Ever wondered what the term "Clinically Proven" means, it gets bandied about by marketers alot. Here's an article that delves into what it really means.
Link
Friday, June 27, 2008
"Clinically Proven"
Subscribe to:
Post Comments (Atom)
About Me
- Yves Matson
- Yves hoovers stuff from all corners of the net and shares his favourite bits here. Also he might rant now and then.
2 comments:
Two comments on things the author states in his article"
First,
"A P value of 0.05 is easier to obtain than a P value of 0.001, but is much less convincing."
The author makes it seem like if you try hard enough, you can "acheive" a lower p-value. Statistics are reflective of the population being studied.
Second,
"When the P value is large, such as 0.05, this suggests there is a 5 percent chance that the result was from chance."
Okay, his explanation of what a p-value stand for is correct, but p = 0.05 is not large. In fact, typically p=0.05 denotes a statistically significant invalidation of the null hypothesis (whatever then null hypothesis may be). p = 1 is a large value.
Thanks for the comment Sid.
Well said.
I took away another point, which was that "Clinically Proven" is interpreted by the general public to mean, "we've proven that this thing works", and further that the last word there, "works" is interpreted as "almost always". So a marketing slogan for skin cream would go "clinically proven to reduce rinkles", interpreted by the general public to mean "this stuff almost always reduces wrinkles", is a sly way of saying "in certain percentages of the population in certain circumstances we've acheived a statistical and reproducable result, which we're declining to elaborate on, because it could be significant or it could be tiny but either way you won't be much impressed by it".
I guess that's the cynical marketer in me.
Yves
Post a Comment