ideas and research results to their own industries.
For researchers at the start of their careers, this sort of proposal under our current system would be pure poison. Even now, many graduate students and others end up getting their research "ripped off" by the unscrupulous. Making things more "open" would just compound this problem. But these sorts of difficulties are not limited to those just starting out, because in an effort to get more productivity out of researchers, many organizations are pushing them harder, reducing job stability, and so on. Under these circumstances, "openness" is not likely to be particularly welcome.
In the case of openness about research accounts, I think this will be unlikely to work very well unless we have massive reforms as well. Already I know of plenty of institutions that keep 2 or 3 sets of books and engage in all sorts of malfeasance, co-mingling of funds, and so on. Often institutions and funding entities encourage this, or at least wink at it. Unless things are drastically changed, the "public" accounting for research funds would just be fraudulent window-dressing, and would bear no resemblance to the real financial state of the research enterprise.
Most of the problems that I heard from the first three speakers sounded as though current checks and balances were deficient. For example: Where was the reviewing? Where was the associate editor and the editor? Some pockets of science insist that the reviewers attach their names to the published article, so their reputations depend on them doing a proper job of reviewing. I also have seen reviewers repeat the analyses of the authors to make sure they can verify the work before publication, and even publish small addenda along with the original paper based on their findings.
Also, I am likely to have more confidence in a result if two separate teams compiled two independent data sets, and used two different sets of code implementing two different techniques to analyze the experiments or observations, and still reached the same conclusions. In fact, for serious medical research, I would prefer that this be the minimum standard. I would feel better with 3 teams and 4 teams, all doing independent work, of course. Just analyzing the same data with the same potentially buggy code means very little, to be honest.--Filll (talk | wpc) 23:47, 17 July 2012 (UTC)[reply]
|
