4 ms·

The paper states that the acceptance rates are only higher when they are not identifiable as women > However, women's acceptance rates are higher only when the

• Sacho 10y ago
  Thanks, this is the relevant part of the paper for that claim: > For outsiders, we see evidence for gender bias: women’s acceptance rates drop by 10.2% when their gender is identifiable, compared to when it is not (2(df= 1;n= 18;540) =131;p < :001). There is a smaller 5.7% drop for men (2(df= 1;n= 659;560) = 103;p <:001). Women have a higher acceptance rate of pull requests overall (as we reported earlier),but when they are outsiders and their gender is identifiable, they have a lower acceptance rate than men Accepting arguendo that this is clear evidence of a gender bias, its effects are not as clearly pronounced as the tweet tries to imply they are.
  • Latty 10y ago
    Sure, except the author of the tweets also states that they believe this project in particular is a particularly bad outlier, with projects varying: https://twitter.com/isislovecruft/status/811510038166695937 https://twitter.com/isislovecruft/status/811510038166695937