I dig into the recently published "Name-based demographic inference and the unequal distribution of misrecognition", by Lockhard, King, and Munsch.

As the paper highlights, what seem like reasonable overall error rates of name-based demographic inference tools quickly degrade into unacceptable (and unpublicized) error rates when looking at intersections of demographics.

#DataScience #SocialScience #SystemicBias

https://www.juliaferraioli.com/blog/2023/influential-articles-jun/