bungler, short answer: yes. "p-value" is something everyone needs to know about a data set. Basic. No conclusions may be drawn without it. Sometimes not even WITH it. There are all different levels of statistics sophistication here on this board, at the FDA, and everywhere else, and those most versed in statistics know all the limitations of "p-values" and other statistical yardsticks.
Today's medical therapy makes an attempt be "evidence based", and essentially all evidence is "anecdotal" until it passes the statistical sniff test. That would be the "p-value" (and/or a dozen others). The most widely accepted bar is at at the p.=.05 level. As mentioned previously, it is like having a 100 chamber pistol with 5 bullets. The chances of these results being merely from chance and not real (if significant at the p.05 level) is the same as you firing a live shot. Most investigators shoot for far higher odds than that, ie 1 in 1000, but with limited sample size we take what we can get. Right now the data looks good, but is anecdotal. We'll see how it looks to the statisticians.
News 
Market Data 
Discover 
AI
