[R] interpreting the output of a glm with an ordered categorical predictor.

kmuller katherine.muller2010 at gmail.com
Sat Mar 3 02:10:25 CET 2012


I'm a Master's student working on an analysis of herbivore damage on plants.
I have a tried running a glm with one categorical predictor (aphid
abundance) and a binomial response (presence/absence of herbivore damage).
My predictor has four categories: high, medium, low, and none. I used the
"ordered" function to sort my categories for a glm.

ah <-
ah1<- ah[ah$date=="110810",] 

aphidOrder <- ordered(ah2$aphidLevelMax,levels=c("none", "low", "med",

ordAph <- glm(chewholebinom~aphidOrder,family=binomial,data=ah2)

When I ran the summary for the glm (output pasted below), I could not tell
which intercept referred to which factor level. My question is, what do .L,
.Q, and .C mean and how can I relate these factors to my original factors
(none, low, med, high)?

Thank you for your help,



glm(formula = chewholebinom ~ aphidOrder, family = binomial, 
    data = ah2)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-1.6512  -0.9817   0.7687   0.7687   1.5353  

             Estimate Std. Error z value Pr(>|z|)   
(Intercept)  -0.05567    0.25097  -0.222   0.8245   
aphidOrder.L -1.36755    0.49366  -2.770   0.0056 **
aphidOrder.Q  0.36824    0.50195   0.734   0.4632   
aphidOrder.C -0.09840    0.51011  -0.193   0.8470   
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 137.99  on 99  degrees of freedom
Residual deviance: 124.05  on 96  degrees of freedom
AIC: 132.05

Number of Fisher Scoring iterations: 4

View this message in context: http://r.789695.n4.nabble.com/interpreting-the-output-of-a-glm-with-an-ordered-categorical-predictor-tp4440383p4440383.html
Sent from the R help mailing list archive at Nabble.com.

More information about the R-help mailing list