.-
help for ^adjprop^                                    (STB-43: sg33.1)
.-

Adjusted proportions (probabilities)  (JMGarrett 02/04/98)
------------------------------------

   ^adjprop^ yvar [^if^ exp]^, by(^xvar1 [xvar2]^)^ [^a^djust^(^covlist^) m^ode
> l
                   ^l^evel^(^#^) g^raph ^b^ar graph_options]

^adjprop^ calculates and optionally graphs adjusted probabilities (risks) and
   confidence intervals from logistic regression estimates for one or two
   nominal X variables, adjusted for covariates. If second X is specified,
   probabilities are calculated for all possible combinations of X categories,
   and an interaction effect is tested. Default shows estimated probabilities
   and confidence intervals only. Optionally, model estimates and/or a graph
   may be displayed. Dummy variables are created for xvar1. The lowest dummy
   variable defaults to the reference group. If xvar2 is specified, dummy
   variables are created for it also, as well as interaction terms. 
 
Variables and options required
------------------------------

    yvar -- dependent variable  (dichotomous 0,1)

    ^by(^xvar1^)^ -- nominal variable for categories of estimated probabilities

    ^by(^xvar1 xvar2^)^ -- categories of all combinations of xvar1 and
                       xvar2; interaction between xvar1 and xvar2 tested 

        
Options allowed 
---------------

    ^adjust(^covlist^)^ -- any covariates (if none specified, unadjusted
                       probabilities reported); Covariates set to their mean,
                       based on observations used in the analysis, or can be
                       set to user specified values (e.g., age=50 gender=1).

    ^model^ -- shows logistic regression table as well as estimated probabiliti
> es

    ^level(^#^)^ -- specifies the confidence level, in percent, for calculation
                    of confidence intervals (default=95%)

    ^graph^ -- if one X (xvar1), graphs probabilities and confidence intervals;
             if both xvar1 and xvar2 specified, points graphed for each
             probability, but confidence intervals are not graphed; xvar1 is
             used for the x-axis with separate points for categories of xvar2

    ^bar^ -- can be used with the ^graph^ option to display a bar graph instead
            of points  
    

Examples
--------

 . ^adjprop died, by(race) adjust(sys age=50)^
        Calculates the probability (risk) of death by race category,
         adjusted for mean systolic blood pressure and age=50

 . ^adjprop died, by(ses) adjust(sys age) model graph xlabel(1,2,3) ylabel^  
        Calculates probability of death by levels of socio-economic status,
        adjusted for systolic blood pressure and age; displays model and graph

 . ^adjprop died, by(gender race) adjust(age smoke etoh) level(90) ^
        Calculates the probability of death for all combinations of gender
        (2 categories) and race (4 categories) for a total of 8 estimates;
        adjusted for age, smoking status, and alcohol consumption; tests for
        interaction between gender and race; 90% confidence intervals

 . ^adjprop died, by(race gender) adjust(age smoke etoh) graph bar ylabel^
        Calculates adjusted probability of death for all combinations of
        gender and race; bar graph of probability, race on the x-axis


Author
------

      Joanne M. Garrett
      University of North Carolina at Chapel Hill
      garrettj@@med.unc.edu


See also
--------

     STB:  STB-43 sg33.1, STB-24 sg33