[R] dummy variable encoding

news at aleblanc.cotse.net news at aleblanc.cotse.net
Thu Mar 5 15:38:18 CET 2009


Hi,
   can anyone tell me why an encoding of 1/2 for a dummy variable for
   two groups (e.g. gender) seems to be preferred over 0/1?
   It's been bugging me for a while, 0/1 seems more natural, but I have
   been told (without explanation) that 1/2 is better. Why?

-- 
aleblanc




More information about the R-help mailing list