R help - Oct 2004 - question about df in linear mixed model function lme

I have looked through Pinheiro and Bates (2000) but have not found an
explanation, so I post my question to the group.  Pages 90-92 explain
how the denominator df are calculated, but not why.  Consider a mixed
model with two predictor variables (Y, Z) in which Y only varies between
groups while Z varies at both levels.  The denominator df for the test
of the marginal effect of Y is based on the number of groups.  The
denominator df for the marginal effect of Z is based on the total number
of observations.  I understand the logic of both.  However, why is the
denominator df for the interaction term (Y:Z) also based on the total
number of observations?

 

Bill Shipley

Subject Matter Editor, Ecology

North American Editor, Annals of Botany

Département de biologie, Université de Sherbrooke,

Sherbrooke (Québec) J1K 2R1 CANADA

Bill.Shipley@USherbrooke.ca

 <http://callisto.si.usherb.ca:8080/bshipley/>
http://callisto.si.usherb.ca:8080/bshipley/

 


	[[alternative HTML version deleted]]