Summer 2014 examination

ST102
Elementary Statistical Theory

Suitable for all candidates

Instructions to candidates

This paper contains seven questions. Answer BOTH questions from Section A, and THREE
questions from Section B. All questions carry equal numbers of marks.

If you answer more than three questions from Section B, only your BEST 3 answers will count
towards the final mark.

This examination counts for 100% of the final grade.

Full working must be shown to gain all marks for each question.

Time allowed - Reading Time: None

Writing Time: 3 hours
You are supplied with: Murdoch & Barnes Statistical Tables, 4th edition
Formulae Sheets (at the end of this paper)
You may also use: No additional materials
Calculators: Calculators are allowed in this examination

©LSE ST 2014/ST102 Page 1 of 10

SECTION A

Answer BOTH questions from Section A. Both questions carry equal numbers of marks.

1. (a) The random variable X has a Poisson distribution with P (X = 1) = P (X = 2).

i. Determine the mean and variance of X.
ii. Find P (X = 4).
(4 marks)

(b) Suppose that commercial aeroplane crashes in a certain country occur at the rate of
2.5 per year.
i. Is it reasonable to assume that such crashes are Poisson events? Briefly
explain.
ii. What is the probability that four or more crashes will occur next year?
iii. What is the probability that the next two crashes will occur within three months
of one another?
(8 marks)

(c) Let X be a continuous random variable with the following probability density function:
(
xe−x for x ≥ 0
f (x) =
0 otherwise.

i. Show that its moment generating function is MX (t) = 1/(1 − t)2 .

ii. Use MX (t) to find the expected value of X.
(8 marks)

©LSE ST 2014/ST102 Page 2 of 10

 2. (a) A random sample {X1 , X2 , . . . , Xn } is drawn from the following distribution:
2
λ2x e−λ
p(x; λ) = .
x!
i. Find the maximum likelihood estimator for λ.
ii. State the maximum likelihood estimator for θ = λ3 .
(9 marks)

(b) Explain how method of moments estimators are derived when sampling from a
population F (x; θ), where θ has p components.
(5 marks)

(c) The mean squared error (MSE) of an estimator is:

 
E (θb − θ)2 .

Prove that, if E(θb2 ) < ∞, the MSE can also be expressed as:
 2
MSE(θ)
b = Var(θ)
b + Bias(θ)
b .

(6 marks)

©LSE ST 2014/ST102 Page 3 of 10

SECTION B

Answer THREE questions from Section B. All questions carry equal numbers of marks.

If you answer more than three questions from Section B, only your BEST 3 answers will count
towards the final mark.

3. (a) A mutual fund company has six funds that invest in the UK market and four that
invest in international markets. An investor wants to invest in two UK funds and two
international funds.
i. How many different sets of funds from this company could the investor choose?
ii. Unknown to this investor, only one of the UK funds and only one of the
international funds will seriously underperform next year. If the investor selects
funds for purchase at random, what is the probability that at least one of the
chosen funds will seriously underperform next year?
(7 marks)

(b) A professor finds that he awards a final grade of A to 20% of his students. Of those
who obtain a final grade of A, 70% obtained an A on the midterm examination. Also,
10% of the students who failed to obtain a final grade of A earned an A on the
midterm examination. What is the probability that a student with an A on the midterm
examination will obtain a final grade of A?
(4 marks)

(c) Consider the following probability function, with the parameter π denoting a
probability: (
(1 − π)x π for x = 0, 1, 2, . . .
p(x) =
0 otherwise.
Show that E(X) = (1 − π)/π.
(9 marks)

©LSE ST 2014/ST102 Page 4 of 10

 4. (a) Consider two random variables X and Y . They both take the values −1, 0 and 1.
The joint probabilities for each pair of values, (x, y), are given by the following table.

X = −1 X=0 X=1
Y = −1 0.09 0.16 0.15
Y =0 0.09 0.08 0.03
Y =1 0.12 0.16 0.12

i. Calculate E(X | Y = 0) and E(X | X + Y = 1).

ii. Define U = |X| and V = Y . Calculate E(U ) and the covariance of U and V .
Are U and V correlated?
(10 marks)

(b) Suppose Xi ∼ N (0, 3), i = 1, 2, 3, 4. Assume all these random variables are
independent. Derive the value of k in each of the following:
i. P (X1 + 3X2 > 4) = k.
ii. P (X12 + X22 + X32 + X42 < k) = 0.99.
p
iii. P (X1 < k X22 + X32 ) = 0.05.
iv. P (X12 + X22 > 39(X32 + X42 )) = k.
(10 marks)

©LSE ST 2014/ST102 Page 5 of 10

 5. (a) Let X1 , X2 , . . . , Xn be independent and identically distributed random variables
with a N (µ, σ 2 ) distribution. Show that, given X̄ and S 2 are independent:

(n − 1)S 2
∼ χ2n−1 .
σ2
You may use the fact that X̄ ∼ N (µ, σ 2 /n) without derivation.
(10 marks)

(b) Five independent samples, each of size n, are to be drawn from a normal distribution
where σ 2 is known. For each sample, the interval:
 
σ σ
x̄ − 0.96 × √ , x̄ + 1.06 × √
n n

will be constructed. What is the probability that at least four of the intervals will
contain the unknown µ?
(5 marks)

(c) A country is considering joining the European Union. In a study of voters’ views on a
forthcoming referendum, 163 out of a random sample of 250 voters agreed with the
statement: ‘The government should seek membership of the European Union.’ Find
an approximate 99% confidence interval for the population proportion of all voters
agreeing with this statement.
(5 marks)

©LSE ST 2014/ST102 Page 6 of 10

 6. (a) A random sample of size n = 20 is taken from N (µ, σ 2 ). Consider the following
hypothesis test:
H0 : σ 2 = 2 vs. H1 : σ 2 > 2
to be conducted at the 10% significance level. Determine the power of the test
for σ 2 = 2, σ 2 = 3 and σ 2 = 4. (You may use the closest available values in the
statistical tables provided.)
(10 marks)

(b) Define the term ‘p-value’ and briefly explain how it is used in statistical hypothesis
testing when testing the value of the parameter θ.
(3 marks)

(c) Class A students were taught using detailed slides and class B students were
required to read textbooks and answer questions in class discussions. The marks in
the final examination for a sample of 13 class A students had a standard deviation of
14.33 marks. The marks in the final examination for a sample of 11 class B students
had a standard deviation of 15.35 marks.
Assuming examination marks are normally distributed, can we infer that the
variances of the marks differ between the two classes? Use a 5% significance level.
(7 marks)

©LSE ST 2014/ST102 Page 7 of 10

 7. (a) The detergent market was dominated by two big brands – Brand A occupied 45%
of total sales, Brand B took 40% of total sales. To boost their sales further, both
companies recently conducted aggressive advertising campaigns. An after-campaign
poll from 300 randomly selected customers yields the counts of customer
preferences below:
Brand A B Others
No. of customers 143 122 35

Can we infer at the 5% significance level that customer preferences have changed
after the advertising campaigns?
(6 marks)

(b) For the simple linear regression model:

yi = β0 + β1 xi + εi

derive the least squares estimators βb0 and βb1 .

(10 marks)

(c) A simple linear regression model where εi ∼ N (0, σ 2 ), is estimated to be:

ybi = 2.1 − 3.6xi

n
(xi − x̄)2 = 0.5. The
P
based on a random sample of 25 observations, for which
i=1
point estimate for σ 2 is 0.79. Test whether the slope coefficient is equal to zero at
the 1% significance level.
(4 marks)

©LSE ST 2014/ST102 Page 8 of 10

Appendix: Formula sheet

Simple linear regression

Model: yi = β0 + β1 xi + εi .
n n
(xj − x̄)2 and:
P P
LSEs: βb0 = ȳ − βb1 x̄, βb1 = (xi − x̄)(yi − ȳ)/
i=1 j=1

n
x2j
P
σ2 j=1 σ2 −σ 2 x̄
Var(βb0 ) = n , Var(βb1 ) = n , Cov(βb0 , βb1 ) = n .
n P
(xi − x̄)2 (xi − x̄)2 (xi − x̄)2
P P
i=1 i=1 i=1

n
b2 = (yi − βb0 − βb1 xi )2 /(n − 2).
P
Estimator for the variance of εi : σ
i=1

Regression ANOVA:
n n n
(yi − ȳ)2 , Regression SS = βb12 (xi − x̄)2 , Residual SS = (yi − βb0 − βb1 xi )2 .
P P P
Total SS =
i=1 i=1 i=1

Squared regression correlation coefficients:

Regression SS (Residual SS)/(n − 2)

R2 = and 2
Radj =1− .
Total SS (Total SS)/(n − 1)

For a given x, the expectation of y is µ(x) = β0 + β1 x. A 100(1 − α)% confidence interval

for µ(x) is:
 1/2
n
(xi − x)2

 P 


 

i=1
βb0 + βb1 x ± tα/2, n−2 × σ
b× n
(xj − x̄)2 
P
n

 
 

j=1

and a 100(1 − α)% prediction interval covering y with probability (1 − α) is:

 1/2
n
2
P
(xi − x) 

 

 

i=1
βb0 + βb1 x ± tα/2, n−2 ×σ
b × 1+ n .
2
P
n (x − x̄)

 j
 

 
j=1

©LSE ST 2014/ST102 Page 9 of 10

One-way ANOVA:
nj
k P nj
k P
(Xij − X̄)2 = 2 − nX̄ 2 .
P P
Total variation: Xij
j=1 i=1 j=1 i=1

k k
nj (X̄·j − X̄)2 = 2 − nX̄ 2 .
P P
Between-treatments variation: B = nj X̄·j
j=1 j=1

nj
k P nj
k P k
(Xij − X̄·j )2 = 2 − 2.
P P P
Within-treatments variation: W = Xij nj X̄·j
j=1 i=1 j=1 i=1 j=1

Two-way ANOVA:
r P
c r P
c
(Xij − X̄)2 = 2 − rcX̄ 2 .
P P
Total variation: Xij
i=1 j=1 i=1 j=1

r r
(X̄i· − X̄)2 = c X̄i·2 − rcX̄ 2 .
P P
Between-blocks (rows) variation: Brow = c
i=1 i=1

c c
(X̄·j − X̄)2 = r 2 − rcX̄ 2 .
P P
Between-treatments (columns) variation: Bcol = r X̄·j
j=1 j=1

Residual (error) variation:

r X
X c r X
X c r
X c
X
(Xij − X̄i· − X̄·j + X̄)2 = 2
Xij −c X̄i·2 − r 2
X̄·j + rcX̄ 2 .
i=1 j=1 i=1 j=1 i=1 j=1

[END OF PAPER]

©LSE ST 2014/ST102 Page 10 of 10