Big science v little science

Funding fewer but larger research grants is likely to decrease the impact of the science base, a study has concluded.

In recent years many funders worldwide, including the UK research councils and the Wellcome Trust, have given out longer and larger grants to fewer applicants in the belief that concentrating funding on the best researchers will maximise the money’s scientific impact.

However, a paper published on 19 June in the journal PLOS One by Jean-Michel Fortin and David Currie, respectively research assistant and professor in biology at the University of Ottawa in Canada, argues that this approach is misguided.

The conclusion is based on analysis of papers funded by the Natural Sciences and Engineering Research Council of Canada – which is also moving towards fewer, larger grants – between 2002 and 2010.

The paper, “Big Science vs Little Science: How Scientific Impact Scales with Funding”, uses four bibliographic measures to assess grant recipients’ scientific impact: the volume of publications they produced, the total number of citations they received, the number of citations for a recipient’s most cited paper and the number of highly cited papers they produced.

The first two measures rose along with the grants, but the gains became less marked as funding increased, so impact per dollar decreased.

The paper also reports only a weak correlation between high funding and high citation. “Rich” researchers’ most highly cited article received on average 14 per cent fewer citations than the most highly cited paper by any random pair of researchers who received only half as much funding, while two smaller grants yielded 20 per cent more highly cited articles than one larger grant.

The paper argues that funders should revert to offering a larger number of smaller grants because “greater scientific diversity, like greater genetic diversity, increases the probability that some researcher (like some genetic mutant) will possess characteristics that will flourish in an unpredictable future”.

Kieron Flanagan, lecturer in science and technology policy at the University of Manchester, said the findings were “plausible” despite the paper’s “narrow” definition of scientific impact. He added that there was a surprising lack of research into the best way to fund science.

“Most of the arguments for concentration are based on common-sense understandings of efficiency or economies of scale, which don’t necessarily make sense for science – and which are perhaps also influenced by the self-interest of the ‘elite’ researchers who have the largest say in funding decisions,” he said.

Jack Stilgoe, lecturer in social studies of science at University College London, agreed that science policy was “ironically, a pretty evidence-free zone”. While cautioning against reading too much into bibliometric analyses, he said he thought that they should form part of a push to build up stronger, ideally nation-specific evidence bases of the kind being sought by the US’ “star metrics” programme.

A spokesman for the Wellcome Trust declined to comment on the paper but said that the aim of the charity’s larger individual-focused grants was “to give our researchers flexibility when answering important biomedical questions”.

paul.jump@tsleducation.com