Hello:
I am using the GEE library under S-Plus 2k Prof, Rel 3. Is there a customary
way of obtaining a test for effects with several df in GEE? In my particular
case, I have 3 factors (say A, B, and C) with 2, 2, and 4
levels respectively. The C main effect and the A:C, B:C, and A:B:C interactions
each have 3 df hypothesis. The standard ouput breaks them down into single df
parameter estimates with associated robust z values.
I was thinking of summing the squared robust z values for each 3 df effect and
then using a Chi-Square test, but I am not sure if that is a good idea (for an
example of what I have in mind, see below). Any advice or references would be
appreciated.
thanks,
Chuck
Coefficients:
Estimate Naive S.E. Naive z Robust S.E. Robust z
(Intercept) 0.72229224 0.07026348 10.279768 0.07037511 10.263462
C1 -0.12440687 0.07928408 -1.569128 0.08133938 -1.529479
C2 0.07290186 0.06045054 1.205975 0.05705350 1.277781
C3 -0.06260184 0.04541103 -1.378560 0.04714258 -1.327926
Estimated Scale Parameter: 1.001538
Number of Iterations: 2
> (-1.529479)^2 + 1.277781^2 + (-1.327926)^2
[1] 5.735418
> 1 - pchisq(5.735,3) # p-value for stat
[1] 0.1252395
http://www.insightful.com/
http://lib.stat.cmu.edu/DOS/S/gee.zip
-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-
Chuck Cleland
Institute for the Study of Child Development
UMDNJ--Robert Wood Johnson Medical School
97 Paterson Street
New Brunswick, NJ 08903
phone: (732) 235-7699
fax: (732) 235-6189
http://www2.umdnj.edu/iscdweb/
http://members.nbci.com/cmcleland/
-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
