(Total Views: 680)
Posted On: 11/06/2020 4:34:00 PM
Post# of 145248
This is for us not so bright investors. We will be scorned for not doing a google search to find out the influence of power. I did a google search, but am hesitant to display confidence on IH for sure.
Power (aka "n" aka "sample size" is used in, and is one of the "factors" in calculating p value. Thus, p value is the bottom line....it takes into account, the number of folks in the trial, thus, is improved at stating the "probability of accuracy", vs no consideration of sample size.
If power was not used in calculating p value, then in theory, you could have very low power but a great p value, and thus "probability of accuracy" would be reduced.
Say you had 6 people in a clinical trial, 3 on placebo and 3 on your experimental drug. All your placebo folks die, none on your drugged folks die. Power would be tiny, p value would be great (if power was not part of the p value formula).
All of this I'm sure is known and obvious to most on the discussion board.
Power (aka "n" aka "sample size" is used in, and is one of the "factors" in calculating p value. Thus, p value is the bottom line....it takes into account, the number of folks in the trial, thus, is improved at stating the "probability of accuracy", vs no consideration of sample size.
If power was not used in calculating p value, then in theory, you could have very low power but a great p value, and thus "probability of accuracy" would be reduced.
Say you had 6 people in a clinical trial, 3 on placebo and 3 on your experimental drug. All your placebo folks die, none on your drugged folks die. Power would be tiny, p value would be great (if power was not part of the p value formula).
All of this I'm sure is known and obvious to most on the discussion board.
(3)
(0)