blog

Who is ultimately responsible for the content of a research paper? Most of us would point at the authors: referees can assess what’s in the paper, but we have to trust that the authors present ‘real’ (not fabricated) data and that they report all results, not a cherry-picked selection. And yet, the expectation generally is that the Editor acts as a gate-keeper and is accountable for everything that is published in his or her journal – once the paper is published in a peer-reviewed journal the reported findings must be true (right?).

F1000Research gives the responsibility back to the authors and has taken the Editor pretty much out of the equation, but coming up with a clearly defined publishing framework is a challenge (more of this below). Can you, our authors and readers, help us define the borders?

There has been much debate about whether journals take their policing role too far, and concern that risk-averse editors rely on (often overworked and perhaps not always thorough) expert reviewers to find reasons to reject a paper, or to make authors do (sometimes a lot of) additional work before they concede to publishing the paper – often without adding much to the validity of the basic findings. We’ve all heard of research that was almost impossible to get published and completely disregarded when it first came out, that then turned out to be correct and defining a whole new area of research.

If we agree with Nobel Laureate Robert Horvitz that “…what is in the paper is fundamentally the responsibility of the authors, not of the reviewers” [1], how much policing should editors do?

By operating a post-publication peer-review model, F1000Research was set up as a publishing platform with the explicit aim of helping authors to share all of their findings (large and small) very quickly and without having to ‘get past’ an Editor. Authors are also closely involved in identifying independent expert reviewers, making it a truly ‘author-led’ publishing model that allows authors to take full and sole responsibility for the content of their paper.

Where does this leave F1000Research’s in-house team of Developmental Editors? We ensure that submitted papers are within F1000Research’s broad scope (which includes all life sciences) and that they meet certain ethical and editorial requirements, such as, for example, the availability of source data. We will point out to authors where we think they might run into trouble with the referees, but we deliberately make no editorial judgement on whether a paper is interesting. Whether the work is scientifically sound is left for invited expert reviewers to decide as they are best placed to assess whether the study design and methods used are appropriate and the conclusions supported by the findings, and whether sufficient data and detail are provided to allow others to repeat the experiment.

There is no question that F1000Research publishes papers that have flaws. Traditional journals publish such papers, too – as is evident from the growing number of retractions published each year [2].

Our job is not to aim for perfection, or to censor work, but to assist publication and ensure that the paper provides sufficient information and everything around its publication is transparent, so the scientific community can determine the work’s value and validity over time. By making the F1000Research peer-review process completely transparent and publishing the discourse between referees and authors, we allow readers to make up their own minds.

The lines remain fuzzy and the question of what is ‘proper’ science frequently causes discussion in our office: Most scientists and clinicians (and editors) will have a clear view on this, but strictly objective criteria on ‘good scientific practice’ that can be applied to ‘borderline papers’ are hard to establish. What about studies that are clearly statistically underpowered? What about experiments that have no controls? Should we publish these, knowing that they won’t get approved by the referees and make readers wonder about the lack of ‘quality control’ provided? What about topics that are at best ‘unconventional’, at worst considered to be ‘pseudoscience’ by many readers?

As in all other domains of life, ‘freedom of speech’ in science comes at a price. What price are our authors and readers willing to pay and where should we draw the line? Let us know what you think!

For the time being, we generally allow all researchers and clinicians to publish their research and express their views in F1000Research – as long as they are affiliated with a recognised institution and hence part of a community of ‘serious’ scientists and clinicians. Inevitably, we will sometimes feel uneasy about specific papers published in F1000Research, but we believe that as part of a transparent process, we should not veto what genuine scientists can say and share.

We rely on the research community to take back control over what is a truly valid contribution to the scientific literature – by commenting on articles and openly engaging with authors, in particular when you disagree with their approach or conclusions!

 

1. Robertson M (2009). What are journals for? J Biol, 8:1 doi:10.1186/jbiol111
2. Steen RG, Casadevall A, Fang FC (2013). Why has the number of scientific retractions increased? PLoS ONE, 8:e68397 doi: 10.1371/journal.pone.0068397