Homing in on the 'Homer Simpson Effect': reply to Dugdale et al
Journal Contribution ResearchOnline@JCUAbstract
We agree with Dugdale and colleagues [1] that the continuing underrepresentation of women in the higher echelons of science is both a serious problem and one that merits far more in-depth study than was attempted in our recent analysis [2]. Our unexpected finding that recognized male experts did not profess greater scientific expertise than did female experts with comparable years of experience (failing to support our so-called 'Homer Simpson Effect') was merely an opportunistic observation within a much wider investigation of the long-term fate of tropical nature reserves. Dugdale et al. suggest that a broader assessment of scientists, especially including younger researchers, might yield different results than those we found. This is plausible but is a much more general question than the one we posed. In taking care not to overgeneralize, we emphasized that our findings applied merely to the sample of researchers we surveyed. We also underscored the fact that our experts were strongly biased in favor of males and concluded that 'the scientific playing field is not yet fully level' [2]. Dugdale et al.'s point about optimal questionnaire design would be taken in a slightly more favorable light had our study been designed to assess gender issues in science. In terms of statistical analysis, the brief format of a TREE Letter (and clear instructions from the journal editor to minimize our statistical treatment, given that TREE does not normally publish original research findings) limited our capacity to provide much quantitative detail. In the ANCOVA that we summarized [2] we did, in fact, conduct the normal safeguards, such as ensuring that our residual-error structure was roughly Gaussian and homoscedastic [3] and that there was no significant gender–expertise interaction; however, it was not possible to squeeze all such ancillary details into our brief Letter. Although our ANCOVA was quite adequate, we agree with Dugdale et al. that statistical analyses ideally should accommodate the specific distribution of the response variables. In this vein, an ordinal logistic regression [4] provided almost identical results to our ANCOVA [the top-ranked information-theoretic model (AICc weight=0.64) included years of experience only, whereas gender and the gender×years of experience interaction explained almost zero additional deviance]. Overall, the effect size comparing the self-assessment of expertise between males and females was negligibly small (0.012), and slightly favored males. These statistical details aside, we agree wholeheartedly with Dugdale and colleagues that it is important to investigate why fewer women than men are ascending to the highest ranks of science. If our simple and slightly irreverent Letter [2] should help advance this aim, then we would be delighted.
Journal
N/A
Publication Name
N/A
Volume
26
ISBN/ISSN
1872-8383
Edition
N/A
Issue
12
Pages Count
1
Location
N/A
Publisher
Elsevier
Publisher Url
N/A
Publisher Location
N/A
Publish Date
N/A
Url
N/A
Date
N/A
EISSN
N/A
DOI
10.1016/j.tree.2011.07.013