SOLUTION FOR HOMEWORK 4, STAT 6331

Welcome to your fourth homework. Reminder: if you find a mistake/misprint, do not

e-mail or call me. Write it down on the first page of your solutions and you may give yourself
a partial credit — but keep in mind that the total for your homeworks cannot exceed 20
points.

Now let us look at your problems.

2
1. a) Here f (x|θ) = (2π)1/2 e−(x−θ) /2 I(−∞ < x < ∞) and −∞ < θ < ∞. Write

f (x|θ) Pn 2 2
L(x, y|θ) := = e−(1/2) l=1 (xl −yl )+θn(x̄−ȳ) .
f (y|θ)

We see that L(x, y|θ) ≡ K(x, y) for all θ and all pairs (x, y) iff x̄ = ȳ. Thus T = X̄ is
Minimal SS (MSS).
b) Here f (x|θ) = e−(x−θ) I(x > θ) and θ ∈ (−∞, ∞). Write,
Pn
f (x|θ) e− l=1 xl I(x(1) > θ)
L(x, y|θ) := = − Pn y . (1)
f (y|θ) e l=1 l I(y
(1) > θ)

(i) Suppose that L(x, y|θ ≡ K(x, y) for all θ ∈ (−∞, ∞). This yields that x(1) = y(1) .
(ii) Suppose that x(1) = y(1) . Then (1) implies that L(x, y|θ) does not depend on θ.
The properties (i) and (ii) establish, according to Theorem 6.2.13, that T = X(1) is the
minimal sufficient statistic.
e) Here f (x|θ) = (1/2)e−|x−θ| I(x ∈ (−∞, ∞)) for θ ∈ (−∞, ∞). Write for ordered
observations:
f (x|θ)
L(x, y|θ) :=
f (y|θ)
P P
l: x(l) ≤θ
(θ−x(l) )− l: x(l) >θ
(x(l) −θ)
e
= P P
(θ−y(l) )− l: y >θ (y(l) −θ)
.
l: y(l) ≤θ
e (l)

Now note that if at least one pair of ordered observations is such that x(l) 6= y(l) then
L(x, y|θ depends on θ. Observing this, Theorem 6.2.13 and some straightfirward steps
(similar to the previous case), show that T = (X(1) , . . . , X(n) ) is MSS.

2. Exerc. 6.11. (a) All the considered distributions (random variables) are from a
so-called location family where
D
X = Z + θ, Z ∼ fX (x|θ = 0). (2)

You can aslo think about (2) in the following way: there is a random variable Z with
distribution corresponding to θ = 0, and then any other RV from the same distribution with
parameter θ is generated as shown in (2).

1
 Let us prove (2) using moment generating function approach (another approach is to
calculate fX (x|θ) using our techniques in Chapter 2). Recall that MX (t) = E(etX ) and X
has the same distribution as Y iff MX (t) ≡ MY (t) for all t in some vicinity of t = 0, say
t ∈ (−δ, δ), δ > 0.
Thus, to check (2), we need to show that

E(etX ) = E(et(Z+θ) ), t ∈ (−δ, δ) (3)

for the case of a location family f (x|θ) =: ψ(x − θ). Here fZ (z|θ) = ψ(z) is the pdf of Z.
To prove (3) we write
Z ∞
MX (t) = E(etX ) = ψ(x − θ)etx dx
−∞

[make change of variable z = x − θ and continue]

Z ∞
= ψ(z)et(z+θ) dz
−∞

[now we continue for all t that the integral exists]

= E{et(Z+θ) } = MZ+θ (t).

What was wished to show.

Now, if Yi = X(n) −X(i) then according to (2) we have Yi = Z(n) −Z(i) , and this shows that
the distribution of Yi does not depend on θ because the distribition of Z does not depend
on θ.
We conclude that in all examples of a location family of distributions, statistics Yi are
ancillary for the location parameter θ. By the way, can you propose several other ancillary
statistics?
(b)

(i) For the case (a) - normal distribution - the MSS is X̄, and it is complete because
normal distribution belongs to an exponential family. Indeed, we can write

f (x|θ) = C(θ)h(x)eθnx̄

and then use Theorem 6.2.25. Then Basu’s Theorem yields that X̄ and Yi are independent.
(ii) For the case (b) - location exponential family - the MSS is X(1) . Is it complete? Let
us check. Suppose that for a function g

Eθ (g(X(1) ) ≡ 0, θ ∈ (−∞, ∞). (4)

This is equivalent to Z ∞
fX(1) (t|θ)g(t)dt ≡ 0, θ ∈ (−∞, ∞). (5)
θ

2
Well, now is the time to remember the density for X(1) . While I know that you remember the
formula, let us one more time deduce it. We begin with the corresponding cdf (cumulative
distribution function)

FX(1) (t|θ) = P(X(1) ≤ t) = 1 − P(X(1) > t)

[remember that the last probability is called the survivor function]

= 1 − P(X1 > t, X2 > t, . . . , Xn > t) = 1 − [P(X1 > t)]n

Z ∞
= {1 − [ e−(x−θ) dx]n }I(t > θ) = [1 − e−n(t−θ) ]I(t > θ).
t
Taking the derivative of the cdf we get the pdf

fX(1) (t|θ) = ne−n(t−θ) I(t > θ).

Now we plug-in this density in (5),

Z ∞
n e−n(t−θ) g(t)dt ≡ 0, θ ∈ (−inf ty, ∞).
θ

The last relation is equivalent to

Z ∞
q(θ) := e−nt g(t)dty ≡ 0, θ ∈ (−∞, ∞).
θ

From here we get dq(θ)/dθ ≡ 0 as well. In its turn, this yields (take the derivative)

e−nθ g(θ) ≡ 0, θ ∈ (−∞, ∞),

which finally yields g(θ) ≡ 0 for all θ ∈ (−∞, ∞). We established that X(1) is complete!
Conclusion: By Basu’s Theorem the CMSS X(1) is independent of the ancillary statistic
Y := {X(n) − X(i) , i = 1, 2, . . . , n − 1}.
REMARK: Do you see that this exponential distribution is not from the exponential
family? This is the reason for the direct proof!
(iii) For the double-exponential distribution the statistic T = (X(1) , X(2) , . . . , X(n) ) is
MSS. Clearly it is not independent of Y := {X(n) − X(i) , i = 1, 2, . . . , n − 1} because if you
know T then you know Y .

3. Exers. 6.19. Let us begin with distribution 1. Let ψ(X) be such that Ep (ψ(X)) ≡ 0
for all p ∈ (0, 1/4). This means that

pψ(0) + 3pψ(1) + (1 − 4p)ψ(2) ≡ 0,

or equivalently
p[ψ(0) + 3ψ(1) − 4ψ(2)] ≡ −ψ(2), p ∈ (0, 1/4).
This is possible for ψ(2) = 0 and ψ(0) = −3ψ(1) 6= 0. We conclude that ψ(k) is not
necessarily equal to 0, so the family of distributions of X is not complete.

3
 For the second distribution, similarly assume that Ep {ψ(X)} ≡ 0 for p ∈ (0, 1/2). This
yields
pψ(0) + p2 ψ(1) + (1 − p − p2 )ψ(2) ≡ 0, p ∈ (0, 1/2).
The last relations yields

p2 [ψ(1) − ψ(2)] + p[ψ(0) − ψ(2)] + ψ(2) ≡ 0, p ∈ (0, 1/2).

A polynom is identically equal to zero on an interval iff all its coefficients are zero. This
yields that ψ(k) ≡ 0.
Conclusion: The distribution is complete.

4. Exerc. 6.20.
(b) We have
θ
f (x|θ) = , x ∈ (0, ∞), θ > 0.
(1 + x)1+θ
Note that here Ω = (0, ∞). Write,
n Pn
f (x|θ) = θn (1 + xl )1+θ = θn e(1+θ) ln(1+xl )
Y
l=1 .
l=1

We see that the distribution is from an exponential family and Ω := (0, ∞) contains an open
set, say (1, 2). Thus, by Theorem 6.2.25 the statistic T := nl=1 ln(1 + Xl ) is CSS.
P

(c) Here
f (x|θ) = log(θ)(θ − 1)θx , x ∈ Ω := (1, ∞).
Write,
n Pn
f (x|θ) = [log(θ)/(θ − 1)]n θxl = [log(θ)/(θ − 1)]n eθ xl
Y
l=1 .
l=1

This is the distribution from an exponential family and Ω contains an open set, say (2, 5).
Thus, by Theorem 6.2.25 the statistic T = nl=1 Xl (or X̄) is CSS.
P

(e). Here we have

!
2
f (x|θ) = θx (1 − θ)2−x , x = 0, 1, . . . , θ ∈ [0, 1].
x

Write,
n
!
2
θxl (1 − θ)2−xl
Y
f (x|θ) =
l=1
xl
n
!
2 Pn
eln(θ/(1−θ)) xl
(1 − θ)2n .
Y
= l=1

l=1
xl
This is again the distribution from an exponential class with Ω containing an open interval.
This X̄ is CSS by Theorem 6.2.25.

4
 5. Exerc. 6.21. X is distributed according to

f (x|θ) = (θ/2)|x|(1 − θ1−|x| I(x ∈ {−1, 0, 1}), θ ∈ [0, 1].

(a) X is sufficient (of course - this is the observation) but not complete. Let us show this.
Suppose that Eθ ψ(X)) ≡ 0, θ ∈ [0, 1]. This is equivalent to

(θ/2)ψ(−1+ (1 − θ)ψ(0) + (θ/2)ψ(1) ≡ 0, θ ∈ [0, 1],

or
θ[(1/2)ψ(−1) − ψ(0) + (1/2)ψ(1)] + ψ(0) ≡ 0, θ ∈ [0, 1].
A polynom is zero over an interval iff all its coefficients are zero. This implies that
(
ψ(0) = 0
(1/2)[ψ(−1) + ψ(1)] − ψ(0) = 0

Solving this system yields ψ(0) = 0 and ψ(−a) = −ψ(1). As a results, ψ(k) is not necessarily
identical zero. For instance, one may choose ψ(−1) = 5, ψ(0) = 0 and ψ(1) = −5.
b),c) Let us begin with c). Write

f (x|θ) = e|x| ln(θ/2(1−θ)) (1 − θ)I(x ∈ {−1, 0, 1}).

This is the distribution from an exponential class. Then by Theorem 6.2.25, the statistic |X|
is CSS. Clearly it is minimal as well (what can you reduce here?), or if unclear use Theorem
6.2.28.

6. Exerc. 6.30. We have f (x|µ) = e−(x−µ) I(x > µ), µ ∈ R.

(a) Write Pn
f (x|µ) = e− l=1 xl +nµ I(x(1) > µ).
Here X(1) is clearly SS and minimal as well. But completeness we should verify via its
definition because this is not the distribution from an exponential class, that is, we cannot
use Theorem 6.2.25.
Let us find the distribution of X(1) ; we did it several times already so I just briefly note
that
P(X(1) ≤ t) = [1 − e−n(t−µ) ]I(t ≥ µ),
and this yields
fX(1) (t) = ne−(t−µ) I(t ≥ µ).
Then is Eµ {ψ(X(1) )} ≡ 0 then
Z ∞
e−n(x−µ) ψ(x)dx ≡ 0, µ ∈ (−∞, ∞).
µ

Now, the last identity implies that ψ(µ) ≡ 0. To prove this, note that we have µ∞ e−nx ψ(x)dx ≡
R

0 and then take the derivative of the integral with respect to µ. Thus X(1) is complete.

5
 (b) Using notation in the text, S 2 := [1/(n − 1)] nl=1 (Xl − X̄))2 .
P

We are dealing with the location parameter here, so we can again use the same trick as
before. Set X = Z + µ where Z is distributed with pdf fZ (z) = e−z I(z > 0). Then
n n
S 2 = (n − 1)−1 (Zl + µ − (Z̄ + µ))2 = (n − 1)−1 (Zl − Z̄)2 .
X X

l=1 l=1

This shows us that S 2 is ancillary with respect to µ. Then Basu’s Theorem implies that X(1)
and S 2 are independent.