Bayesian ANCOVA (non-informative prior)

DATA:

file = probsolve.dat;

VARIABLE:

names = id school tx frlunch stanmath pretest probsolv;

usevariables = tx frlunch stanmath pretest probsolv;

missing = all (-99);

centering = grandmean(frlunch stanmath pretest);

ANALYSIS:

estimator = bayes;

bseed = 32504;

fbiterations = 10000;

MODEL:

probsolv on tx pretest frlunch stanmath;

! IV variances (necessary with incomplete predictors);

tx pretest frlunch stanmath;

! IV covariances (necessary with incomplete predictors); 

tx pretest frlunch stanmath with tx pretest frlunch stanmath;

OUTPUT:

standardized tech8;

PLOT:

type = plot2;

Questions or suggestions? Email Craig Enders