.- help for ^adjprop^ [STB-24: sg33] .- Adjusted proportions -------------------- ^adjprop^ yvar [^if^ exp]^,^ ^by(^xvar^)^ [^adj^ust^(^covlist^)^ ^m^odel ^p^lot graph_options] ^adjprop^ calculates and optionally plots adjusted proportions and 95% confidence intervals from logistic regression estimates. By default, only the estimated proportions and confidence intervals are displayed. Optionally, the model estimates and/or a plot may be displayed. ^yvar^ is a dichotomous variable. Dummy variables are created from xvar and renamed x1, x2, etc. The highest category defaults to the reference group. Options ------- ^adjust(^covlist^)^ specifies the covariates. If this option is omitted, unadjusted estimates will be reported. ^by(^xvar^)^ specifies the nominal variable that defines the categories for the estimated proportions. This option is required. ^model^ requests the display of the logistic estimates ^plot^ requests a plot of proportions along with their 95% confidence intervals Examples -------- . ^adjprop died, by(race) adjust(sys age)^ Calculates the proportion of deaths by race category, adjusted for systolic blood pressure and age . ^adjprop died, by(ses) adjust(sys age) model l2(Proportion of Deaths)^ ^plot xlabel ylabel^ Calculates the proportion of deaths by levels of socio- economic status, adjusted for systolic blood pressure and age; displays model and plot . ^adjprop died, by(hosp) adjust(age race sex) plot xlabel(1,2,3,4,5,6) ylabel^ Calculates the proportion of deaths for 6 hospitals, adjusted for age, race, and sex; displays plot Note ---- Observations with missing values for yvar, xvar, or any covariates are dropped from the analysis before estimation. Covariates are set to their mean based on observations used in the analysis. Author ------ Joanne Garrett, University of North Carolina FAX 919-966-2274 Also see -------- STB: sg33 (STB-24)