Re: ANOVA and Interaction


On 25-Aug-2006, Bernhard Laback <Bernhard.Laback@xxxxxxxxxx> wrote:

The data are shown in the following table:
The particular question I would like to answer is at what levels of A
the diffences between the levels of factor B are significant.

A B Dependent
200 1 25.28
200 1 30.49
...

I missed the beginning of this thread. Is the goal to build a model to
predict Dependent = f(A,B), or is your goal strictly related to finding the
significance of A and B? Do you want to treat both A and B as ordered
values? Are you open to other types of models such as decision trees and
SVM?

--
Phil Sherrod
(phil.sherrod 'at' sandh.com)
http://www.dtreg.com (decision tree and SVM predictive modeling)
http://www.nlreg.com (nonlinear regression)
.