Donald Burrill wrote on 12/7/06 11:00 AM:
>And you didn't point out to them that a t-test (or any other statistical
>test, for that matter) is not applicable to a census?
The t-test (and other inferential tests) may still be applied even to a
census data set.
If we take the view that the observed census is just one particular
outcome taken from a theoretical pool of possible outcomes, then
statistical test is still possible.
Paul Bernhardt
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
