# Computing required sample size for paired t-test

I have some paired t-test data (40 pairs) and I would like to use these data to estimate the sample size required if I were to conduct the study again with a new set of requirements.

I'm using minitab's power & sample size calculator for paired t. I need to enter the difference, the power (I have both of these) and also the standard deviation of paired differences.

Regarding the last one, does this just mean, take the differences of the pairs and compute the SD? Or do I have to compute SDd as above and use the formula: $SDbar = SD/\sqrt n$?

What you are calling "$SDbar$ is the standard error of the mean difference for the paired t-test. At any rate, no that isn't what you need to enter. What Minitab wants is the regular old standard deviation.