Dear SPMers,
I have a rather basic question about the statistical power of random
effects analysis. I understood that the power of fixed effects analyses
depends largely on the number of scans, while in random effects analyses it
depends on the number of subjects in the study. My question is, whether the
number of scans has any impact on the statistical power at the second level
at all and if so, whether there is a way of quantifying it.
Greetings,
Stefan
|