PakMediNet Discussion Forum : Biostatistics : sample size
sample size
Hi every body;
I just wonder if anybody knows the formulas used to calculate the sample size for a medical study.. or any web sites that clearly state how to calculate the sample size...
thank you
Posted by: palestinePosts: 7 :: 11-10-2006 :: | Reply to this Message
Re: sample size
A.S.A.
You need to consider the following:
(1) Event rate in control group.
(2) Relative treatment effect (ie difference in treatment effect between two groups).
(3) Type 1 error (also knows as alpha): usually 5% (a 5% chance we will conclude that there is a treatment group difference when, in truth, treatment is not better or worse than a placebo).
(4) Type 2 error (also knows as beta). Power of a study is 1-beta.
Try the following online calculators:
http://statpages.org/#Power
http://statpages.org/proppowr.html
If unsure, I suggest you get help from a medical statistician.
Good luck,
Rehman
Posted by: asiddiquiPosts: 26 :: 11-10-2006 :: | Reply to this Message
Re: sample size
thank you so much rehman for your resonse to my question.but i wonder if you have an idea bout some web sites that show some examples for sample size calculations.
thanx
Posted by: palestinePosts: 7 :: 12-10-2006 :: | Reply to this Message