NOT TO BE TAKEN AWAY

THE UNIVERSITY OF HONG KONG

DEPARTMENT OF STATISTICS AND ACTUARIAL SCIENCE

STAT2602 PROBABILITY AND STATISTICS II

?? ??, 2018 Time: ??:?? ?.m. - ??:?? ?.m.

Only approved calculators as announced by the Examinations Secretary can be used in

this examination. It is candidates’ responsibility to ensure that their calculator operates
satisfactorily, and candidates must record the name and type of the calculator used on
the front page of the examination script.

Answer ALL questions. Marks are shown in square brackets.

S&AS: STAT2602 Probability and Statistics II

1
1. Recall that if X ∼ χ2ν , MX (t) = 1
ν for t < 2 , where MX (t) is the moment
(1 − 2t) 2
generating function of X. Let X1 , X2 , X3 be an independent random sample such
that Xi ∼ χ2i for i = 1, 2, 3.

∑
3
(i) Calculate the moment generating function of Y = Xi . [5 marks]
i=1
(ii) Specify the distribution of Y . [2 marks]
(iii) Calculate E(Y s ) for s = 1, 2. [8 marks]
[Total: 15 marks]

2. Suppose that a person takes the subway to school everyday, and the waiting time
(in minute) for the subway follows a uniform distribution U [0, 5]. What is the
approximated probability that the total waiting time in 100 days is larger than 5
hours?
[Total: 15 marks]

3. Let X1 , X2 , · · · , Xn be an independent random sample from N (0, θ), where θ > 0.

(i) Find θ̂1 , the MLE of θ on the space Ω1 = {θ : θ > 0}. [5 marks]
(ii) Show that θ̂1 is an unbiased estimator of θ. [2 marks]
(iii) Show that θ̂1 is a consistent estimator of θ. [3 marks]
(iv) Recall that EXi4 2
= 3θ . Will θ̂1 be the UMVUE of θ? Explain it. [5 marks]
(v) Find θ̂2 , the MLE of θ on the space

Ω2 = {θ : θ > θ∗ },

where θ∗ is a given finite positive constant. [10 marks]

(vi) Show that θ̂2 is a biased estimator of θ. [5 marks]
[Total: 30 marks]

4. Let X1 , X2 , · · · , Xn be a random sample with Xi ∼ N (µi , θi ), where θi > 0 for all

i.

(i) If X1 , X2 , · · · , Xn are independent, provide a general condition so that X is a

consistent estimator of µ, where

1∑ 1∑
n n
X= Xi and µ = µi .
n i=1 n i=1

[5 marks]

P. 2 of 4
S&AS: STAT2602 Probability and Statistics II

( )i
1 1
(ii) If X1 , X2 , · · · , Xn are independent, µi = , and θi = , will X be a con-
i 2
sistent estimator of µ ? Explain it. [5 marks]
(iii) If X1 , X2 , · · · , Xn satisfy that
{
0, if |i − j| > 10,
Cov(Xi , Xj ) = 1
2
, if |i − j| ≤ 10,

will X be a consistent estimator of µ ? Explain it. [5 marks]

[Total: 15 marks]

2
5. Let X1 , X2 , X3 , X4 be a random sample of STAT2601 scores, assumed to be N (µX , σX ),
and let Y1 , Y2 , Y3 , Y4 , Y5 be a random sample of STAT2602 scores, assumed to be
N (µY , σY2 ). Suppose X1 , X2 , X3 , X4 , Y1 , Y2 , Y3 , Y4 , Y5 are independent, and the fol-
lowing data are observed:
x1 = 75 x2 = 95 x3 = 80 x4 = 65
y1 = 80 y2 = 75 y3 = 90 y4 = 70 y5 = 70
2
(i) Find a 90% confidence interval estimate for σX . [3 marks]
2
(ii) Find a 90% confidence interval estimate for σX /σY2 . [5 marks]
(X1 − X2 ) /σX
2 2
(iii) Find the distribution of . [2 marks]
(Y1 − Y2 )2 /σY2
(iv) Based on the result in (iii), construct an alternative 90% confidence interval
2
estimate for σX /σY2 . [3 marks]
(v) Is the confidence interval estimate in (ii) better than the one in (iv)? Explain
it. [2 marks]
[Total: 15 marks]

6. Let X1 , X2 , · · · , Xn be an independent random sample from N (µ, 1). Suppose that

a test for hypotheses:
H0 : µ = 0 versus H1 : 1 ≤ µ ≤ 2
{ n }
∑
has the rejection region Xi ≥ c for some constant c.
i=1

(i) Determine the value of c such that this test has size 0.05. [5 marks]
(ii) Based on the value of c in (i), find the maximum and minimum power of this
test under H1 . [5 marks]
[Total: 10 marks]

P. 3 of 4
S&AS: STAT2602 Probability and Statistics II

A LIST OF STATISTICAL FORMULAE

( )
tX r dr
1. MX (t) = E(e ). E(X ) = MX (t) .
dtr t=0

1∑ 1 ∑( )2
n n
2. X = Xi . S = 2
Xi − X .
n i=1 n i=1

X −µ
3. CLT: √ ≈ N (0, 1) for large n.
σ/ n

nS 2 X −µ
4. Normal population =⇒ ∼ χ2n−1 , √ ∼ tn−1 .
σ 2
S/ n − 1
[ ] [ ]2
5. Bias(θ̂) = E(θ̂) − θ. E (θ̂ − θ)2 = Var(θ̂) + Bias(θ̂) .
[( )2 ] [ ]
∂ ln f (X; θ) ∂ 2 ln f (X; θ) 1
6. I(θ) = E =E − . Var(θ̂) ≥ .
∂θ ∂θ2 nI(θ)
( )
σ s ns2 ns2
7. x ± zα/2 √ . x ± tα/2,n−1 √ . , .
n n−1 χ2α/2,n−1 χ21−α/2,n−1
√ √ ( )
σy2
σx2 nx s2x + ny s2y 1 1
8. x − y ± zα/2 + . x − y ± tα/2,nx +ny −2 + .
n1 n2 nx + ny − 2 nx ny
( )
nx (ny − 1)s2x 1 nx (ny − 1)s2x
· , · Fα/2,ny −1,nx −1 .
ny (nx − 1)s2y Fα/2,nx −1,ny −1 ny (nx − 1)s2y
1
9. F1−α,m,n = .
Fα,n,m
( )
∑ (Oi − Ei )2 ∑ O2 ∑ ∑ Oi,j 2
10. −2 ln Λ ≈ = i
− n. −2 ln Λ ≈ n −1 .
i
Ei i
Ei i j
ni. nj.

********** END OF PAPER **********

P. 4 of 4