Re: st: a question about -desmat-

[Joan Holand <Joan.Holand@gmx.net>]

Hello Maarten,

thank you very much for your answer.

Actual, the effect differences came from the different coding schemes.
After several tries I can now say: my experience is that the results are
being correctly reproduced by -desmat- as far as the deviation coding
and indicator coding is concerned. Concerning the backward coding for
one variable in conjunction with deviation coded variables (with the
highest category as the reference) I found a few divergences and I still
do not know, what is the reason for them (e.g. the frequency
distributions of the manually and automatically produced dummies are
identic, but the respective coefficients are not! ?).

Thank you anyway,
Cheers
Joan


Maarten buis writes:
> --- Joan Holand <Joan.Holand@gmx.net> wrote:
> > I used -desmat- as a prefix (one can use it as a command, too) and 
> > then compared the output with the results of my manually produced 
> > model. The both results (including the algebraic sign of some 
> > effects) are different! The log likelihood value, the Pearson Chisq
> >  and the AIC value is identic but the deviance and BIC is higher in
> >  the desmat output than in the manual model.
> -desmat- uses the first category in each variable as the reference, 
> did you do the same for your manually produced dummies? If not then 
> that would explain the different parameter estimates but the same log
>  likelihood. Are the differences in BIC and deviance large, or so 
> small that they could be rounding errors? Hope this helps, Maarten
>
>
>
>
>
>
> ----------------------------------------- Maarten L. Buis Department
>  of Social Research Methodology Vrije Universiteit Amsterdam
> Boelelaan 1081 1081 HV Amsterdam The Netherlands
>
> visiting address: Buitenveldertselaan 3 (Metropolitan), room Z434
>
> +31 20 5986715
>
> http://home.fsw.vu.nl/m.buis/ 
> -----------------------------------------
>
>
> ___________________________________________________________ All New 
> Yahoo! Mail – Tired of unwanted email come-ons? Let our SpamGuard 
> protect you. http://uk.docs.yahoo.com/nowyoucan.html * *   For 
> searches and help try: * 
> http://www.stata.com/support/faqs/res/findit.html * 
> http://www.stata.com/support/statalist/faq * 
> http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/