Skip to content
Prev 381340 / 398502 Next

stats::power.t.test error

You don't really want the diddling, since it gives meaningless values anyway...

For a pragmatic strategy, how about this?:

(a) calculate the power at n=2, if bigger than target power, done, else
(b) calculate n to reach target power, now guaranteed to have n > 2. Round upwards.

Peter D.