Power Rules
I have
an estimate of SD(Y)
the SE from an existing study
the SE from my pilot
and I want to find
the MDE
the power
the sample size
for
a treatment effect
an interaction
.
Power level:
80%
90%
95%
Custom
SD(Y) =
SD of the outcome in a population like yours, e.g., from a census or prior study
R² =
proportion of variance explained by covariates; leave blank for no adjustment
SE (existing) =
standard error of the treatment effect from that study
n (existing) =
number of respondents per condition in that study
SE (pilot) =
standard error of the treatment effect from that study
n (pilot) =
number of respondents per condition in that study
n =
number of respondents per condition in the planned study
tau =
the smallest treatment effect you'd consider meaningful, in units of the outcome