table. Also included in the output is a likelihood ratio test statistic (91.33) and
correspondingp-value (0.0000) for a likelihood ratio test comparing the full model
with 8 regression parameters to a reduced model containing only the intercept. The test
statistic follows a chi-square distribution with 8 degrees of freedom under the null.
Theoroption for thelogitcommand is used to obtain exponentiated coefficients
rather than the coefficients themselves. In Stata, options appear in the command
following a comma. The code follows:
logit chd cat age chl ecg smk hpt ch cc, orThelogisticcommand without theoroption produces identical output as thelogit
command does with theoroption. The output follows:
Logit estimates Number of obs ¼ 609
LR chi2 (8) ¼ 91.33
Prob>chi2 ¼ 0.0000
Log likelihood¼173.61476 Pseudo R2 ¼ 0.2082
chd Odds Ratio Std. Err. z P>jzj [95% Conf. Interval]
cat 3.08e-06 9.57e-06 4.09 0.000 7.02e-09 .0013542
age 1.035582 .0167127 2.17 0.030 1.003338 1.068862
chl .9945599 .004161 1.30 0.192 .9864378 1.002749
ecg 1.443587 .4732125 1.12 0.263 .7593048 2.74454
smk 2.166718 .709095 2.36 0.018 1.14086 4.115025
hpt 2.848091 .9445266 3.16 0.002 1.486845 5.455594
ch .0971222 .0721295 3.14 0.002 .0226547 .4163692
cc 1.071618 .0153883 4.82 0.000 1.041878 1.102207
The standard errors and 95% confidence intervals are those for the odds ratio
estimates. As discussed in the SAS section of this appendix, care must be taken in
the interpretation of these odds ratios with continuous predictor variables or inter-
action terms included in the model.
Thevcecommand will produce a variance–covariance matrix of the parameter esti-
mates. Use thevcecommand after running a regression. The code and output follow: