Researchers ‘plot promotions using conference collusion rings’
Massive academic conferences are falling victim to “unethical” groups of peer reviewers who are secretly sharing the titles of publications to ensure preferred researchers are invited to these events, a leading computer scientist has claimed.
Warning about the dangers that “collusion rings” pose to the integrity of scholarly meetings, Michael Littman, Royce family professor of teaching excellence in computer science at Brown University, said he was keen to raise awareness of the “growing problem” blighting conferences in his discipline, which had seen senior researchers team up to manipulate the “blind” peer-review systems of some conferences.
At some large computer science meetings, committees use as many as 5,000 reviewers to assess which submissions – numbering up to 10,000 – are worthy of presentation, with reviewers placing “bids” to review papers within their field of expertise.
While all papers are anonymised to ensure reviewers do not evaluate those on which they have a conflict of interest, Professor Littman told Times Higher Education that some scholars were secretly passing the titles of submitted works to each other to make sure they were selected.
“The idea that people are gaming these conferences – using their energy and brilliance to thwart the system – really offends me,” he explained.
Since he wrote about the issue in this month’s edition of the Communications of the ACM (Association for Computing Machinery) after three conferences held in 2019 found evidence that these illicit practices had taken place, Professor Littman said many more scholars had contacted him to explain “what was happening behind the scenes”.
“I have become a clearing house for this topic – people are saying they see this behaviour in other parts of computer science,” he said.
Similar corruption of peer review organised by US science funding agencies in relation to grants had also been alleged, said Professor Littman. “If collusion rings are happening for funding, that’s something that doesn’t just affect computer science,” he said.
However, his discipline was more likely to be targeted by such practices given the high premium that it places on presenting at prestigious conferences, with publication in the relatively slow-moving world of journals seen as much less important for career advancement, he said.
“By the time someone has published a paper, everyone will have already heard about it because they will have presented it at a conference,” said Professor Littman, who added that “decisions on promotions and tenure are often made on whether someone has presented at a very good conference”.
With success rates often lower than 20 per cent, this helped to uphold the standard of research presented at conferences, but collusion rings meant low-quality work was more likely to be accepted, said Professor Littman, who would like to see conference suspensions for those caught collaborating in this way.
“People shouldn’t be fired – we perhaps need to treat them as those caught cheating in class,” he said.
The increasing popularity of subjects such as machine learning and artificial intelligence was also challenging for conference organisers, he added.
“That large scale brings benefits as a lot of people are trying out new ideas which can be shared quickly, but the disadvantage is that you lose that sense of belonging to a community and the feeling that it’s not in your interest to screw someone over to get ahead,” said Professor Littman.