Thanks for your response.
Could you please give me more details about your post?
> I like an approximate randomization approach (or exact if the sample is
> small) and testing whether the SSerror is smaller than X% of the random
> permutations' SSerror for the test statistic.
1. What is the approximate randomization approach?
2. What does the sample mean? The size of the dataset used for model fitting?
3. How to determine the value of X%?
> With different models I believe that that would translate to simply taking
> the one with the smallest p-value (for tests done with, say, 5000
> permutations)
4. What kind of p-value is it? How to get it? What test shall I do?
5. What does "test done with ... permutations" that mean?
Sorry for so many questions and thanks a lot for your help!
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
