Y
Hacker News
new
|
ask
|
show
|
jobs
by
apl
5242 days ago
Nitpick: You'd have to test the difference between before and after at a certain level of significance, not just establish estimates for before and after. Different things.