Bayesian mediation analysis (informative prior)

DATA:

file = healthyeating.dat;

VARIABLE:

names = txgroup coworker knowldge intent friends partner;

usevariables = txgroup coworker intent;

missing = all (-99);

ANALYSIS:

estimator = bayes;

bseed = 77463;

fbiterations = 10000;

MODEL:

coworker on txgroup (apath);

intent on coworker (bpath);

intent on txgroup (cpath);

MODEL PRIORS:

apath ~ N(.750,.301);

bpath ~ N(.187,.038);

cpath ~ N(.160,.340);

MODEL CONSTRAINT:

new(ab);

ab = apath * bpath;

OUTPUT:

standardized tech8;

Questions or suggestions? Email Craig Enders