did the AC really focus in on the difference in medians being only 3 weeks in the final look?
Several members did just that. They would give a nod to the HR and then go on to talk about the effect using only the median. Of course that might just have been them talking their position, but I don't think so. For whatever reason, the median seems much more concrete to all the non-statisticians than the HR.
Merck did produce a "computed" median based on the HR - I think it was around 6 weeks, but nobody seemed to pay that much attention.
Market Data 
Markets 
AI
