[R] pairwise comparisons with multcomp package

[gaiarrido]

Hi,
I've got this model and following Hothorn et al advices, I used glht for a
post hoc comparison


> modezqM<-glm(rojos~estacion*zona3,quasipoisson,subset=(edadysexo=="M"))
> anova(modezqM,test="F") 
                                 Df      Deviance      Resid. Df     Resid.
Dev       F                   Pr(>F)    
NULL                    293         41148                      
estacion                  1         3647.0           292             37501       
23.6250          1.927e-06 ***
zona3                      2           931.4           290             36569         
3.0167           0.050510.  
estacion:zona3      2         2421.7           288            34148         
7.8438            0.000482 ***	


> summary(glht(modezqM,linfct=mcp(zona3="Tukey",interaction_average=T))) 
                                   Estimate       Std. Error      z-value      
Pr(>|z|)  
norte - turist == 0     -0.05182    0.23701      -0.219             0.9727  
occid - turist == 0     -1.05970    0.41618      -2.546             0.0271 *
occid - norte == 0     -1.00788    0.45830      -2.199             0.0665 .

I'm trying to understanding the way this command made pairwise comparisons,
but, where does this z-value comes from?  
I'm suspected this not come from Tukey's test, but i don't know it's Wald or
something i've never use
I know it is not an habitual doubt.
Sorry and thanks in advance

I suspect it


-----
Mario Garrido Escudero
PhD student
Dpto. de Biología Animal, Ecología, Parasitología, Edafología y Qca. Agrícola
Universidad de Salamanca
--
View this message in context: http://r.789695.n4.nabble.com/pairwise-comparisons-with-multcomp-package-tp4384499p4384499.html
Sent from the R help mailing list archive at Nabble.com.