An analysis of medical research grant applications in Sweden suggests that the peer-review process is not immune to sexism. After a court battle to gain access to reviewers' evaluations, two Swedish researchers report in today's issue of Nature that women applicants were consistently ranked as less competent than men with similar publication records.
Like most funding agencies, Sweden's Medical Research Council (MRC) relies on a scientific peer-review process to determine who gets grants or fellowships. Microbiologist Christine Wenneras and immunologist Agnes Wold, both at Göteborg University in Sweden, applied for research fellowships. After being turned down, the pair noticed that of the 20 spots available, only four went to women. "We thought the MRC discriminated against women scientists," says Wenneras. But when she and Wold asked to look at the peer reviewers' evaluation sheets, the MRC refused.
Swedish law states that all unclassified state documents must be made public, so after a skirmish in court, Wold and Wenneras gained access to the evaluation sheets for 114 applicants--of whom 52 were women--for 1995 postdoctoral fellowships. In each case, five peer reviewers gave an applicant three scores--competence, relevance of the proposal, and the quality of the experiment. Wenneras and Wold compared the scores with the applicant's publication record, using such measures as the number of articles published and the prestige of those publications to gauge an applicant's scientific standing.
Wenneras says the results confirm her suspicion: Reviewers gave women consistently lower scores than men with an equivalent publication record, especially in the "competence" rating. In fact, according to the analysis, a woman needs three extra papers in a prestigious journal like Science or Nature to get the same ranking as a male researcher.
"Our results are pretty depressing," says Wenneras. "It goes a long way to show why women are less successful than men in the biomedical field." She suggests that more objective guidelines for reviewers might reduce the bias. Other agree that the evidence is disturbing. "Something is going on there, clearly," says Annette Flanagin, a nurse and the associate senior editor of The Journal of the American Medical Association. She notes that seven journals, including JAMA, are currently investigating whether bias might be avoided by not revealing to reviewers the identity of a paper's authors.