Lessons from a year's worth of hiring data

I like this study.

However, perhaps you should do it again with some Bayesian analysis. It seems as though you neglected the base rate on multiple occasions.

For instance, you showed fewer offers with higher degrees, but you know fewer people have higher degrees. You have to adjust the effect by the applicant pool.

In addition, it seems that people who got offers had a lot more errors than other candidates. I wonder why this might be?