Dr. Barry Sears Dr. Sears Comments:

This is why there is a big difference between clinical relevance and statistical significance.  In other words, bad science can be made to look good by tweaking the data.  Drug companies do it all the time.