AI
Sunday, July 30, 2006 7:11:19 PM
Re : GTOP - Thanks Dew.
>>> In a strictly arithmetic sense, the appropriate hurdle when there is only one pivotal trial would be 0.0025 (0.05^2), not 0.025. However, the FDA rarely holds a sponsor to such a stringent threshold and usually gives at least 0.01 of cumulative p-value for a trial no matter what. <<<
That extra decimal point makes a big difference --- I need to get myself some more of that " arithmetic sense ".
Even though the penalty is only to a .01 p instead of .0025 , it still seems like the prudent thing to do is run two trials if at all possible , unless the results are highly predictable.
>>> In a strictly arithmetic sense, the appropriate hurdle when there is only one pivotal trial would be 0.0025 (0.05^2), not 0.025. However, the FDA rarely holds a sponsor to such a stringent threshold and usually gives at least 0.01 of cumulative p-value for a trial no matter what. <<<
That extra decimal point makes a big difference --- I need to get myself some more of that " arithmetic sense ".
Even though the penalty is only to a .01 p instead of .0025 , it still seems like the prudent thing to do is run two trials if at all possible , unless the results are highly predictable.
Discover What Traders Are Watching
Explore small cap ideas before they hit the headlines.
