[R] Best Regression Technique to Use

Michael Friendly friendly at yorku.ca
Mon Mar 14 13:46:51 CET 2016

On 3/13/2016 6:42 PM, Ajay Andrews wrote:
> I have a set of independent variables that are all BINARY, and my dependent
> variable is also BINARY. Should I use the logistic regression for this -
> using the glm function?
> 	[[alternative HTML version deleted]]

glm(..., family=binomial) will give you what you want. When the 
independent variables are all categorical, this is sometimes
called a logit model.  Alternatively, you could also fit
a loglinear model, but the logit form is probably simpler.

More information about the R-help mailing list