xoxoxoBruce • Jun 27, 2020 10:29 pm
Many types...
The article says that’s not statistically significant?? :unsure:
Clodfobble;1054559 wrote:Whether it's statistically significant will depend on how many people were in the trial ...
I can see where 1100 people would not constitute a statistically significant trial result to researcher/tester types. However paying $3120 for a drug that works better than a placebo for 5 days but not for 10, seems a little steep. Especially since the 14 to 11 day improvement was for a 10 day trial. To my mind apples, oranges and bananas too.Clodfobble;1054559 wrote:Whether it's statistically significant will depend on how many people were in the trial--if two people get the drug and two don't, and both placebo folks die but one of the drug folks lives, that's a 50% improvement in outcomes, which sounds amazing but could just be chance. If I give it to 100,000 and still half the drug folks live and all the placebo folks die, then that's much more statistically significant.