Standard errors for the adjusted factor level means from a GLM model

[yiwu ye <yiwu21111958@yahoo.com>]

Dear list,

I fitted the following GLM model,

rating_c(20,24,28,28,15,18,23,24,18,19,24,23,26,26,30,30,22,24,28,26,19,21,27,25)
judge_c(1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4,5,5,5,5,6,6,6,6)
wine_c(1,2,3,4,1,2,3,4,1,2,3,4,1,2,3,4,1,2,3,4,1,2,3,4)

dat_data.frame(rating=rating,judge=judge,wine=wine)
dat$judge_as.factor(dat$judge)
dat$wine_as.factor(dat$wine)

options(contrast=c("contr.treatment","contr.poly"))
glm.1_glm(rating~judge+wine,data="">> summary(glm.1, cor = F)

Call: glm(formula = rating ~ judge + wine, family = gaussian, data = "">Deviance Residuals:
       Min         1Q     Median        3Q      Max
 -1.333333 -0.3333333 -0.1666667 0.6666667 1.666667

Coefficients:
                  Value Std. Error     t value
(Intercept)  23.6666667  0.2108185  112.260857
     judge1  -2.5000000  0.3651484   -6.846532
     judge2  -0.5000000  0.2108185   -2.371708
     judge3   1.5000000  0.1490712   10.062306
     judge4   0.3000000  0.1154701    2.598076
     judge5  -0.1333333  0.0942809   -1.414214
      wine1   1.0000000  0.2981424    3.354102
      wine2   1.8888889  0.1721326   10.973453
      wine3   0.7777778  0.1217161    6.390097

The level means for factor judge would be:
> c(coef(glm.1)[1], coef(glm.1)[1] + coef(glm.1)[2:6])
 (Intercept)   judge1   judge2   judge3   judge4   judge5
    23.66667 21.16667 23.16667 25.16667 23.96667 23.53333
However, I don't know how to calculate standard errors for those means.
Could any one help?

Thanks,

Yiwu

---

Looking for earth-friendly autos? 
 Browse Top Cars by "Green Rating" at Yahoo! Autos' Green Center.