Dear members, I would be very grateful if you could advise relating to
the following issue:
When testing for a difference in population means (variances known), the
z-test is used, with sample means xbar1, xbar2, and sample sizes n1 and
n2. The test is straightforward.
When one has information relating to the entire populations under study
(i.e. random sampling not required), the population means are mu1 and
mu2 and population sizes N1 and N2. In this case, is it still
appropriate to use the same z test, but substituting population
statistics for the sample statistics in the formula?
Please advise, pointing me in the direction of appropriate references if
possible.
Thanks,
Fen Scott.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|