|
|
 |
2 Sample T-Test (Sample Size)
Posted by: Darth Posted on: Tuesday, 3rd November 2009, 11:25 AM.
10% is a proportion. You would have to use a quantitative difference to use a t test...a difference of so many inches. You will still need an estimate of the s.d. to get the sample size. Usually you would do a small sample to get the estimate and then do your full blown experiment. If it turns out that the s.d. was off a little then you will either need a few more to keep the alpha and power where you want or you will have more alpha and beta than you expected which is good. Message Thread:  Return To Discussion ForumPost A New MessageRead the Forum Guide to Good Etiquette
"The Bottom Line" Links
|
 |
|