Daniel Marovitz 'falou e disse': publicações com revisão por pares são uma ameaça à ciência!!!

sexta-feira, novembro 28, 2014

Science Journals Have Passed Their Expiration Date -- It's Time for the Publishing Platform

Posted by Daniel Marovitz on November 13, 2014 at 12:00pm

Technology has helped so many industries evolve over the past few decades, but scientific publishing, surprisingly, has hardly changed since the first journal article in 1665. They must; their day is done. Their continuing existence damages science.

Several factors create this situation:

The use of an arbiter (Editor) in science – There is nobody qualified to do the job. How can that be? Isn’t science publishing full of editors who evaluate articles to see whether they are worthy of inclusion under a prestigious name? The problem is there is NOBODY up to the job. How can someone have enough knowledge to understand that a newly described insight is big or small, impactful or irrelevant, correct or incorrect? History is full of examples of “great” papers that turned out to be wrong, while many papers describing huge leaps forward had trouble finding anyone to publish them. Editors must not try to decide, because in trying to select for “interest” or “importance”, they might not publish something that changes everything. Or maybe it doesn’t, but what benefit is created by not having the information flow? This effect is particularly exacerbated in the “glamour” mags where mere 28 editors cover every single aspect of science from agriculture, to astronomy, to chemistry, and all aspects of medicine. Even if they were all Nobel prize winners at the top of their game (and let’s recognize they have been away from the lab for years and have become publishers), you would be hard pressed to find 28 people who are truly capable of selecting the Most Unexceptional papers from thousands of submissions across every scientific discipline that exists.

Extensive delays – It very commonly takes a year for a paper to be published. In the current world, that means that fundamental insight about Ebola could take a year for others to benefit from it! These delays serve no purpose and are simply a result of the machinery of traditional science publishing hopelessly mired in thinking from an analog age.

The public trust - Much scientific output has public funding behind it and people deserve to be able to see the work they paid for – immediately. After all, it’s theirs in effect – why should an editor make an arbitrary decision about whether it should be seen and when? The publishing morass serves to keep the public from the research they funded with their taxes. Crucially, most journals refuse to publish negative results that are not deemed “news worthy.” Labs that try and fail simply can’t publish in the large traditional journals. Most experiments necessarily fail, it is the very nature of science, and scientists are condemned to repeat those failures around the world because editors keep negative or ambiguous results out of their journals. The work has been done. The work has been paid for by the public trust - publishing everything benefits everyone and advances science.

Anonymous peer review – Peer review is a core building block of formal science publishing. In reality, anonymous review is not effective in fulfilling the intended purpose of assuring accuracy and credibility; its nature promotes fraud and is an excuse for delay. Authors must know who is evaluating their work and what they have said. Peer reviewers must be forced to stand up before the community with the same bravery as the author. Peer review makes science better with openness and by promoting dialog. Being reviewed in the open, subjected to a public critique is hard. Reviewing someone when you know THEY know who you are is also difficult, but it is the only way that there can be a true dialog in science.

Missing data – If a study's underlying data is published and shared, the scientific community can better assess the article’s accuracy. In an age of paper, including reams of data was impractical; in the digital age, not publishing all relevant data is indefensible. If the experiment generated gigs of data, gigs of data should be available for download; it is as simple as that. Science must rise up against bold conclusions without equally bold and complete accompanying datasets. If you have said it, prove it.


It is time for publishing -- the sole window through which science is made visible to the world -- to embrace the open culture of the Internet. In an age of serious global challenges, science will continue to be the source for answers. Those answers must be delivered with speed, fairness, transparency, and freedom from bias. The scientific journal is a relic of another age.
Daniel Marovitz is CEO of F1000.

Read more here/Leia mais aqui: Wired