Conversation

Replying to
It doesn't seem like this is normalized for the number of people in each cohort. The clean lines may be hiding some unpleasant details.
1
2