I am a begginer with "R" and am fitting a multiple logistic regression using the command "glm". I am testing the effect of "Species" (categorical factor, 5 levels) and two continuous regressors (forest complexity indices: Index1, Index2) on the mortality of mangrove seedlings, a binary response (0=survival, 1=death).
I am using:
glm(y~Species+Index1+Index2, family=binomial)
In the summary results I obtain coefficient estimates and associated statistics for the two continuous regressors, but in the case of the categorical factor I don't get estimates for the factor "Species", instead I obtain estimates for four of the species separately. Why R do this? can anyone help me with this?
Try
if(!require(car)){install.packages("car")}
model = glm(y~Species+Index1+Index2, family=binomial)
library(car)
Anova(model)
And read
library(car); ?Anova
When you run a logistic regression with a categorical explanatory variable, the estimated coefficients will show the change in the log-odds of each category versus one (fixed) other category. See this for example.
http://statistics.ats.ucla.edu/stat/r/dae/logit.htm
- Badges
- Science topic