Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: dealing with 0 in cells in 2x2 tables


From   "Beynon, Caryl" <C.M.Beynon@ljmu.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: dealing with 0 in cells in 2x2 tables
Date   Wed, 9 Jul 2008 11:05:34 +0100

Thanks for your help. I seem to get a different output to yours - any suggestion why?
(Using exlogistic depvar indepvars)


. exlogistic spoonB_vis_writ homeless2

[iterations snipped]

Exact logistic regression                        Number of obs =       163
                                                 Model score   =  3.943866
                                                 Pr >= score   =    0.0846
---------------------------------------------------------------------------
spoonB_vis~t | Odds Ratio       Suff.  2*Pr(Suff.)     [95% Conf. Interval]
-------------+-------------------------------------------------------------
   homeless2 |   5.628941*         22      0.0645      .9202888       +Inf
---------------------------------------------------------------------------
(*) median unbiased estimates (MUE)


I actually want to model the joint effects of a number of independent variables (those significantly associated to the dependant variable in the bivariate analyses). I believe I need the terms() command. 

I typed:

. exlogistic spoonB_vis_writ inj_withwho2  ever_NEX2  evershared3  sex_other_user2 homeless2, terms(t1= inj_withwho2  ever_NEX2  evershared3  sex_other_user2 homeless2)

And I got this message: 

note: P-values cannot be computed for the joint tests of the terms() option using the default method of sufficient statistics; to obtain p-values replay with either test(score) or test(probability)

Despite increasing my memory, there was insufficient to carry out the command - clearly because I am trying to model the joint effects of too many variables. 

Can anyone suggest how I derive a model which provides individual estimates for each variable, adjusted for the others, when I have this data?

Thanks


*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index