input
command (such as the following for EMS Table 18.2)
clear input gender area antibody freq 1 1 1 36 1 1 0 14 1 0 1 50 1 0 0 50 0 1 1 24 0 1 0 126 0 0 1 10 0 0 0 90 end
mhodds antibody area gender [weight=freq], compare(1,0)
will generate the results in EMS 18.4. The first variable after the command mhodds
is the outcome; the second variable is the risk factor (the variable of interest); the third variable is the stratifying variable. (Note that logistic regression logistic antibody area gender [weight=freq]
will generate similar, but different, results.)
cc antibody area [weight=freq], by(gender)
generates the results in EMS 18.5. (Note that test of heterogeneity is equivalent to test of interaction effect between area
and gender
.)
logistic
or logit
. For example, logistic mf area
(or logit mf area, or
) generates output in odds scale (EMS 19.2). logistic mf area, coef
(or logit mf area
) generates output in parameter scale (i.e. log(odds) scale). The first variable after the command logistic
(or logit
) is the outcome variable, then followed by input variable(s).
xi: logistic mf i.agegrp
. To analyze area and age group simultaneously, xi: logistic mf area i.agegrp
generates the results in EMS 20.2.
mcci 184 54 14 63
will generate the McNemar's test result and OR estimate in EMS 21.2-3, but its CI calculations are based on formulae slightly different from those in EMS.
cii 873 57, poisson
will generate results in EMS 22.5-6, with numbers in count per child-year, not per 1000 child-years. In the latter scale, the CI will be from 49.45 to 84.59, slightly different from that in EMS 22.6, because the calculations are based on formulae slightly different from those in EMS.
iri 33 24 355 518
gives results in EMS 23.2.