2017 is shaping up to be the year that preprints in biomedical sciences go mainstream.
At the beginning of the year MRC and Wellcome Trust both moved to accept preprints in grant applications and scientific reviews. Another major UK biomedical funder is likely to follow suit. In the USA the NIH has recently done the same. The ASAPbio initiative has issued its call for proposals for a central preprint service and it will be interesting to see what comes out of that. Meanwhile the Chan Zuckerberg Initiative has made a major commitment to support bioRxiv financially. Several new discipline-specific preprint servers have been launched. All this and there’s still nearly 8 months of 2017 to go!
The more I think about and learn about preprints, the more questions that seem to crop up. Maybe you can help me find some answers.
Last year’s SPOTON conference on the future of peer review featured a session on preprints and peer review which I chaired and introduced. The discussion was interesting although it felt inconclusive. The recently-issued report from the conference has a piece by me: “The history of peer review, and looking forward to preprints in biomedicine” – now available as a post on the BMC blog. It’s nothing profound but I speculate slightly:
We may be moving to a world where some research is just published ‘as is’, and subject [only] to post-publication peer review, while other research goes through a more rigorous form of review including reproducibility checks. This [change] will be a gradual process, over a period of years. New tools … using artificial intelligence, will help by providing new ways to discover and filter the literature. A new set of research behaviours will emerge around reading, interpreting and responding to preprint literature. The corridors of science will resound with caveat lector and nullius in verba.
Whether / when to post?
Last week I gave a short talk about preprints to the postdocs at the MRC LMCB at UCL (here’s my slides). They were a lively group, already knowledgeable about preprints and open access but keen to learn more. I focused on a quick history of preprints and general stuff about them. The other speaker, Fiona Watts, provided real insight into the experience of a researcher who posts preprints, also drawing on her long experience as a senior editor on various top journals. She suggested that researchers should carefully consider which papers they preprint (is it a verb yet? I think it needs to be). Papers that have a heavy data component are appropriate for preprinting, but those with a high conceptual element might not be. I wonder if that is a generally-accepted notion, and whether there are other criteria that people use for deciding whether to post a preprint?
Scooping and citation
Related to this, the question of being scooped came up. Audience members shared their different experiences. Some believed it to be really uncommon, others reported having been scooped. In theory posting a preprint can give you priority over anything published subsequently. Some felt that would not be much comfort if someone else publishes a peer-reviewed paper before your preprint is published in a journal, particularly if the author of that paper fails to cite your preprint. Maybe peer reviewers need to be reminded that they should include preprints in their search for relevant literature?
There is some useful discussion of strategies about whether and when to post a preprint in a post on Niko Kriegeskorte’s blog.
One person at the LMCB meeting raised the question of preprint citeability. This had recently been discussed on Twitter (it has been usefully summarised here on the Jabberwocky blog). Apparently NAR does not permit preprints to be cited.
Are there other journals that have a similar policy?
Clearly preprint citation is necessary if we are to give appropriate credit/priority to work posted as a preprint. But I can understand why some may be wary of allowing citation of work that’s not been peer-reviewed. I’m not altogether persuaded by the argument that because physicists do it it must be OK for biologists to do it. There are differences between disciplines and their cultures of knowledge-sharing.
I would like to know how much difference there is between a preprint version and a finally-published version of an article. (Indeed, I’d also like to know how much difference there is between a ‘green’ version of an article and the version of record, but that’s another story).
In the comments on the Jabberwocky blogpost mentioned above, Martin Jung suggests that the key issue is “how to support and improve scientific peer-review to better deal with grey literature and non-published sources” (such as preprints). We need to think harder about this.
I’ve also been thinking about institutional policies around preprints. Judging by the ASAPbio page on University policies these are not yet widespread.
In talking to people about appropriate policies someone raised the idea of establishing an institutional preprint server. I can imagine a world where that would be a good idea, but I don’t think we are living in that world. I fear that an institutional server would be an irrelevant backwater, ignored by most in favour of the big disciplinary preprint servers. What do you think – is there much benefit to an institutional preprint server?
Two other useful things I read about preprints this week are a piece purporting to show increased citation rates for articles which started life as preprints, and a ten-point gospel highlighting the benefits and properties of preprints.
To recap my six questions:
- Is ‘preprint’ a verb yet?
- What criteria (if any) do you use for deciding whether to post a preprint?
- Should peer reviewers be told that they should include preprints in their search for relevant literature?
- Do you know of any journals which do not preprints to be cited as references?
- How much difference is there between a preprint version and a finally-published version of an article?
- Is there any benefit to an institutional preprint server?
Please give your answers in comments below, or on Twitter.