Prisoners of a system

Share this on social media:

Vitek Tracz describes his perspective on the changes in the way researchers have communicated their findings over the last 100 issues of Research Information

It is not clear what is more dangerous, to summarise the past or to predict the future. To do that on a 16-year time span in both directions may end badly. Still, to celebrate the 16 years of Research Information, I will try.

This anniversary coincides with the most significant moment in my life as a science publisher (more than 40 years). I will try here to give a description of the significant changes in the way researchers communicate their findings, and consume the findings of others, that have happened over the last 16 years. I will follow by presenting the momentous changes in this process I expect to happen in the next 16 years; changes that we at F1000 propose and experiment with now. 

There are, of course, many others inventing and proposing new ways in this area, and I admire many of them, but here I will concentrate on giving my personal perspective on where we have been and where we are going.

Outcomes of the research by millions of researchers worldwide is one of the greatest treasures of humanity. (UNESCO Science Report,https://en.unesco.org/node/252277, states that: ‘There were 7.8 million full-time equivalent researchers in 2013, representing growth of 21 per cent since 2007. Researchers accounted for 0.1 per cent of the global population.’ The benefit of all this research depends on the ability of researchers to communicate their findings effectively and responsibly, and the ability of others (researcher and anyone interested) to have easy, practical, full access to reports of research findings. 

I believe this process is badly flawed. Researchers are not free to communicate their findings, and there are a multitude of barriers to accessing the findings of others. As a result, I suspect that a significant proportion of published research is false and should not have been published, and a significant proportion of research findings that should be published are not. The main reason for this sorry state of affairs is that researchers are prisoners of a system of reporting findings that is almost completely dependent on research journals, and the decisions that journals and their editors make.

The reasons why this is so are (or should be) clear to all. The career of a researcher depends significantly on which journals they publish their findings in, because funding for research and research jobs are decided by the easy-to-use check on the Journal Impact Factor. This almost absolute dependence on research journals is the prison in which most researchers, as authors, are forced to live, and it profoundly distorts published research.

The life-changing role of the Impact Factor to a researcher’s career skews what is published, because articles are selected by journals based significantly on the potential for generating citations, the measure by which the Journal Impact Factor is calculated, and on which a journals’ success depends. Furthermore, access to the research published in research journals is commonly blocked to anyone who is not in a subscribing institution. The income of many major publishers of research journals depends on blocking access, and this creates both inequality and profound damage to the progress of research.

Until now: helping the readers

A little over 16 years ago, I was involved with a group (including Harold Varmus and David Lipman of NIH, and researchers Mike Eisen, Patrick Brown, Paul Ginsparg and others) in initiating an alternative publishing model now known as open access. The idea was simple: previously, the cost of publishing the research findings was collected from the reader (or rather the institution representing the reader – the library) at the end of the process, which was the only way possible when the only practical distribution was a printed journal. With the appearance of the World Wide Web, when one copy of an article could be accessed by all, we proposed that the costs be collected at the beginning of the process from the authors (or rather funders or institutions representing them) and make it available to all without restrictions.

The first step was the creation of the first open access publisher, BioMed Central by me, and the first repository for open access published research reports, PubMed Central by NIH (National Institutes of Health). BioMed Central started with more than 30 journals in every major subject in biology and medicine. These were the first Open Access journals, which editorially operated exactly like any other research journal, but the difference was all articles published where available immediately to all, free and without restriction. A special licence was developed that protected the intellectual rights and integrity of the publication, but they could be accessed and shared freely. 

The beginning was very slow and difficult. While many leading scientists supported the idea, it was hard to convince researchers to publish in new journals with no impact factors (it takes some years for a new journal to get an impact factor). Furthermore, publishers and societies whose profits depended on the income from subscriptions were trying to prevent the growth of the open access model. The major change happened when some major funders (NIH and Wellcome were the first) mandated that the research they fund had to be available freely to all after a year delay (this was in response to pressures from traditional publishers). But the benefits of open access to research and to society became increasingly clear over time. About two years after I established BioMed Central, the second open access publisher, PLOS, started – and, over time, many others. 

There seems little doubt to me that open access will become the universal mode to publish research, even though there is undoubtedly still some way to go, and there are many problems created by manipulating the scheme by publishers, big and small. The recently announced Plan S is a major step to improve the way open access works, by mandating sensible rules by a coalition of funders. 

The future: helping the authors

Over the years, I have become convinced that while open access has a real chance to solve the important issue of access to research findings, it does not address the problems research journals bring to the process of reporting research findings, whether open access or not. These problems are the results of the essential aspects of the operation of a research journal – the selection of what to publish and what to reject. This process, which seemed natural in the early days of printed journals, now dominates and distorts the way research is published. This is due to the immense growth of research outputs, the growth of the publishing industry (about 30,000 research journals support an industry with profits of billions of dollars) and the intense competition between journals, where success is significantly dependant on the highly problematic measure of the Impact Factor. 

These problems include, among others:

  • The delay in making new findings known, a problem now being partially addressed by the growth in preprint servers;
  • The opaque process journals operate to decide what to accept and what to reject;
  • The influence that the desire of journals to increase their impact factor has on the decision on what to publish; and
  • The range of rules and formats that different journals operate.

We have over the last six years developed a new publishing process, which we call open research publishing. We believe that this process, which completely bypasses the research journals, has a real chance to remove many of the problems currently damaging and distorting the publishing of research. The objective of these developments is three-fold: to pass the control of the process from research journals and their editors to the researchers themselves; to simplify and speed up the availability of new findings; and to introduce as much transparency to the process as possible. 

 To aid understanding of what it is we have proposed to replace the scheme operated by research journals, I have detailed a broad description of the steps of the process we have developed. The basic aim is to allow every researcher to report findings they feel should be reported without delay, and then submit themselves to a full and transparent scrutiny by peers after publication.

How open research publishing works

Researchers who want to report any findings can place them on an open research publishing platform. The decision to publish is in the hands of the researcher, and will be published immediately, after a check that it is research, by a researcher, it is legal, readable, and that the data underpinning the findings are made accessible. The publication will be fully citeable and available on the open research platform and indexed by search engines etc. This is similar (though the checks are more thorough) to a preprint scheme. Unlike a preprint, an article on an open research platform cannot be published elsewhere. 

The initial publication includes in the citation 'Version 1, awaiting peer review'. The peer review status will always be visible in the article citation. There is no editorial decision to accept or reject the submission. All submissions that pass the checks are published in a few days.

The moment it is published, expert reviewers are invited for transparent peer review. The authors are given a list by the platform of appropriate peer reviewers generated algorithmically. Authors and reviewers have to declare conflict of interests. Authors can suggest additional reviewers, whose appropriateness is checked by the operator of the platform.

Reviewers who respond to the request from authors (transmitted by the operator of the platform) make a general determination of acceptance (approved, approved with reservations, not approved) and write a reviewer report. This is published immediately alongside the article, with the names of the reviewers and their affiliations.

Authors are free to respond openly to the reviewer comments and to write a new version of the article. The new version is published above the previous version, with a summary of the changes. The peer review status is amended to (for example) “version 2, one approved, one approved with reservations”. This process continues for as long as the authors wish to continue. 

The advantages of speed and lack of editorial bias from the preprint-like component are combined into one simple and rapid process with the independent formal peer review, archiving and indexing components from the traditional journal model. We also work with authors to ensure that the source data underpinning any published findings are made available to support reuse and reproducibility. 

This model makes it possible and desirable to bypass traditional publishing, and the associated problems of delay, non-transparent selection, and high costs to the research community. It has already been adopted experimentally by some major funders and organisations such as Wellcome, the Gates Foundation and many others, and I believe will become broadly adopted and, in time, mandated by funders, societies and organisations. If this happens, it will finally free researchers to communicate any findings they think are worth reporting, while having to submit themselves to transparent scrutiny by peers. The great treasure our societies have – the results of research efforts by millions of researchers – will finally be able to flow freely, available to all.  

Vitek Tracz is founder and chairman of the F1000 Group