<>
See http://www.stata.com/statalist/archive/2008-12/msg00843.html and Cameron
and Trivedi (2009), chap. 17, http://www.stata-press.com/books/mus.html
The cut-off for overdispersion depends on your chosen significance level for
the "Likelihood-ratio test of alpha=0" at the bottom of -nbreg- output in
Stata. What is the "1" you are referring to in your last sentence?
HTH
Martin
-----Ursprüngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Mamopeli
Matooane
Gesendet: Dienstag, 18. August 2009 09:47
An: [email protected]
Betreff: st: Poisson and negative binomial regression
I have count data - time-activity data (dependant variable) and categorical
data e.g. gender, education, employment and others (Independant variables).
I have applied poisson and NBR and get similar results in terms of
coefficients, and p-values. There's a slight improvement in standard errors
with NBR.
Applying NBR does not however eliminate the problem of underdispserion in
the data. How does one address underdispersion and what is the cut off point
from 1 to say there's underdisprsion/overdispersion
--
This message is subject to the CSIR's copyright terms and conditions, e-mail
legal notice, and implemented Open Document Format (ODF) standard.
The full disclaimer details can be found at
http://www.csir.co.za/disclaimer.html.
This message has been scanned for viruses and dangerous content by
MailScanner,
and is believed to be clean. MailScanner thanks Transtec Computers for
their support.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
