The treatment arm in m/m trial just has to have a one point improvement in the mean score over placebo for the primary end point to just reach significance. I run several simulations and sensitivity analysis with 5 different statistical tests and they all show the threshold as 1 point improvement to get to p < 0.05.
The question is do you believe LL arm will on the average perform at least one point better than placebo arm on a grading scale of 0 to 12?
Market Data 
Markets 
AI
