Retrospective or "observed" power is defined as the power
of the test just performed, based on the observed effect
size and the same sample size and significance level.  

It is a useless exercise -- see Hoenig and Heisey (2001),
"The Abuse of Power: The Pervasive Fallacy of Power Calcu-
lations for Data Analysis", The American Statistician, 55, 
19-24.

This user interface is based on the fact that power is the 
probability of rejecting the null hypothesis.  Retrospec-
tively, the outcome of the test is known; hence if we use 
*all* available information, the retrospective power is 1 
if the test was significant, and 0 otherwise.  

This formulation of retrospective power is easy to calcu-
late, and impossible to abuse.
