Replies to post #1351 on Dendreon Corp fka DNDNQ

[iwfal]

But I have never seen the statistician reply that he/she is not the least bit surprised, given the lower hurdle.

A) Again, it wouldn't surprise me in the least if >50% of practicing biastatisticians didn't know this factoid. You give too much credit to technical folks actually groking what they are doing (vs pluggind and chugging).

B) If the added assumptions that go into a Cox Regression are correct then it isn't that Cox Regression is, per se, a lower hurdle than the normal 0.05, but that the unadjusted Log Rank is too high a hurdle I.e. Q: If the Cox Regression assumptions are correct, which provides a p value that is a closer measure of the chance that the trial seen could have occured by chance alone? Unadjusted Log Rank or Cox Regression? Ans: The Cox Regression.