Alexander von Humboldt Foundation adopted randomised process for whittling down applications
A state-sponsored academic exchange funder in Germany has been forced by the Covid-19 pandemic to use a lottery process to award 10 per cent of its prestigious fellowships in 2020.
The secretary general of the Alexander von Humboldt Foundation, Enno Aufderheide, told Research Professional News that some fellowships had been chosen through a randomised procedure for the first time this year, because the full review committee was unable to meet due to Covid-19.
Lottery-based procedures for awarding grants are rare but not unheard of among funders. Germany’s Volkswagen Foundation and New Zealand’s Health Research Council have established programmes which include a randomised element, with proponents arguing that they increase diversity of ideas and reduce unconscious bias among reviewers. But most existing lottery schemes are for blue-skies research, and it is unusual for fellowship programmes not to be fully decided by review panels.
Alexander von Humboldt Foundation research fellowships bring postdoctoral and experienced researchers from around the world to work in Germany. The foundation awarded 621 fellowships in 2019 and has a 2020 budget of €145.5 million, largely from the German government. The research fellowship programme accounts for around 30 per cent of the foundation’s annual budget.
Aufderheide said that the foundation split the review committee, which numbers more than 40, into several smaller committees for the last two rounds of fellowships. These assessed applications virtually via videoconference. In meetings held in March, the subcommittees eliminated weak applications and approved 149 for funding. But 33 applications which were strong enough to be considered for funding remained, of which 17 were chosen at random.
“It was not possible to make a decision,” said Aufderheide.
Further review meetings were held in July where the same process was used. The foundation was not able to confirm exact numbers chosen by lottery in July, but Aufderheide said the numbers were similar to the March meeting.
Applicants to the grant scheme were not informed that any of the winners were awarded through a lottery, rather than via full peer review. “This has not been publicised,” said Aufderheide.
Aufderheide said that for applications near the financial cutoff point for funding, a lottery process is “as just” as full peer review.
“It’s very hard to say who is a little bit better or not,” he said, “but I think people hesitate at having randomised elements in a scientific decision process.”
Nonetheless, the governing board of the foundation approved the process in the face of Covid-19 restrictions. With the next meeting due in October and the pandemic showing little sign of abating, Aufderheide said “we’ll probably have to do it another two times”.