Replies to post #180742 on Cytodyn Inc (CYDY)

[kgromax]

It’s a good illustration of this important investment principle:
If statistics were so « obvious », there would be no need to pay statisticians to crunch the numbers.

In that case:

- The DSMC could already see based on interim data that there was an issue: either the hypothesized effect was absent (=Leronlimab is saline) or the trial was underpowered (=not enough additional patients remaining in the second half to reach statistical significance with a good probability)
- The DSMC knew that, ONCE CORRECTED FOR DATA MINING, the 14 days numbers and other numbers were non significant (=pure random noise)

As a reminder, Nader’s pvalues at 14 days have NOT been corrected for data mining. Nader has recognized it in a conf call. That is total clown show at best and fraud at worst. Any first year student in statistics knows that data mining WITHOUT pvalue corrections allows you to always come back with completely fake « statistical significant » pvalues. But once corrected, they are not stat sig any more. I did some rough corrections for the data mining based on the scarce data published and, as if by magic, the 14 days numbers were non stat sig (pvalues waaay above 5% in all cases).

Shareholders have been scammed, as every time with Cytodyn.