Bayesian Methods in Health Economics

Introduction to Bayesian inference 39

a suitable probability function, in order to describe the uncertainty, or in

other terms the state of the available knowledge on them. This distribution is

indicated by p(θ).

For the observed variables (the data), imperfect knowledge depends es-

sentially on the sampling selection. Typically we assume that the data y =

{y

1

,...,y

n

} are a random sample of observations from a reference population,

made up by N>nunits. The uncertainty on the observed variable is given

by the fact that we only observe a single sample (sampling variability).

Conversely, the uncertainty about parameters is generated by our igno-

rance, or ...