F1000 Research: a pre-print server?
| 7 February, 2012 | Rebecca Lawrence |
4
|
|
First, a big thank you for all the supportive comments on Twitter and in the blogosphere to our plans at F1000 Research. We are under no illusion this is going to be a tough project and need all the feedback you can offer!
Unsurprisingly, the greatest concerns have been around the idea of immediate publication followed by post-publication peer review. We have received a wealth of views – supportive comments and valid concerns – on this site and elsewhere, including Retraction Watch and at ArsTechnica, both of which have sparked long debates. As we will do when we start publishing, we have linked those debates to the original article so others that may not be aware can see the discussion going on elsewhere. We will revise our plans where necessary based on the outcome of these debates.
The first thing I would like to do is address some of the recurring themes in these discussions. This post will focus on the comparison some respondents have made between F1000 Research and pre-print servers such as ArXiv and Nature Precedings (or indeed our own F1000 Posters). I think this has ended up a slightly confusing analogy; F1000 Research will be much more than a pre-print server and in fact is more similar to traditional publishing approaches except for 2 major differences: immediate publication and open post-publication peer review.
- Immediate publication
- Open post-publication refereeing
Once an article is submitted, the article is made visible and published, i.e. priority stamped immediately with the time and date. It will then be clearly labelled as ‘submitted for refereeing’ and peer reviewers will be formally invited for all papers (just as on a traditional journal). This should encourage authors to think carefully about what stage they submit their work (just as they do now when submitting to a journal) to ensure they do not open themselves up to public criticism from the referees due to their work being too preliminary in nature.
We will be encouraging additional comments from registered users but we are certainly not going to be relying on those. As others have noted, uninvited commenting on the publisher’s site is often sporadic, although there is evidence that articles are increasingly discussed on Twitter and other social media outlets (just as with our recent announcement) and we will be connecting those comments to the original article.
As referee reports come in, these will be made immediately visible to all. The authors will be encouraged to engage in discussion with the referees (in the open) and to amend and improve their article, with all versions of the article stored and accessible. The peer review status of the article will be clearly displayed.
My next post will address some of the concerns we have heard about content quality. As always, we welcome your thoughts on everything we say here. If you want to discuss this issue on your own blog, then please do add a trackback here so others can follow the trail.
|
|
previous post
One major concern I have with F1000 is that it is a for-profit company. Quite aside from what it decides to charge for its services to start with, what guarantees do we have that if successful it won’t change its charging practices in the future?
This is a very real concern. If a large scale open access service becomes very successful and therefore influential and then decides to “cash in” on that success, it can be very difficult for the academic community to shift to an alternative model or service. We see this with the current difficulties in shifting away from traditional, very expensive but high reputation journals.
I believe that F1000 is owned by Science Navigation Group, which previously started a number of other journals and services including BioMed Central, Current Biology and others. They have sold these and similar journals/services to the likes of Thomson Reuters, Springer and Elsevier. What guarantees do we have that they won’t do the same with F1000 Research?
Why would academic researchers take the risk of investing in such a for-profit exercise if in the future it could be sold to exactly the type of publishing companies that we’re trying so desperately to escape?
The concern expressed by Tom Johnstone, which recurs from time to time, is a peculiar one, and one that we find hard to understand coming as it does from a scientist who should be able to think through the argument and see the nonsense of it. Is he really suggesting that all new ideas in publishing must come from people who do not want to make money?
He hints darkly at the possibility of our selling F1000 Research to a larger company, as happened with BioMed Central when it was sold (by us) to Springer. An often-expressed opinion in this context is that it would be better if a non-profit took on such a development, as if this would somehow be morally preferable. But Open Access publishing, which BioMed Central started (quite a long time before PLoS, though very much with the support of PLoS’s founders), was established against the strong opposition of many big commercial publishers, learned societies (some not-for-profit), and individual scientists. It required an investment of tens of millions of dollars, and persistence for close to ten years, to demonstrate that it is a sustainable business model – one that was subsequently adopted by PLoS and others – and the sale to Springer was seen by most active proponents of Open Access as confirmation of its long-term viability and a laudable success.
Publishing ideas will simply die, however beautiful they might be, if they do not find a sustainable business model. Not to mention of course that non-profit organisations (like PLoS) need to make money just as much as commercial organisations do if they want to survive without donations or taxpayer support.
We are entering into the F1000 Research project knowing that it will require a very high level of investment and has significant risks. It is the sale of BioMed Central that makes this possible. The benefits of our proposed scheme are significant and not difficult to see. If we succeed – that is, if we show that the F1000 Research model is sustainable – and if we then sell the company to a large institution capable of maintaining it long term, I hope Tom Johnstone will rejoice with us.
From what I can see, this sounds like potentially a very good idea. Although there is undoubtedly much to discuss about the technicalities of how this should work, I am more concerned with how to bring this publication model into the scientific mainstream.
In other words, I’d personally love to publish under this model – but how do I make sure that this does not compromise my research compared to publishing in a more conventional journal – not even Cell or Nature, but say, Molecular Cell Biology or Genome Biology.
As I’m sure you know, Biology Direct have been practising a somewhat similar publishing model for quite a while, and so far it seems to be working very well for theoretical studies, but not as much for the wet-lab fields. I’m not quite sure why this is the case (not sure that simply because the most reputable editorial board members are from that field), but this is still concerning.
It’s true that time is always needed to attract popularity to any media platform, especially in such a conservative field as science publishing.
But I believe that there’s a certain time window after which the media, if they fail to gain popularity, become firmly stuck on the margins. We celebrate PLoS ONE’s impact factor of 4, but we all know how biased this metric is. I cannot formally prove this, but I have the feeling that most of the highly-cited PLoS ONE papers would come from those narrow fields for which the middle-rank journal segment is nearly absent, and the choice is really between Nature/Science and some obscure highly specialised journal. In the larger fields such as genomics, immunology and developmental biology, publishing in PLoS ONE is unfortunately still viewed by many as a last resort (reserved, for example, for cases where that big competing paper is about to come out). And this is not just due to a given author’s conservatism: a PI may not want to do that because it would compromise the career of a junior first author, or reduce their own chances of getting a grant. It takes a combination of a very secure PI and a brave lead author to push good papers into experimental media.
It would be great if in addition to developing an exciting publishing model, F1000 Research could think of some specific ways to de-marginalise publications from the very start. An obvious suggestion would be to grant a near-automatic post-publication review by F1000, but I would argue against introducing this bias as it may potentially affect F1000’s own credibility based on the idea that the reviewers genuinely pick those papers they are interested in, theoretically without giving preference to any specific journal. And this is perhaps an aspect where F1000’s being a company rather than a group of enthusiasts, despite the obvious caveats, could help at this stage: they simply have more time to do the PR. I will be keeping a close eye on this project and hope it will take off. There already seems to be some emerging competition on this market, with openrepub.org also promising to launch this year, which I hope both projects will find equally stimulating.