[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: xi: sw poisson, p-values of variables rather than categories of them
At 15:57 12/12/02 +0100, A.M. Vos wrote:
If you want to do chi-squared tests, then use -testparm- to test multiple
dummy variables at the same time. Type -help test- and/or read -[R] test-
to find out more. For instance, you might type
Now here's another question about my poisson regression.
again, I'm performing something like:
-xi: sw poisson patient i.origin i.age i.durstay etc... forward-
than what I get is a list of p-values that are below -pe- of CATEGORIES
included in the model. (e.g. _Iorigin_2, _Iorigin10, _Iage1, _Iorigin7
But what I would like to know is which VARIABLES (origin, age, durstay,
etc) are significant, rather than values (representing categories) of
to do a chi-squared test for origin.
However, it must be stressed that there are issues involved here other than
simply using the right Stata command. For instance, if you do a lot of
significance tests, then you run the risk of selecting the 5% of results
that are significant at the 5% level. I don't know much about the
background to Annelies's project. However, quite often, if a customer wants
to calculate P-values, then there is often a confidence interval that the
customer really wanted.
I hope this helps.
Lecturer in Medical Statistics
Department of Public Health Sciences
King's College London
5th Floor, Capital House
42 Weston Street
London SE1 3QD
Tel: 020 7848 6648 International +44 20 7848 6648
Fax: 020 7848 6620 International +44 20 7848 6620
or 020 7848 6605 International +44 20 7848 6605
Opinions expressed are those of the author, not the institution.
* For searches and help try: