more info on methodology?

[cathydeng]

hi! can you elaborate more on the methodology here? specifically wondering about US names - what's the source data for the confidence interval calculations?

[jeremybmerrill]

Hi Cathy,

That's a great question. It's been a few years since I used this code, but it looks like I used a formula called Agresti-Coull: https://github. com/jeremybmerrill/beauvoir/blob/master/lib/beauvoir/statistics.rb

Why I chose that, I don't remember, but I was likely copying from an OpenGenderTracker library in another programming language. I believe it was an R package.

The source data is from the Social Security Administration. I believe it's supposed to be of the population (I.e. not derived from a sample). Obviously it's not exactly the entire population (maybe just US-born? Idk) but it's pretty close.

Jeremy B. Merrill Sent from my mobile device

On Feb 13, 2017 6:03 PM, "Cathy Deng" [email protected] wrote:

hi! can you elaborate more on the methodology here? specifically wondering about US names - what's the source data for the confidence interval calculations?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/jeremybmerrill/beauvoir/issues/3, or mute the thread https://github.com/notifications/unsubscribe-auth/AAhdmp4T0vil0XK2GmHXpAsRo5s7YhNNks5rcOFBgaJpZM4L_1rm .