[R] logistic regression

Kevin E. Thorpe kevin.thorpe at utoronto.ca
Sat Sep 27 14:21:22 CEST 2008

Darin Brooks wrote:
> Good afternoon
> I have what I hope is a simple logistic regression issue.
> I started with 44 independent variables and then used the drop1,
> test="chisq" to reduce the list to 8 significant independent variables.  
> drop1(sep22lr, test="Chisq") and wound up with this model:
> Model: MIN_Mstocked ~ ORG_CODE + BECLBL08 + PEM_SScat + SOIL_NUTR +
> cSEEDSRCE_SW + cMSP + ceFFP + cEXT_Cold
> 4 of the remaining variables are categorical and 4 are continuous.
> However, when I run a glm and then a summary on the glm - some of the
> categorical data is missing from the output.  
> The PEM_SScat is missing only one variable ... the BECLBL08 is missing
> several variables ... the ORG_CODE is missing 4 .. and the SOIL_NUTR is
> missing 1 variable.
> It seems arbitrary to the number of variables missing.  Is there something
> wrong with my syntax in calling the logistic model?  Am I not understanding
> the inputs correctly?   
> Any help would be appreciated.

I'm not sure I fully understand your question.  It sounds like you
created your own dummy variables for your categorical variables. Did
you?  Or did you use factor variables for your categorical variables?
If the latter, then I REALLY don't understand your question.


Kevin E. Thorpe
Biostatistician/Trialist, Knowledge Translation Program
Assistant Professor, Dalla Lana School of Public Health
University of Toronto
email: kevin.thorpe at utoronto.ca  Tel: 416.864.5776  Fax: 416.864.6057

More information about the R-help mailing list