But I have not heard that maximum likelihood asymptotic results
systematically underestimates the standard error, as seems to be the case
here.
Tim
"Svend Juul" <[email protected]>
Sent by: [email protected]
08/08/2005 14:46
Please respond to
[email protected]
To
<[email protected]>
cc
Subject
Re: st: Question on stir
Tim wrote,
I wonder if I could ask a more theoretical question: I've calculated the
confidence interval using the Poisson regression method as suggested. The
confidence intervals were consistently smaller than the ones given by
stir.
---------------------------------------------
-poisson- calculates an approximate (asymptotic) confidence interval, and
especially if you have few events there can be some discrepancy between
the "exact" interval calculated by -ir- and -stir-, and the result from
-poisson-. The exact interval should, I think, be preferred. But you can
only obtain it in simple situations, not when more independent variables
are involved.
Svend
________________________________________________________
Svend Juul
Institut for Folkesundhed, Afdeling for Epidemiologi
(Institute of Public Health, Department of Epidemiology)
Vennelyst Boulevard 6
DK-8000 Aarhus C, Denmark
Phone, work: +45 8942 6090
Phone, home: +45 8693 7796
Fax: +45 8613 1580
E-mail: [email protected]
_________________________________________________________
*
* 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/
