This month the open access journal PeerJ launched a new preprint server where scholars can post papers prior to submission for formal peer review and publication. Preprints are common in many disciplines, but have been unusual in the biology and biomedical areas that PeerJ focuses on. The culture of biomedicine and the academic overlap with highly competitive and potentially lucrative biotechnology and biopharm firms have retarded pre-publication release of results.
Pre-print servers are part of a growing trend. Over the last few years, the breadth of scholarly communication has begun to dramatically expand to support a life-cycle trajectory extending from the publication of small pieces of the research process in “nanopublications,” to the publication of pre-prints, and subsequently publications of record, often with post-print versions. With the launch of its preprint server, PeerJ hopes to capitalize on the growing comfort with pre-publication review and commentary that is increasingly accepted as a normal part of the publication lifecycle.
I was able to do a Q+A with the founders of PeerJ this last week, Pete Binfield and Jason Hoyt, to ask them more about their motivations.
PW: Why are you launching a pre-print server now?
PeerJ: Three reasons really:
Firstly, “Green Open Access” and the role of repositories are very important issues these days. The demand for Green OA is coming from both the top and bottom, and if you look at it, then the peer-reviewed portion of Green OA is covered by institutional repositories, but the ‘un peer-reviewed’ or draft versions of articles (i.e. the pre-prints) really have no major venues (at least not in the bio/medical sciences). So we view PeerJ PrePrints as one solution to that demand.
Secondly, academic journals themselves started out as non peer-reviewed venues for the rapid communication of results. Peer-review came about later on, evolving over centuries, to create something which has certainly introduced many positives for science. Still, ‘preprints’ also have many benefits that we no longer get to enjoy, because peer-review has come to dominate peoples attitudes towards what deserves to see the light of day. Now that more and more scientists are comfortable with the sharing attitude of the Internet (in part encouraged by the rise of Open Access), and as the costs of ‘preprinting’ are really quite low, it seemed like a good time to return to the roots of scholarly communication. Both peer-review and preprints have important roles to play in the ecosystem.
And thirdly, we believe that we are finally seeing a desire from the Bio and Medical communities for a service like this, but with no viable venue to meet that need. Just in the last year or two, we have seen biologists start to use the arXiv preprint server more (even though it really isn’t set up for their areas); we have seen services like FigShare and F1000 Research launch; and we have heard from many academics that they are eager to submit to something like this.
PW: In biomed, particularly, there has been a marked reluctance to pre-publish findings or early stages of papers because of the highly competitive nature of the domain. Do you think that is changing, or do you think you will attract a certain audience?
PeerJ: We do think that this is changing. It is common, of course, for early adopters to prove the value of a new way of doing things before the rest of a field will follow, and we believe that there is now a sufficient ‘critical mass’ of engaged academics who will use this service, to the extent that the rest of their communities will see what they are doing and give it a try as well. In this respect, we believe that earlier ‘failed experiments’ in the preprint space may have been simply (and unfortunately) too far ahead of their time to gain wide enough adoption.
In addition. although the default state for a PeerJ PrePrint will be to be ‘fully open’, future developments of the site will allow authors to apply ‘access’ and ‘privacy’ controls to create what we call ‘private preprints’. Specifically, in future authors will be able to limit the audience for a specific preprint (e.g. to just collaborators) or only make part of the preprint visible (e.g. just the abstract). In this way, we hope to make people comfortable enough to share to an extent that they might previously have been uncomfortable with.
PW: Researchers are increasingly publishing smaller bits of their research workflow, e.g., as data or even specific queries or lab runs. In some ways, a pre-pub server could be seen as a very conservative component of academic publishing. What do you regard as the “MVP” (Minimum Viable Product) for a pre-print publication?
PeerJ: We’re focusing on what we support best, long-form writing that is still a necessary step (for the time-being) on the road to a formally peer-reviewed publication. It’s a good fit for the lifecycle of a manuscript. Therefore, as a general rule, the MVP could be considered as “something which represents an early draft of a final journal submission”. On the other hand, there are no restrictions to the exact format used for these preprints, so we are actually hoping to see the use cases evolve due to an intentional lack of what is or isn’t allowed.
In terms of content, the only thing we don’t allow as a PrePrint for the moment are clinical trials or submissions which make therapeutic claims (as well as things which don’t fit within our Aims and Scope, or adhere to our Policies).
PW: How does pre-print fit into the economic model that peerj is running?
PeerJ: First it should be noted that authors who ‘preprint’ with PeerJ have no obligation to submit that item for formal peer-review – they can go to any journal that accepts preprints that have not been peer reviewed. Our membership plans allow for one public preprint per year for Free Members, and paying users can have unlimited public preprints. Paid memberships also have different levels of private preprints, but that isn’t available just yet. This is a similar model to several repository type services, such as GitHub, with a mix of public and private options. We expect private preprints to be attractive to those who want to test the waters of preprints, but restrict access to groups that they choose themselves.
PW: Are you requiring a CC-By license on pre-pub contributions? If so, do you think it discourages submissions from researchers who are sensitive to potential commercial gains from their work?
PeerJ: Any ‘public’ preprint will be published under a CC-BY license. The PeerJ Journal is also under the same license, and so if this license dissuades researchers, then they would also have been dissuaded from submitting a Journal article for the same reason.
PW: How do you imagine a work flowing within the PeerJ environment from a prepub status into an official publication?
PeerJ: The submission form that PeerJ PrePrint authors use is basically the same submission form that PeerJ Journal authors use (although there are some missing questions which are relevant to a preprint, and some fields are “Suggested” rather than “Required”). Because of this, it will be quite easy for an author to take their preprint submission and ‘convert’ it into a journal submission (they would simply have to supply a few extra bits of metadata and perhaps a fuller set of original files). Therefore, we expect a PeerJ PrePrint author to publish their preprint, get feedback, perhaps publish revisions etc, before then deciding it is ‘ready’ to be submitted to a Journal. If they choose PeerJ as their journal then it will be a simple matter to submit it to the journal for formal peer reviewed and eventual publication (assuming it passes peer review). If the preprint version had already gathered any informal comments then clearly those could also be used in the formal evaluation by the Journal’s Editors and Reviewers.
PW: How do you imagine a pre-print server generating additional traffic or “buy-in” for PeerJ? Will a pre-print server be able to increase the overall conversation that happens at peerj.com?
PeerJ: Our first focus is to make sure that we’ve built a service that researchers enjoy and engage with. We look at metrics such activity rates and engagement time as a barometer for whether what we are building is actually benefiting anyone. We are not worrying about traffic levels, but rather engagement levels. The traffic and new members will follow if we build something that researchers love.