In regards to this point - "It's interesting to note that though WT treated patients showed numerically bigger response over the WT placebo than did ITT treated patients over ITT placebo, the p values are bigger [less significant] in WT than in ITT, e.g. for ADAS-Cog data, p=0.015 in WT patients vs. p=0.008 in ITT." From what I understood the p value was higher in WT despite the stronger effect size due to the 'n' only being 70% of the ITT 'n' and larger populations tend to mean lower p value.
Market Data 
Markets 
AI
