On 28 May 2003 at 8:26, Paul R Swank wrote: > This is another good illustration why one should not use automatic > procedures for research. The problem of interactions in stepwise procedures > is that the compute doesn't know it is an interaction and that you can't > drop the main effects while the interaction is strill in the model.
Not to promote automatic procedures, but this is not a problem with stepwise per se, it is a problem with inadequate implementations of stepwise. You can tell the computer about marginality relations, and some implementations do that. Look for stepAIC in package MASS for R/S-Plus. Kjetil Halvorsen . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
