Borel Fields, Thanks for your comments. Howeve
Post# of 155328
(Total Views: 621)
Posted On: 09/23/2020 1:34:55 PM
Re: Borel Fields #57548
Borel Fields,
Thanks for your comments. However, I don't see where you disagree with me. By the contrary, you are saying the same thing:
Please note that I am just making assumptions and speculating about the p-values. At the end of the day, it all translates on what "price" we will be paying by doing a "full interim", so to speak.
We agree that p will be around 0.0492 (O’Brien-Fleming criteria) or 0.04 & 0.045 as you point out).
The problem is the stoppage in the interim, this can probe to be a difficult task if we don't have the right number of deaths (horrible thing to say). But we agree as well that the number needs to be much smaller than 0.05. We can argue how smaller but is impossible to have "hard" data to get into this as we would making assumptions that make our calculations moot.
GLTU
Thanks for your comments. However, I don't see where you disagree with me. By the contrary, you are saying the same thing:
Quote:
This latter criteria would require a 0.0054 at interim look and 0.0492 at the end of trial. So, maybe Nader was referring to the END value but need to account what the interim would look like.
Please note that I am just making assumptions and speculating about the p-values. At the end of the day, it all translates on what "price" we will be paying by doing a "full interim", so to speak.
We agree that p will be around 0.0492 (O’Brien-Fleming criteria) or 0.04 & 0.045 as you point out).
The problem is the stoppage in the interim, this can probe to be a difficult task if we don't have the right number of deaths (horrible thing to say). But we agree as well that the number needs to be much smaller than 0.05. We can argue how smaller but is impossible to have "hard" data to get into this as we would making assumptions that make our calculations moot.
GLTU
(0)
(0)