Exercises 7
Uploaded by
danExercises 7
Uploaded by
danESTIMATORS Exercises
Answers
[8.12]
a.
1 1
For the uniform distribution E(Yi ) = θ + 2 and so E(Ȳ ) = θ + 2 and the bias is
E(Ȳ ) − θ = 21 .
b.
An unbiased estimator is Ȳ − 12 .
c.
1
The variance of Ȳ is 12n and so
MSE(Ȳ ) = V (Ȳ ) + B 2
1 1
= +
12n 4
[8.13]
a.
Y Y 1
E n 1− = E(Y ) − E(Y 2 )
n n n
1
V (Y ) + E 2 (Y )
= np −
n
= np − p(1 − p) − np2
= (n − 1)p(1 − p)
b.
An unbiased estimator is
n Y Y
n 1−
n−1 n n
1
ESTIMATORS Exercises
[9.3]
a.
1 1 1
E(θ̂1 ) = E(Ȳ − ) = (θ + ) − = θ
2 2 2
1
V (θ̂1 ) = V (Ȳ ) =
12n
The probability density function of Y(n) is given by
fn (y) = nF n−1 (y)f (y)
= n(y − θ)n−1 θ ≤y ≤θ+1
and by the method of transformations the variable Z(n) = Y(n) − θ has a beta
distribution, B(n, 1).
n n n
E(θ̂2 ) = E Z(n) + θ − = +θ− =θ
n+1 n+1 n+1
n n
V (θ̂2 ) = V Z(n) + θ − = V (Z(n) ) =
n+1 (n + 2)(n + 1)2
and so θ̂1 and θ̂2 are unbiased and the efficiency of θ1 relative to θ̂2 is
n
V (θ̂2 ) (n+2)(n+1)2 12n2
= 1 =
V (θ̂1 ) 12n
(n + 2)(n + 1)2 .
2
ESTIMATORS Exercises
[9.74]
θ
2
Z
E(Y ) = 2 (θy − y 2 ) dy
θ 0
θ
2 h θy 2 y 3 i
= 2 −
θ 2 3 0
θ
=
3
θ̂
and setting the sample mean equal to the population mean gives Ȳ = 3 or θ̂ = 3Ȳ .
[9.85]
a.
The likelihood function is
n
Y 1
L= α
yiα−1 e−yi /θ
i=1
Γ(α)θ
n
1 P
yi /θ
Y
= n nα
e −
yiα−1
Γ (α)θ i=1
and P n
yi Y
ln L = −n ln Γ(α) − nα ln θ − + ln yiα−1
θ i=1
Differentiating with respect to θ gives
P
∂ ln L yi nα
= 2 −
∂θ θ θ
and equating the derivative to zero gives
Ȳ
θ̂ = .
α
3
ESTIMATORS Exercises
b.
Given that the gamma variable has mean αθ and variance αθ2 ,
1 αθ2 θ2
E(θ̂) = θ and V (θ̂) = = .
α2 n nα
c.
θ2
The estimator θ̂ is unbiased for θ and limn→∞ V (θ̂) = limn→∞ nα = 0 and so θ̂ is
a consistent estimator of θ.
