Dear all,
In lme, models in which a factor is fully contained in another lead to
an error. This is not the case when using lm/aov.
I understand that these factors are aliased, but believe that such
models make sense when the factors are fitted sequentially. For example,
I sometimes fit a
Pascal A. Niklaus pascal.niklaus at ieu.uzh.ch writes:
In lme, models in which a factor is fully contained in another lead to
an error. This is not the case when using lm/aov.
I understand that these factors are aliased, but believe that such
models make sense when the factors are fitted
2 matches
Mail list logo