Reviewers to demand authors share their data

Peer Reviewers’ Openness Initiative billed as a ‘grassroots’ response to concerns over reproducibility

一月 13, 2016
Research lab
Source: iStock

A coalition of scientists has launched an initiative to make sure research reviewers demand that authors openly share their data and methods.

The Peer Reviewers’ Openness Initiative (PRO) aims to make it easier to verify and test the reproducibility of scientific findings, following growing concerns about the reliability of some research.  

So far 225 reviewers have declared that from 2017, they will demand that papers make things public, such as data, materials such as survey questions and how to interpret files or code.

The initiative was launched by 13 scholars from the UK, United States, Australia, Germany and the Netherlands.

The idea behind the scheme is that authors need to have incentives to make their data public, because currently they fear doing so will benefit their competitors without any guarantee of reciprocation.

This does not mean that researchers are opposed to the principle of openness, but simply that they do not want to be the first to reveal their data and methods, the founders believe.

Richard Morley, a cognitive scientist at Cardiff University, said that “science should be an open and transparent process, but in reality it often falls short of this ideal”.

“Data and materials are often unavailable to other scientists or the public, hindering replication and verification,” he said.

Instead of changing policies, the PRO was a “grassroots” initiative, he said.

“Under PRO, reviewers provide complete reviews only for papers that share data and materials, or where the authors justify their lack of sharing,” Dr Morley explained.

“The power of PRO lies in the unity of collective action: since science depends on reviews, reviewers hold the key to opening up science.”

david.matthews@tesglobal.com

请先注册再继续

为何要注册?

  • 注册是免费的,而且十分便捷
  • 注册成功后,您每月可免费阅读3篇文章
  • 订阅我们的邮件
注册
Please 登录 or 注册 to read this article.