Inference for theta = P(heads) with binomial data.

1) Using a beta(5,5) prior.

Doodle for conjugate inference for a coin tossing problem.

BUGS code


model;
{
   theta ~ dbeta(a,b)
   y ~ dbin(theta,n)
   y.pred ~ dbin(theta,n.pred)
}

DATA

list(a=5,b=5,n=12,y=7,n.pred=10)


2) Inference with a truncated normal prior

model;
{
   theta ~ dnorm(mu,prec)I(0,1)
   y ~ dbin(theta,n)
   y.pred ~ dbin(theta,n.pred)
}

DATA

list(mu=0.5,prec=1,n=12,y=7,n.pred=10)