Researchers based in China or having a Chinese surname were far less likely than their US and Canadian counterparts to have papers published in the world’s most prestigious journals, according to a new study.
Just 2.5 per cent of submissions from China-based researchers were accepted, compared with 8.3 per cent overall for those from North America, a team found after analysing a rare dataset of more than 110,000 manuscripts submitted to Science and Science Advances between 2015 and 2020.
Even among US-based scientists, those with Chinese surnames had lower publication rates than those without Chinese surnames, they reported in Science Advances on July 15.
Sam Zhang, the study’s lead author and a computational social scientist at the University of Vermont, said one reason might be that China’s research system offered career and financial incentives for scientists to submit to elite journals such as Science.
That could encourage submissions even when some studies were not the best fit for the journals’ broad scientific scope.
The study could not determine whether the gap reflected differences in research quality and novelty or in how editors perceived authors and their affiliations. “Realistically, it is likely both,” he said in a university statement.
According to the paper, such a gap appeared throughout the review process, from the initial editorial screening to the final publication decision.
In a statement, Science said the editors of both journals were “grateful” to the study’s authors for providing a detailed analysis of trends in the journals’ peer review and publication pipelines.
“As the data shows, there is still some work left to do,” the journal said, noting that the editors were “looking more closely at the data and considering strategies to interrogate and address potential biases further”.
As part of that effort, the journals were hiring two new editors based in China, one for each journal, it added.
Zhang said the study was among the first to open a window into the editorial process at elite journals – a system that plays a pivotal role in shaping scientific careers but has long remained hidden from outside scrutiny.
By analysing a large dataset of anonymised submissions rather than only published papers, the study gave “really a first look at this process that was previously opaque”, Zhang said.
The team hoped it would encourage other journals to become more transparent and allow scientists to examine whether their review systems worked as intended, he added.
The researchers broke the publication process into several stages, from editors deciding whether to send a manuscript for peer review, to external review, revision and the final publication decision.
This approach allowed them to pinpoint where papers were filtered out along the way.
Their analysis also highlighted how selective top journals were.
Between 2015 and 2020, only about five out of 100 submissions were accepted by Science, compared with about eight out of 100 for Science Advances.
Institutional prestige also mattered. Papers from more highly ranked universities were consistently more likely to advance through the review process and ultimately be published, they found.
Meanwhile, the gender gap was much smaller, with acceptance rates of 6.4 per cent for male corresponding authors and 5.3 per cent for female corresponding authors.
Zhang said the rise of AI-generated junk papers was putting unprecedented pressure on elite journals and that greater transparency in the editorial process could help journals understand how their review systems needed to adapt. -- SOUTH CHINA MORNING POST
