Publishing a peer-reviewed scientific manuscript is considered to be the most important indicator of research productivity for scientists today. While peer-review is fairly effective at keeping the published science reliable and high quality, there are a number of concerns with our current scientific publication system. Some of these concerns, such as barriers to access for the public, were alleviated by the spread of online and open-access journals in the last decades; however, others were not. For example, the peer-review and publication process can be slow, taking several months or up to even a year, especially if manuscripts have to be evaluated by several sets of reviewers at different journals. Publishing in open-access journals can be expensive, potentially preventing authors who lack external funding from publishing their findings. Even though most scientific journals are no longer constrained by physical page limits, editors and peer reviewers still place a premium on studies presenting novel results, with a prejudice against studies that are mostly confirmatory. Finally, studies that demonstrate the lessons learned while failing to replicate results or test hypotheses (the so-called negative results) are difficult to publish, unless these results are very counter-intuitive.
All these inefficiencies lead to scientific work and findings to be published slowly, or go unpublished, relinquished to filing cabinets, drawers or their digital counterparts. While some of these research products are not of sufficient quality to merit publication, others constitute a loss for the scientific community. For example, a routine experience of researchers getting into a new field is that they can’t get a published protocol to work until a colleague shares with them a tiny trick that everyone else knows but no one has published. Another type of research that often goes unpublished are preliminary studies that constitute important supporting information for grant proposals or other time-sensitive purposes, unless incorporated into subsequent publications. Finally, while there are journals specifically designated for undergraduate research, most research conducted by students in class settings, and even as part of M.Sc. thesis, is lost to the scientific community. One reason for this is that the amount of effort, energy, time and expense required to publish such studies in peer-reviewed journals is considered to be too high relative to the probability of them being accepted in such journals.
Preprints constitute one potential solution to share preliminary, confirmatory or contradictory studies easier and faster with the scientific community. Preprints allow researchers to make their manuscripts publicly available on servers when they consider them ready for dissemination, without a formal editorial or peer-review process. Quality control is provided by subject-matter experts providing publicly available comments, potentially anonymously, or in an open peer-review format. Authors can incorporate such comments and suggestions into subsequent revisions of their manuscript, which they can update on the preprint server, creating a clear history of changes. This makes the peer-review process transparent, and can provide more thorough peer-review than the traditional peer-review process.
Preprints have a long history with scientists distributing their own papers yet-to-be-published at conferences. In fact, most scientific communications were conducted without a formal peer-review process before 1960s. In 1991, Paul Ginsparg at Los Alamos National Laboratories created arXiv to distribute theoretical high-energy physics preprints. In 2001, arXiv was moved to Cornell and expanded to physics, mathematics, computer science, quantitative biology and other fields. The Nature Publishing Group ran a pre-print server named Nature Precedings to host posters, manuscripts and unpublished observations in the biomedical sciences, chemistry and earth sciences between 2007 and 2012. In 2013, Cold Springs Harbor Laboratory started biorXiv to host preprints in the life sciences. Since then, a slew of different preprint servers have launched, across many fields, including for the medicine and health sciences, as well as regional servers such as AfricArxiv. While originally it was assumed that preprints will always be submitted to peer-reviewed publications eventually, this is no longer assumed to be the case.
While the acceptance and evaluation of preprints as trustworthy indicators of research productivity has recently increased, there are still remaining doubts of the value of preprints, especially in the medical sciences. In a recent publication, Sarabipour and her coauthors summarized the benefits of publishing preprints for early career researchers and made an attempt to dispel these lingering doubts. Benefits of preprints include accelerating scientific communication; facilitating career progression; increasing visibility; facilitating networking; optimizing research design and quality; allow publishing at low cost; boosting productivity; developing reviewer skills; and facilitate corrections via revisions. As a concrete example, funding agencies, such as the US National Institutes of Health and the Medical Research Council in the UK accept preprints as valid references to preliminary studies, which helps early career researchers apply for grants even before they had a chance to get their results published in peer-reviewed journals. In turn, several funding agencies, such as the Bill and Melinda Gates Foundation require grantees to submit preprints on their results before submitting peer-reviewed publications.
One of the concerns that researchers voice about preprints is the risk of being scooped. This scenario involves rival research groups with more resources repeating experiments published in preprints faster and publishing the same results in peer-reviewed journals before the original author of the preprint could publish. However, preprints have a timestamp associated with the submission which proves priority in case of a dispute over who produced the results first. Another area of concern has to do with patents. According to Patent Law, inventions that have been made publicly available before constitute prior art, which may invalidate patent claims on such inventions. Therefore, publishing preprints for marketable and potentially patentable inventions would not be advisable; however, the same goes for peer-reviewed publications.
Many researchers worry that publishing their results as preprints will preclude them from submitting their manuscript to peer-reviewed manuscripts. In the last couple of years, there has been a cultural shift towards accepting manuscripts already published as preprints among many publishers and individual journals. Submission to most publishers and journals, including BMC and Parasites and Vectors, is now compatible with preprints, with some of them even allowing one to directly submit their manuscript from biorXiv, or deposit a manuscript to a preprint server at the time of submission. It is now the exception, and not the rule, that a journal would not accept a manuscript that has been previously shared as a preprint; some of them have restrictions that bar authors from revising their preprints based on the formal peer-review process. However, by and large, we don’t have to worry about precluding publication in peer-reviewed journals following publication as a preprint.
Finally, authors worry that results they publish as preprints will have low visibility, relative to as if they published them in peer-reviewed journals. While peer-reviewed publications are generally more reliable, they do not necessarily have higher visibility. Preprint servers such as biorXiv are indexed by Google Scholar, Crossref, and other search engines, but not by e.g. Web of Science. Therefore, preprints are getting quite a lot of visibility.
Let me illustrate these points with my own experience. I published a preprint last summer on biorXiv on the habitat suitability of a new invasive tick in North America. I needed to develop a predictive map using simple methods quickly for a previous blog post, without a lot of specific expertise. Once I completed this task, I wanted to publish it somewhere so that my methods were clearly laid out for anyone to see. It would have taken way too long to try to publish in a peer-reviewed journal, and given my simplistic methodology, it might not have been good enough to get accepted. Therefore, publishing on biorXiv provided me with a platform to share my work in a much better way than putting it on a personal website. The process of publishing the preprint was fast, easy and free, and I could easily share and refer to the manuscript in my blog post. Over the next several months, as this new invasive tick was found at additional locations, I was able to quickly and easily update and revise my preprint. Along the way, my preprint got noticed by other researchers working on this same tick, and I was invited to participate in a USDA Working Group. In addition, when a peer-reviewed publication was finally published on the same topic with more sophisticated and reliable methodology, I was asked to provide my insight in a newspaper article, increasing my visibility. I also connected with the author of this new paper, increasing my professional network. While I ultimately decided not to submit my preprint for publication to a peer-reviewed journal, I don’t feel like I was scooped, and I’m still proud to share my humble attempt. Finally, as I’m coming up for tenure evaluations next year, I will include it in my tenure packet, along with the peer-reviewed publications that I’ve authored as well. Without the preprint server, this work would have just been lost on my computer to oblivion.
All in all, I wholeheartedly agree with Sarabipour and her coauthors on the value of preprints, especially for early career researchers, and I believe that they will become an even bigger part of our scientific discourse. Personally, I intend to keep submitting preprints on my results, and will encourage students in my lab and classes to do so as well. Do you agree? Let me know what you think! I’d love to hear your own experience!