CONCEPTS OF BAYESIAN DATA ANALYSIS

June 2011

Name: ...............................................................

1. Suppose we are given the twelve observations from a normal distribu-

tion:
15.644, 16.437, 17.287, 14.448, 15.308, 15.169,
18.123, 17.635, 17.259, 16.311, 15.390, 17.252.

and we are told that the variance is σ 2 = 1.

(a) Find a 90% HPD interval based on the posterior distribution of

the mean, assuming a normal prior with mean 10 and variance 25.
(b) What is the predictive distribution for a possible future observa-
tion x?
2. Suppose that you have a prior distribution for the probability π of
success in a certain kind of gambling game which has mean 0.4 and
that you regard your prior information as equivalent to 12 trials. You
then play the game 25 times and win 12 times. What is your posterior
distribution for π?
3. Explain the Bayesian p-value.
4. When is a 95% credibility interval equal to a 95% HPD interval?
5. Suppose that you have prior beliefs about an unknown quantity θ which
can be approximated by a normal distribution N (µ1 , σ12 ), while my be-
liefs can be approximated by a normal distribution N (µ2 , σ22 ). Suppose
further that the reasons that have led us to these conclusions do not
overlap with one another. What distribution should represent our be-
liefs about θ when we take into account all the information available
to both of us?
6. Consider the scaled inverse χ2n distribution
−a
 
p(θ) ∝ θ−n/2 exp
2θ
and suppose we wish to simulate draws from this distribution using
Metropolis-Hastings algorithm, for n = 5 and a = 4. Suppose we take
1
as our proposal distribution q(θ1 , θ2 ) = q(θ1 ) = 100 (for 0 ≤ θ ≤ 100),
a uniform distribution on (0, 100). We take θ0 = 1 as starting value,
and the uniform distribution returns a candidate value of θ∗ = 39.82.
What is the probability to accept this value θ∗ as a sample from the
distribution p(θ)?