Re: power calculation

To do it properly you'll also need more context. Where are the mean
difference, sd and sem of the groups from? Most likely they are from a
pilot study or from previous research and using them as estimates of
the mean difference and sd for a future study. If the future study
differs in design or if the pilot/earlier study have small N that'll
influence the estimate (although in some cases you won't know in
advance how the estimate will change in other cases you can make an
educated guess).