Sample size calculation

[sternchen]

Hi all,

if got the a question concerning sample size calculation:
When doing a sample size calculation for a randomized clinical trial for comparing to group means I know the following formula:
\( n= \frac{r+1}{r} \frac{\sigma^2 (Z_{\beta}+Z_{\alpha/2})^2}{difference^2} \)

Now my question: How does the sample size change when I have a non-randomized trial instead of a randomized one? Which formula to use?

Thanks in advance for your help.

Best regards
sternchen

 

[ted00]

statistical test, and hence sample size formulas, don't typically account for random vs. non-random sample selection; can be a good idea to try your calculations for different values to get a sense of how sensitive the required sample size is to your hypothesized values, e.g. see first 3 here http://powerandsamplesize.com/Calculators/