It’s about time that we took advantage of the full power of modern age digital technologies to innovate science publishing. Not just for the sake of innovation but for a much better cause: to address some of the core problems in scientific publishing such as irreproducibility, withheld data, the endless and excessive demands of reviewers, the shortcomings of the single blinded peer-review system, frequent rejections, slow publishing, and, foremost, the unreasonable need for sexy and cutting-edge scientific stories.
There have been numerous recent attempts to deal with some of these problems. However, one issue remains unaddressed: This is the pressure that the science publishing system exerts on scientists – to develop and submit flashy and fuller stories out of simple and sometimes powerful observations. Universities and funding agencies compound this pressure by judging and rewarding the ability of scientists to create and publish such stories in high-impact journals. Storytelling, thus, has become the prevailing paradigm of scientific publishing. While everyone loves a good story, the fact that we need to submit fuller stories (already during the initial submission) to get key scientific observations published is also accompanied by delayed reporting of major observations, non-reporting of ‘inconvenient’ facts and orphan observations and strong positive result-oriented publication bias. These systemic flaws, contribute, at least in part, to the alarming increase of incidences of irreproducibility and scientific misconduct. Not surprisingly, many scientists including Peter Higgs and Sydney Brenner, contend that current publishing practices discourage scientific exploration and affirm that their seminal studies would not have conformed to the current demands of science publishers and might well have stalled their careers.
To create some balance, we developed ScienceMatters, a new scientific publishing initiative that aims to redirect the focus of science publishing away from storytelling and back to the observation, a fundamental unit of science. Many of us have observations that matter and that are exciting – observations that we believe others should see and think about, observations that do not necessarily fit into a story, that might even be negative. Such single neglected, unpublished observations could be the missing pieces for numerous scientific puzzles but remain hidden because we do not have a story around them nor the means and resources to develop such a storyline. So what do we do? We file them, and hope to develop them some day into a story at some later date. However, all too frequently such observations are never published, because many a times the observation remains just that – an observation.
Not anymore. With ScienceMatters, there is now a space dedicated for publishing single, well validated observations. ScienceMatters eliminates the need to spin big stories from simple observations as there is no pressure to create stories, no reason to delay publication, no reason to omit ‘inconvenient truths’, a free path to publishing orphan and negative observations and direct experimental reproductions.
ScienceMatters is truly innovative and we briefly describe a few main features:
1. Single observation publishing: Rather than publishing stories, ScienceMatters publishes single, well-validated and properly controlled observations. The advantages of single-observation publishing are myriad. It eliminates the need for scientists to delay disclosing their results while they collect months or years worth of additional data in pursuit of a long, drawn-out story. It reduces the pressure on scientists to omit ‘inconvenient truths’ that might otherwise interfere with the story.
It also allows scientists to publish those intriguing observations that might otherwise lie forever unpublished, whether for lack of resources (personnel, funding, sometimes even intellectual) or because they cannot currently be explained. Most important of all, single observation publishing places the emphasis where it should be: on the observation and de-emphasizes the role of storytelling which often stretches science well beyond the data.
2. Triple-Blind Peer-review: Peer review is a cornerstone of science publishing. However, the prevailing practice of peer review is often biased, as editors and reviewers know the identity of the authors, and editors and reviewers know one another. This opens peer review to all manner of personal, professional, and social bias. ScienceMatters eliminates this problem by instituting – for the first time ever – triple-blind review, in which the authors’, editors’ and reviewers’ identities are unknown to all.
As a result, each observation is evaluated solely on its merits, without regard to the authors’ reputations, identities, institutions, gender, race, or nationality. ScienceMatters also improves the peer review process by scoring observations on a 0-10 scale rather than the non-quantitative and somewhat meaningless ‘thumbs-up, thumbs down’ approach of other journals.
These scores guide our publication decisions, with the very best observations published in Matters Select (score 8-10), all other technically sound observations published in Matters (score 4-8), and other “low-scored” observations (score <4) available on our Matters Archive website. Furthermore publishing the pre-publication peer review scores and our post publication peer-review process emphasize rapid and transparent communication of data. All of our journals adhere to the principles of Open Access, Open Data and Open Science and employ post-publication public/peer review, thus bringing the “public” into publication.
Real-time Continuum of Publishing: Once authors publish at Matters, they can return and extend their observations using our real-time continuum format. Others can also extend the authors’ original seed observation thus contributing to the natural and communal development of science. This should result in an honest and open picture of the science that is done, rather than the pruned or an embellished version that meets the demands and visions of authors, reviewers, editors and publishers.
Once a narrative emerges from a string of linked, independently reviewed observations, authors are welcome to publish their perspective on the scientific storyline that emerges from these linked observations in Matters Narratives, the review journal from ScienceMatters. So, you get what we mean with our tagline: Stories can wait. Matters Narratives will publish perspectives on storylines that emerge once the component observations are peer-reviewed and post-publication reviewed. We encourage different perspectives by different authors who have contributed to the emergent narrative, in the case that different authors have different interpretations of the scientific narrative.
A better metric for evaluation: If an observation is of import, it will have life and interest and should be made available for examination and use. Einstein’s early theoretical work on quantum phenomena and relatively are key to many modern-day applications, such as lasers or global positioning systems, that are far beyond what he and his contemporaries could envisage.
With the benefit of such hindsight, we aim not only to lower the barriers for publishing well-validated observations, but also to go beyond peer-review rankings and create community-interest metrics: how many scientists replicated the observation, how far was it extended, in how many directions was it extended, and how often was it linked to other observations? For this very reason, we developed metrics of an author or an observation based on reproducibility of the said work and connectivity of their data rather than the impact factor of the publishing journal, as it is done today. Our metric – MATTERIC – measures the impact of each observation: its connectivity.
Fairness to all: Resonating with the voice of many scientists, students, and the tax-payers who often fund the science that we do, ScienceMatters is Open Access. Authors deserve access to a journal that can guarantee a rapid decision and rapid publication of their observations at a low price. Our submission, article processing and open access publishing altogether cost is only $150 USD for academics, $300 for companies, and is (currently) free for all researchers from rural institutions in the developing world and we would like to keep this way, should funds be available for us to do so.
In addition, we pledge to review and render our first decision within 2 weeks of submission, and if acceptable, to publish it within the following week. Reviewers and editors who are scientists themselves also deserve fair treatment, for they do the real work of science publishing. For that reason, we pay our editors and reviewers half of the article processing charges ($75 USD divided equally between the editor and two reviewers).
Readers deserve access to the science that is published, and ScienceMatters offers free access to its content through its Open Access publication policy. By publishing raw data and encouraging sharing views and even reagents through our post-publication public review, we encourage Open Data and Open Science policies.
We have even more innovations in the pipeline for the coming months, which we believe will make publishing more enjoyable, fun and productive.
Today, 24 February 2016, we launch ScienceMatters with the publication of 13 novel observations in the area of biology and medicine, with 12 published in Matters, the flagship journal and one in Matters Select, the journal for outstanding observations. Future expansion of ScienceMatters into fields of physics, other natural sciences, and even the social sciences will be announced as we expand our editorial scope in these areas.
In summary, ScienceMatters will bring much desired speed, reproducibility, democracy and honesty to science publishing; accelerate the pace of research, and thereby improve the scientific enterprise. Join us in changing science by submitting your observations today.
It is an experiment, an exploration and an adventure. Be a part of this movement. Stories can wait. Science can’t.
Lawrence Rajendran Stephen Gould Stephen Curry John Ioannidis Volker Haucke Gisou van der Goot Anthony A. Hyman George Lake Graham Warren Kai Simons Rudolph E. Tanzi Satyajit Mayor Iva Tolic Michel Goedert Richard Hahnloser Magdalena Strzelecka Mirko Bischofberger Oliver Cooper Michael Hengartner Brian Nosek Paul Ayris Marcel Salathe Tom Südhof