[R] Binary logistic regression with binary predictors

F.j. Berg f.j.beeg at student.rug.nl
Wed May 2 19:07:16 CEST 2012


Ey R-people

Trying to do a binary logistic regression with only 
categorical(age,ballot) and binary predictors for a binary response 
variable.  I can model them at least at least i treat the binary 
predictors as categorical with the use of as.factor(etc) and use glm 
with binomial distribution and logit as link function. The problem is 
with interpretation of results and the use of the Wald-test to see if 
certain predictors are significant. When i do the wald test with 
standard code i see in help i can only use my whole model as input and 
terms 0,1 but then i take it all terms 0,1 are taken.
  Is there a better method for this or maybe a package or maybe an other 
approach?
Any help would be welcomed.

Frank



More information about the R-help mailing list