Keith - I would very much like to see a built-in capacity for overdispersed
models in Stata. In fact, just yesterday, I had to write my own kluge to
estimate parameters for a set of overdispersed binomial data.
Al Feiveson
-----Original Message-----
From: [email protected]
[mailto:[email protected]]On Behalf Of Keith Dear
Sent: Monday, August 09, 2004 6:51 PM
To: [email protected]
Subject: st: beta-binomial regression, ebb.ado
Does anyone do this? It seems not to be supported by Stata. Brad Palmquist
wrote the ebb program, but he no longer supports or provides it. It seems
to work fine, but I am wondering:
* should I not be doing beta-binomial regression for some reason?
* should I not be using ebb.ado?
* are there other methods (models) I should be using instead?
Beta-binomial regression can be used to model over-dispersed binomial-like
data (i.e. bounded non-negative integer). I frequently handle such data,
arising from psychological and psychiatric scales -- typically each subject
is asked a number of questions (9 or 12 perhaps) and the outcome variable
to be analysed is the count of "yes" responses. The "e" in ebb allows for
Extension to under-dispersed data.
Thanks
Keith Dear
==========================================================
Dr Keith B.G. Dear
Senior Fellow
* Centre for Mental Health Research (CMHR)
* National Centre for Epidemiology and Popn Health (NCEPH)
The Australian National University
Canberra ACT 0200 Australia
T: +61 2 6125 4865 (NCEPH)
T: +61 2 6125 8439 (CMHR)
F: +61 2 6125 0740
CRICOS Provider #00120C
www.anu.edu.au/nceph/dear
*
* 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/
