As a follow-up comment to my email for Stata Corp, I received an email from
Juul Svend indicating that Juul had emailed Stata about these points and
others. One point that Juul emphasized was to include the count in the
output, I want also to emphasize this. Although one can recover the count
from r(N), this cannot be done if ci is run with the -bys- prefix. These
would be very useful fixes.
Thanks,
Fred
At 05:58 AM 8/14/2004, you wrote:
-ci- displays exposure time as an integer. This is not always desirable,
as in the instance when exposure time has been divided by 1000 so that
output will be in 1000 person-years.
In the example below, the only way to get exact exposure time is to run
the analyses by years and by 1000 years.
Perhaps Stata Corp might consider presenting exposure in more details than
just integer. Also useful would be to add a format option to format the #
of decimal places overall.
Even more useful would be an option to express the output with different
divisors so that it would be presented in persons-units, 100 person units,
1000 person units, etc.
-- Poisson Exact --
Variable | Exposure Mean Std. Err. [95% Conf. Interval]
-------------+---------------------------------------------------------------
pulmhosp2 | 7 40.35625 2.3698 35.8447 45.2785
Variable | Exposure Mean Std. Err. [95% Conf. Interval]
-------------+---------------------------------------------------------------
pulmhosp2 | 7186 .0403562 .0023698 .0358447 .0452785
Thanks,
Fred Wolfe
Fred Wolfe
National Data Bank for Rheumatic Diseases
Wichita, Kansas
Tel (316) 263-2125 Fax (316) 263-0761
[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/
Fred Wolfe
National Data Bank for Rheumatic Diseases
Wichita, Kansas
Tel (316) 263-2125 Fax (316) 263-0761
[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/
