I would like to use the Sequential Sum of Squares test. But the degree of freedom for the denominator is (n - p - 1), where n = number of samples, and p = number of variables in the full model.
What do I do when p > n?
I would like to use the Sequential Sum of Squares test. But the degree of freedom for the denominator is (n - p - 1), where n = number of samples, and p = number of variables in the full model.
What do I do when p > n?