Saturday, October 18, 2014
A week and a half ago, I pointed out the gender imbalance apparent in the September issue of Geology. My particular gripe was that it would be hard to achieve gender balance in my ongoing geopoetry series if issues (like the September one) had three or fewer papers by women authors. With encouragements from commenters and the geotwitter rock stars, I had a slightly deeper look into what is going on with gender in geology, by recording the given name-assumed gender and author order for a year’s worth of Geology articles.
In total, this included 239 papers with a total of 1164 authors. The number of authors per paper ranged from 1 to 19. Of these authors, 64% were male, 19% were female, and 16% were initials. Initial authors excluded from the analysis; Most (57%) of them were on papers with six or more authors, so I assume that initialization was generally a space-saving exercise.
Looking only at uninitiated papers, the M/F ratio is 76.9% to 23.1%. This is not too different to the professional gender balance quoted here (76% M) and is slightly better than the decade-old numbers on assistant professor hires (23% F), but is substantially worse than the (similar era) graduating PhD student ratio (38% F). So the implication is that the Geology gender ratio mostly reflects post-grad school anti-female filtering.
As for author order, the observed vs expected ratios (given the gender ratio) are shown in the figure below. Due to the small size of the data set and the large number of individual categories, none of these deviations are statistically significant; the probability of sole author papers being seven M to zero F is about 14%- not high, but not enough to convict either. The M/F of first authors, second authors, etc. was generally within a few percentage points of the mean ratio, and always within counting stats.
And a quick
Carlo* suggests that the probability of getting three
or fewer female first authors in any particular issue is about 28% (see below),
based on 10,000 random author list generations for 20-paper issues.
This is only a simulation, of course. It will take the Geological Society of America just shy of 800 years to put out their 10,000th issue. Let’s hope that gender equity in academia has been achieved by then.
* Yes, I know there is an analytical solution, but simulations are more fun and quicker.