Plos launches open science data collection push

The Public Library of Science is beginning a project to track open science behaviours across scientific publishing, calling the lack of such data a critical barrier to making meaningful advances in research-sharing.

Plos, the pioneering non-profit open-access publisher founded in 2000, said that its new Open Science Indicator project would measure and report three characteristics of published articles: how many appeared in a preprint format, shared their research data, and made available the computer code underlying that data.

“It’s a really, really important evidence piece that helps us understand what the current state of open science practices is,” said Iain Hrynaszkiewicz, the director for open science solutions at Plos.

Other publishers and researchers have collected such data, Mr Hrynaszkiewicz said, but in limited ways and for limited time frames, such as a one-time analysis covering a particular set of journals or scientific field.

Plos plans to begin its project with an initial sample size of nearly 70,000 articles, with about 61,000 coming from Plos journals and 6,500 from non-Plos titles, then adding each quarter another 5,000 from Plos’ dozen journals and 500 from elsewhere.

“I don’t know of any other effort that has tracked these three things in this way, of any sample size,” Mr Hrynaszkiewicz said.

Universities, funders, scientists and journals have made major strides toward an environment in which government-funded science is available to anyone regardless of cost. The US took a transformative leap in that direction in August when the Biden administration gave government agencies until 2025 to adopt rules that would require all published results of federally funded research be made immediately and freely available to readers worldwide.

But progress has been slower in other dimensions of the problem. Even Plos, which was created with the express mission of pushing forward open publishing, has only 21 per cent of its papers appearing in a preprint format – a tool that speeds the sharing of research findings.

A central idea of the Plos initiative, Mr Hrynaszkiewicz said, is that participants throughout the scientific ecosystem will be able to see more clearly where they stand on the pathway to open sharing of scientific research, and test how much improvement particular interventions manage to create.

A national leader in the pursuit of research sharing, Brian Nosek, co-founder and director of the Centre for Open Science, said he backed the concept. “The availability of this data for publishers and funders,” said Professor Nosek, who was not part of the Plos initiative, “will help them fine-tune their policies and procedures to maximise adherence by authors and grantees so that the policy isn’t just talk, it also leads to action.”

Plos announced its effort at the same time as the Biden administration outlined a separate plan aimed more generally at improving equity in science. The White House hosted a broad gathering of universities, companies, and others interested in ways they could work across their organisational lines to improve minority participation in science and technology, and end the discrimination and biases that lead many to leave. Biden officials identified “generating better and more comprehensive data” on the scientific enterprise as a key element of its approach.

paul.basken@timeshighereducation.com