Sharing open data: sometimes costly but mostly good
Researchers gain by openly sharing data, but men and junior scientists are more likely than women and senior scientists to object, an UdeM survey suggests.
Many Canadian researchers share their data but some – often young men at the start of their career – say competitors have ‘scooped’ them and misused their data, according to a new survey.
Done as part of a master’s thesis in Université de Montréal’s quantitative and computational biology program, the survey of faculty members at Canada’s top 20 universities was published today in the journal BioScience.
It suggests that, while the overwhelming majority of researchers support data-sharing and think it’s good for science, roughly one in five has had a negative experience from doing so.
In particular, some male researchers in the early stages of their careers resent competitors using their data for their own ends, and one in three researchers think data-sharing is too time-consuming when required by scientific journals.
Funded by Canada’s National Sciences and Engineering Coucil, the survey was completed using open-source software in January and February 2020 by 140 Canadian faculty members in the biology fields of ecology and evolution.
They were asked 14 multiple-choice questions and invited to comment.
“The survey had to be short and to the point because professors are incredibly busy,” said Dominique Roche, a research associate in biology at UdeM when the study was done, and now a research fellow at Carleton University.
A broader view
Nevertheless, Roche and co-author Sandrine Soeharjono believe, the survey hints at a broader view among faculty and provides insight into researchers' data-sharing behavior. It will be used to improve data-sharing policies increasingly common among scientific journals and research funding agencies.
“Earlier studies have looked at the fears and attitudes of researchers towards open data,” said Roche, whose current research is funded through the European Commission’s Marie Skłodowska-Curie Global Fellowship program.
“This is the first time that the actual costs and benefits of data-sharing to researchers were measured,” he said.
“Most people will agree that data-sharing is good for science and society, but many researchers don't want to share because they fear personal costs in terms of career advancement,” he said.
“Until now, it was unclear whether these fears were warranted or overblown.”
Roche’s previous work has shown that there are many barriers to scientists sharing their data and that those who share often don’t do it well.
“Unfortunately, most open data out there are incomplete and not reusable,” he said. “This is true not just in ecology and evolution, but also in many other fields, such as psychology and cognition research.”
“But the truth is, researchers have little to fear from data-sharing,” he said.
Ninety per cent of researchers in the survey answered that sharing their data had led to beneficial or neutral outcomes, with only one in five saying they had experienced some form of cost.
“Most often, that meant others misusing their data, but that is very subjective,” he noted. “It might be upsetting if someone publishes a paper that challenges your work, but it doesn’t necessarily mean that the data were misused.”
Sharing during the pandemic
At a time when the COVID-19 pandemic has raised the importance of sharing research data to speed up scientific discoveries, he added, “we hope our results will encourage researchers to continue sharing, even after the pandemic is over.”
Soeharjono, a master’s student at UdeM when the survey was done and now a junior data scientist at the Montreal tech startup My Intelligent Machines, agreed:
“I was excited to work on this project because I believe that scientific data should be open access in order for society to reap the full benefits of research,” she said.
“We are social creatures, after all, and should aspire to thrive from collaboration instead of competition,” she added. “I am glad to see that many researchers are on-board.”
Others who cited negative experiences from sharing data said they had been ‘scooped’ in their discoveries, were not included as co-authors on a paper when they felt they should have, and were generally frustrated from having to share data they would have preferred to keep private.
To help researchers overcome their hesitancy to share open data, Roche and Soeharjono suggest that university departments and scientific journals make an effort to:
- create better guidelines, standards and training for data-sharing;
- provide greater support for research data management and equity in sharing practices;
- and provide better incentives for sharing and better protection from potential negative outcomes.
“If our study can prompt a single conversation about the movement towards open sharing of data, I consider that a step forward," said Soeharjono.
About this study
“Reported individual costs and benefits of sharing open data among Canadian academic faculty in ecology and evolution” by Sandrine Soeharjono and Dominique Roche, was published April 20, 2021 in BioScience.