Hacker News new | ask | show | jobs
by soco 2012 days ago
Aren't we doing here statistics on a one-surgeon sample size?
1 comments

Where is the "one surgeon" sample size?
It's about the success of surgery on that surgeon's birthday, and calculating the implications of switching the surgery types etc etc, while in a real world situation you'd just use the other surgeon - which hopefully has another birthday. I agree there's gonna be some smaller effect even then, but less and less the more surgeons you have and the more randomly distributed their birthdays are.