Anavex Life Sciences Corp (AVXL)

[pcguy]

I think we may be saying the same thing. A statistical analysis should be performed to determine required sample size. Regardless of the amount of data, a statistical analysis will determine the likelihood a sample size will produce a desired result given an acceptable amount of variance. The desired result is whether the drug works is an adequate sample of the target population to account for variables in the target population. The target population may only be a subset of the entire population. If the trial design can account for the variables, a smaller sample size would be warranted. Taking a pot shot at sample size is a huge risk with investors money.

The FDA or any other regulator authority is not going to just be interested in whether statistical significance was achieved in a sample population, they are going to also want to know what the target population is and how the trial represents that population.

Example... I am making a drug that grows hair in balding males. We take a sample of 100,000 males 18-21 and they happen to all have brown hair. 90% of the sample regrew a full head of hair. What does this tell us about blondes or red heads, and anyone over 21? What is the confidence this drug would work in those people as well? Should the company be able to advertise that thier product can regrow hair in men?