Hi, useful and interesting study!

About this comment: I (author of the Holman et al reference) do realise that many academics (me included) work in countries in which they were not born. However the majority DO work where they were born, so I decided it was best to use the name-country associations where known. To see why, let's use the example that in Denmark, Kim is a name most often used by men. Using the worldwide association for Danish authors would therefore result in more misclassifications (since Kims are usually women in most other countries). I guess you're saying that ideally one should use the person's country of origin rather than residence (and I agree), but this information is not known in most/any of the studies I have read. So, would you agree that it's best to use country of residence (as opposed to nothing) if this is the only information known? Also, can you explain how this would add "bias", as opposed to random inaccuracies that don't systematically alter the estimated gender ratios? I can't see how. Cheers!