Thanks for the feedback. Let me ask a more specific question. If
using a Tukey Endcount test, you choose a confidence (or significance,
I'm not sure which is correct) level, and look up the required
encount. If you run the test, and then have a higher than required
endcount, can you then change the significance level of the test, and
determine a higher level? If not, Why?
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
