Scribd
Upload
3 views

Tut2 Toe

Uploaded by

mashiringwaninyasha2003
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

Tut2 Toe

Uploaded by

mashiringwaninyasha2003
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Tutorial 2 HSTS201

int
March 22, 2019

1. Show that the mean X̄ of a random sample of size n from a distribu-


tion with pdf f (x; θ) = 1/θe−(x/θ) , 0 ≤ x ≤ ∞, zero otherwise, is an
unbiased estimator of θ and has variance θ 2 /n.

Solution

Let f (x; θ) = 1/θe−(x/θ) , 0 ≤ x ≤ ∞, this is an exponential distribution


with mean θ and variance θ 2
Thus,

1X 1 X 1X 1X 1
E(X̄) = E( X) = E( X) = E(X) = θ = nθ = θ
n n n n n

1X 1 X 1 X
V ar(X̄) = V ar( X) = 2 V ar( X) = 2 V ar(X)
n n n
1 X 2 1
= 2 θ = 2 nθ2 = θ/n
n n
2. Let X1 , X2 , . . ., Xn denote a random sample from a normal distribu-
Xi2 /n
P
tion with mean zero and variance θ, 0 ≤ x ≤ ∞, . Show that
is an unbiased estimator of θ and has variance 2θ2 /n.

Solution
Xi2 /σ 2 χ2 (n).
P
It is known that is Hence
X X
E[ Xi2 /σ 2 ] = n and V ar[ Xi2 /σ 2 ] = 2n

1
Thus, X
E[ Xi2 /θ] = n
n X 2
E[ Xi /θ] = n
n
1 X 2
E[ Xi /θ] = 1
n
1 X 2
E[ Xi /n] = 1
θ
X
E[ Xi2 /n] = θ

X
V ar[ Xi2 /θ] = 2n
n2 X
2
V ar[ Xi2 /θ] = 2n
n
1 X
V ar[ Xi2 /θ] = 2/n
n2
1 X
V ar[ Xi2 /n] = 2/n
θ2
X
V ar[ Xi2 /n] = 2θ2 /n

3. Let Y1 ≤ Y2 ≤ Y3 be the order statistics of a random sample of size


3 from the uniform distribution having pdf f (x; θ) = l/θ, 0 ≤ x ≤
∞, 0 ≤ θ ≤ ∞, zero elsewhere. Show that 4Y1 , 2Y2 , and 34 Y3 are all
unbiased estimators of θ. Find the variance of each of these unbiased
estimators.

Solution
It can be show that Y1 /θ, Y2 /θ, Y3 /θ follows a beta distribution with
parameters (α, β )=(1,3), (2,2) and (3,1) respectively. Thus

1
E[4Y1 ] = 4θE[Y1 /θ] = 4θ × =θ
4
Similarly,
1
E[2Y2 ] = 2θE[Y2 /θ] = 2θ × =θ
2

2
4 4 4 3
E[ Y3 ] = θE[Y3 /θ] = θ × = θ
3 3 3 4

3 3
V ar[4Y1 ] = (4θ)2 V ar[Y1 /θ] = (4θ)2 × = θ2
80 5
Similarly,
1 4 1
V ar[2Y2 ] = θ2 & V ar[ Y3 ] = θ2
5 3 15
4. Let X1 , X2 , . . ., Xn denote a random sample from a Poisson dis-
P
tribution with parameter θ , 0 ≤ θ ≤ ∞. Let Y = Xi and let
L[θ, δ(y)] = [θ − δ(y)]. If we restrict our considerations to decision
functions of the form δ(y) = b + θ/n, where b does not depend on y ,
show that R(θ, δ) = b2 + θ 2 /n. What decision function of this form
yields a uniformly smaller risk than every other decision function.

Solution
Since Xi follows Po(θ ), Y follows Po(nθ )
We have E(Y ) = nθ and V ar(Y ) = nθ Thus

R(θ, δ) = E(L[θ, δ(y)]) = E[(θ − b − y/n)2 ] = ........ = b2 + θ2 /n

You might also like