Re: GLM vs unbalanced designs

[John Darrington]

On Sun, Nov 13, 2011 at 04:04:24PM +0000, John Darrington wrote:
          
          I just checked Neter, Wasserman and Kutner's "Applied Linear
          Statistical Models", and I think you are right. The model PSPP uses
          has different terms in its model by virtue of dummy coding. The
          discrepancy could be fixed by adding the "effects coding" to
          categoricals.c.
          
     I did try that, but got completely way out results.  
     
     Perhaps I made an error somewhere - I'll have another go this week, and 
     I'll be a bit more careful about it.
     
As a first step, I've pushed a change which adds an undocumented /SHOWCODES
subcommand to GLM.  It gives a table of the dependendent variables and the
corresponding factor encodings.  

Hopefully this will help in the debugging process.

J'



-- 
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://pgp.mit.edu or any PGP keyserver for public key.

signature.asc
Description: Digital signature