Quote:
Originally Posted by FloridaMan88
Trying to make blanket projections with a small sample size… not an accurate basis for comparison.
|
The issue with the sample size argument remains that the outcomes are SO bad that you could double the sample size and (for some reason) give the new sample a 50% return rate and STILL only get to a 25% net.
The N is big enough and the output poor enough that increasing the sample size isn't going to move the needle a ton.
The only way to defend the numbers are to distinguish them. Essentially dismiss them altogether because...reasons. You can't reasonably sample size your way out of this one.
You have to just say it's inapplicable because....whatever. He had a different procedure or we have some fancy new imaging that establishes his blood flow is unprecedented and it takes him out of that study altogether.
Sample size arguments don't work when the outcomes are THAT bad.