[R] anova on binomial responses data

[Stephane Dupas]

Hi ! how to perform anova on binomial response data using glm procedure, 
could anyone help us with this problem ? we are trying to test the a 
model of binomial response with non-ordered categorical factors. I am 
not getting the significance of each factors but, the departure from the 
binomial model within each category.

For example : Two factors, A and B, one binomial matrix response Ymat
I want to get the effect of F1, F2, F1*F2
If I test the model Ymat ~ F1*F2, with glm :

fml <- glm(Ymat ~ F1*F2, family = binomial(link=probit), data = data.df)
summary(fml)

I get levels of significance within each level of the different factors 
and all their interactions, I cannot figure out how to get the effect of 
the factors themselves and their interaction on the binomial response.

something like that:

Coefficients:
            Estimate Std. Error z value Pr(>|z|)  
(Intercept)  0.21972    0.18642   1.179  0.23854  
F1b          0.16444    0.22118   0.743  0.45719  
F1c         -0.06241    0.24396  -0.256  0.79809  
F1d          0.16560    0.23554   0.703  0.48203  
F1e          0.11228    0.21025   0.534  0.59331  
F2b         -0.02052    0.34442  -0.060  0.95249  
F2c          0.27298    0.27000   1.011  0.31200  
F2d          0.21100    0.32371   0.652  0.51451  
F2e          0.97407    0.31176   3.124  0.00178 **
F2f          0.25457    0.23411   1.087  0.27687  
F1b:F2b      0.15497    0.38003   0.408  0.68343  
F1c:F2b      0.22932    0.42740   0.537  0.59158  
F1d:F2b      0.43980    0.40695   1.081  0.27982  
F1e:F2b      0.46955    0.39571   1.187  0.23539  
F1b:F2c     -0.17764    0.36336  -0.489  0.62492  
F1c:F2c      0.43913    0.42142   1.042  0.29740  
F1d:F2c     -0.09235    0.58819  -0.157  0.87524  

etc...

do you how can I get the effect of the factors themselves, not the 
departure from binomial model within each catagory ?

Thanks for your help,

Stéphane

-- 

Stéphane DUPAS

IRD c/o CNRS
Lab P.G.E., Bat 13,
1 av terrasse
91198 Gif sur Yvette
tel 33 1 69 82 37 04


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._