Hello,
are there potential drawbacks in the use of rank (ordinal) variables as
predictors in a "linear" mixed model for continuous dependents?
I am trying this approach to create mixed effects models independent of the
(monotonic) transform of the independent variables. The use of rank
transformed independents seems the logical choice to me.
Thank you for any feedback,
Bruno
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Bruno L. Giordano, Ph.D.
CIRMMT
Schulich School of Music, McGill University
555 Sherbrooke Street West
Montr?al, QC H3A 1E3
Canada
http://www.music.mcgill.ca/~bruno/