0

I have a glm that produces different levels within my summary, I want to use specific levels from this summary as one of my factors.

My code is

Model3<-glm(status ~ month + age + centre + treatment, family="binomial")
summary(Model3)

This comes out with 

Coefficients:
         Estimate Std. Error z value Pr(>|z|)    
(Intercept) -0.487717   0.315126  -1.548  0.12170    
month1       0.606370   0.286407   2.117  0.03425 *  
month2       0.402645   0.284686   1.414  0.15726    
month3       0.606370   0.286407   2.117  0.03425 *  
month4       0.362311   0.284486   1.274  0.20282    
age         -0.020200   0.006931  -2.915  0.00356 ** 
centre2      1.090830   0.189501   5.756 8.60e-09 ***
treatment1   0.822554   0.181614   4.529 5.92e-06 ***

I want to somehow be able to remove month2 from my next glm but I don't see how this is possible without removing them all.

Joshua Ulrich
  • 173,410
  • 32
  • 338
  • 418
Daniel
  • 1
  • 1
  • Maybe the solution to the question posted [here](http://stackoverflow.com/questions/20913379/removing-specific-factor-level-from-factor-variable) is relevant? – horseoftheyear Nov 02 '15 at 21:32
  • I feel like the solution is really relevant but I tried to apply it to my coding and kept getting errors. Will keep trying but perhaps I'm most likely not using it right – Daniel Nov 02 '15 at 21:54
  • If you could include a small [reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example), others can have a crack at is well. Increasing your chances of a solution. – horseoftheyear Nov 02 '15 at 22:46
  • Additionally, the last answer to [this question](http://stackoverflow.com/questions/12326357/how-do-you-remove-an-insignificant-factor-level-from-a-regression-using-the-lm) might be of help. – horseoftheyear Nov 02 '15 at 23:00

0 Answers0