Gender bias in the refereeing process

Our sample of experts included 190 male scientists and 61 female scientists. We note at the outset that this sample is strongly male biased (x 2 = 35.6, df = 1, P < 0.0001; chisquare test). On average, the men had a slightly greater period of experience at their study area (mean AE SD = 19.8 AE 10.1 years) than did the women (16.3 AE 7.6 years), a significant difference (t = 2.44, df = 249, P = 0.015; twosample t-test with log-transformed data). However, we found no difference in the level of perceived expertise between men and women, either in a direct comparison (t = À0.12, df = 249, P = 0.90; two-sample t-test) or when using the number of years at the study site as a covariate to adjust for varying levels of experience. The latter analysis indicated that gender had no significant effect on perceived expertise (F 1,248 = 0.73, P = 0.39), whereas the number of years at the site had a strong and positive effect (F 1,248 = 9.31, P = 0.003; ANCOVA with log-transformed year data). Our analyses suggest that, at least among tropical researchers, men and women rank their professional expertise similarly given comparable levels of field experience. In 1993, a prominent female scientist suggested that 'women tend to make a much lower self-evaluation of their success than do their male colleagues' [8]. Our findings from 2008 to 2010 seemingly provide a more optimistic view, although women were identified less frequently than were men as experts on their respective long-term study sites. We conclude that the Homer Simpson Effect is insignificant among the researchers we surveyed, even if the scientific playing field is not yet fully level