Global Machine Learning & Data Science Research

Nature Bid to Boost Paper Review Transparency Applauded by ML Researchers

Leading scientific publication Nature announced it is launching a trial starting this week that will give authors of newly published papers the option of appending contents of the discussions they’ve had with and reports they’ve received from their reviewers.

There is plenty of back-and-forth that goes on between the scientists who write research papers and the experts who review them for academic journals. Amid growing calls from the research community to bring greater transparency to this process, leading scientific publication Nature announced it is launching a trial starting this week that will give authors of newly published papers the option of appending contents of the discussions they’ve had with and reports they’ve received from their reviewers.

Nature promised readers a glimpse into “the often fascinating and important discussions between authors and reviewers, which are crucial in shaping and improving research and checking its integrity.” The journal also suggested that publishing these exchanges may better frame new research as part of “the scholarly journey.”

Peer reviews in academic journals typically follow a single blind method wherein reviewers know who the paper authors are, but authors do not know who is reviewing their paper. Four years ago, Nature began encouraging reviewers to include their names, and in a 2017 survey of Nature reviewers 82 percent supported standardizing the peer review process, with more than half agreeing the process should be more transparent.

Many in the machine learning community will welcome the behind-the-scenes peek afforded by the Nature initiative. Google Research scientist David Ha tweeted his hope that major machine learning conferences will follow the trend by also opening up their paper review processes.

Toronto Computational genomicist Michael Hoffman tweeted “There are way too many publishers offering substandard peer review and suckering prominent scientists into participating. Can only counter with evidence of a legitimate peer review process.”

Peer review has been a controversial topic in the machine learning community due in large part to the ever-increasing number of papers submitted to conferences like CVPR in recent years and concerns about reviewer qualifications (related Synced story). Many believe publishing reviews may push reviewers to deliver more constructive opinions. The reviewers, under the new arrangement, can choose to be named or remain anonymous.

The EMBO Journal and BMC journals is the first tier that started making their peer reviews more transparent about a decade ago. Nature Communications followed in 2016, and Nature Biomedical Engineering, Nature Cell Biology, Nature Ecology & Evolution, Nature Human Behaviour, Nature Immunology, Nature Microbiology and Nature Structural & Molecular Biology made the change in December 2019.

Nature stressed that their peer reviews will need to meet high standards whether or not they are being published. The journal says that 98 percent of authors who have already published both their paper and peer review reports on Nature Communications reported they were happy with the change. Nature will report back on how this trial proceeds. See the official announcement here.


Author: Reina Qi Wan | Editor: Michael Sarazen

1 comment on “Nature Bid to Boost Paper Review Transparency Applauded by ML Researchers

  1. I like such topics.

Leave a Reply

Your email address will not be published.

%d bloggers like this: