B.

Math II – Statistics-II – Assignment 6

1. A large shipment of parts is received, out of which 5 are tested for defects.
The number of defective parts in the sample, X is assumed to be Binomial
with parameter θ. From past shipments it is known that θ has a Beta(1, 9)
distribution.
(a) Find the HPD estimate of θ if x = 0 is observed.
(b) Find a 95% credible set for θ if x = 0 is observed.
(c) Interpret the results of both (a) and (b).
(d) For testing H0 : θ ≤ 0.10 versus H1 : θ > 0.10, find the posterior odds ratio.

2. Let X1 , . . . , Xn be a random sample from U (0, θ). Let

π(θ) ∝ θ−(a+1) , θ > θ0 ,

where a > 0 and θ0 > 0 are given. Find the HPD estimate of θ and the 100(1-
α)% HPD credible set for θ. Also find the posterior mean and the posterior
variance of θ.

3. Let X1 , X2 , . . . be i.i.d. observations from some population with mean µ and

finite variance σ 2 > 0. Let Yi = X2i−1 , i = 1, 2, . . . and Zi = X2i , i = 1, 2, . . ..
For n ≥ 2, define

1X 1 X 1 X
n [m] [m]
Tn(1) = Xi , Tn(2) = Yi , Tn(3) = Zi ,
n i=1 [m] i=1 [m] i=1

where [m] is the integer part of n/2.

(j)
(a) Show that Tn , j = 1, 2, 3 are all consistent estimators of µ.
(b) Show that these are asymptotically normal.
(c) Find the relative efficiencies of these estimators.

4. Let X ∼ Poisson(λ), λ > 0. Define Y = 1 when X > 0, and 0 otherwise.

(a) Find the Fisher information on λ (say, I (X) (λ) and I (Y ) (λ), respectively)
contained in X and Y individually.
Consider a random sample X1 , X2 , . . . , Xn from the X population and the cor-
responding Y1 , Y2 , . . . , Yn .
(b) Construct consistent estimators T1 (X1 , X2 , . . . , Xn ) and T2 (Y1 , Y2 , . . . , Yn )
for λ based on the MLE of λ from the two data sets.
(c) Find the asymptotic distributions of T1 and T2 .
(d) Find the asymptotic relative efficiency of of T1 w.r.t. T2 .