How to get p-value from z statistic

[Paul Somwe]

Good Morning,

I am a novice STATA user hence kindly bare with me if my question is basic. I executed the following commands:

mhodds tb hiv2 if missing==0, by(bmigrp) c(1,0)

Maximum likelihood estimate of the odds ratio
Comparing hiv2==1 vs. hiv2==0
by bmigrp

-------------------------------------------------------------------------------
   bmigrp | Odds Ratio        chi2(1)         P>chi2       [95% Conf. Interval]
----------+--------------------------------------------------------------------
      <19 |   8.000000          16.53         0.0000         2.42589   26.38208
    19-24 |   2.311694           5.60         0.0179         1.13161    4.72239
      25+ |   3.213415           3.89         0.0486         0.94169   10.96542
-------------------------------------------------------------------------------

    Mantel-Haenszel estimate controlling for bmigrp
    ----------------------------------------------------------------
     Odds Ratio    chi2(1)        P>chi2        [95% Conf. Interval]
    ----------------------------------------------------------------
       3.256096      21.22        0.0000         1.912757   5.542868
    ----------------------------------------------------------------

Test of homogeneity of ORs (approx): chi2(2)   =    3.22
                                     Pr>chi2   =  0.1996

I would like to get the Pr>chi2 value 0.1996 and populate it into an excel cell but am facing difficulties. The return list command does not show the actual Pr>chi2 value but shows I think the z score (i.e. r(chi2_hom))

. return list

scalars:
                 r(or) =  2.796406651543788
              r(lb_or) =  1.71732962901426
              r(ub_or) =  4.55351729142816
               r(chi2) =  18.64835958570303
                  r(p) =  .0000157182454007
             r(df_hom) =  1
           r(chi2_hom) =  .068043938660382

Any help will be greatly appreciated.

Thanks,

Paul

[Joseph Coveney]

You're better off posting questions like this over in the main Stata forum. This one's for Mata-related questions.

Anyway,

display in smcl as result %06.4f chi2tail(r(df_hom), r(chi2_hom))

is probably what you're looking for.

help chi2tail()

for more information.

[Paul Somwe]

Hi Joseph,

Am new to this forum, nevertheless I have taken note of your guidance. Thanks for you help, your code worked, very much appreciated.

Thanks,

Paul