Summer 2012 examination

ST107
Quantitative Methods (Statistics)

Suitable for all candidates

Instructions to candidates

This paper contains eight questions.

Section A is worth 40% and Section B is worth 60%.

Answer all three questions in Section A. All questions in Section B will be given equal weight
(20% each). Section B contains five questions, of which you may attempt as many questions
as you wish, but only your best three questions will count towards your final mark.

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

Time allowed - Reading Time: None

Writing Time: 2 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 LT 2012/ST107 Page 1 of 12

SECTION A

Answer all questions in this section (40 marks in total).

1. Give brief answers to each of the following questions.

(a) Maria has an important job interview in the morning. To ensure she wakes up in time,
she sets two alarm clocks which ring with probabilities 0.97 and 0.99 respectively.
What is the probability that at least one of the alarm clocks will wake her up?

(b) Suppose that the distribution of a set of values has a population mean of 50 and a
population variance of 6. If 4 is subtracted from each score and then each score is
divided by 2, what will be the new mean and variance of the set of values?

(c) If X ∼ Bin(n, π) state, in terms of n and π, what the approximating Poisson and
normal distributions would be.

(d) State the central limit theorem.

(e) When and why can a standard normal distribution be used to approximate a
Student’s t distribution?

(f) In a hypothesis test for a single proportion using a normal distribution, we have
H0 : π = 0.7 and H1 : π < 0.7. The test statistic is computed to be z = −1.676.
For α = 0.10, α = 0.05 and α = 0.01, state whether or not the null hypothesis is
rejected.

(g) Define what a p-value is.

(h) The following least squares regression line was computed:

yb = 51.7 + 3.47x

where x is the number of hours studied and y is examination score. Based on this
predicted relationship, approximately how many hours should a student study to get
at least 85 on the examination?

(16 marks – two marks for each part)

©LSE ST 2012/ST107 Page 2 of 12

 2. A continuous random variable X has the cumulative distribution function:

0
 for x < 0
2
F (x) = 2x − x for 0 ≤ x ≤ 1

1 for x > 1.


Find:

(a) the probability density function

(2 marks)
(b) the mean of X
(2 marks)
(c) the variance of X
(2 marks)
(d) the median of X to four decimal places
(2 marks)
(e) the mode of the distribution
(1 mark)
(f) P (X > 0.75 | X > 0.25).
(3 marks)

3. A well-known fast food retailer forecasts its percentage sales growth for its next financial
year for each of 32 sales regions. The forecasts are shown in the following table:

6.5 7.2 5.6 3.7 7.5 6.1 4.9 4.7

3.3 3.8 5.2 6.3 3.3 2.5 6.8 4.9
6.3 5.5 4.5 3.9 4.5 3.8 5.1 6.4
8.2 4.5 4.7 5.0 4.1 5.1 3.7 5.3

(a) Construct and label carefully a stem-and-leaf diagram for this dataset. Use your
diagram to find the median and lower and upper quartiles.
(8 marks)

(b) From the shape of the stem-and-leaf diagram, state whether the distribution of the
sample is skewed and, if so, give the direction of the skewness.
(2 marks)

(c) Without making any further calculations, state whether you would expect the sample
mean to be the same as, smaller than, or larger than the sample median and give a
brief reason for this.
(2 marks)

©LSE ST 2012/ST107 Page 3 of 12

SECTION B

You may attempt as many questions as you wish from this section, but only your best three
questions will count towards your final mark. Each question carries 20 marks.

4. (a) 20% of men show early signs of losing their hair. 2% of men carry a gene that is
related to hair loss. 80% of men who carry the gene experience early hair loss.
i. What is the probability that a man carries the gene and experiences early hair
loss?
ii. What is the probability that a man carries the gene, given that he experiences
early hair loss?
(5 marks)

(b) Given two events A and B, state whether each of the following is true or false.
Use formulae or words to justify your answer. (No marks will be awarded without a
justification.)
i. If A and B are independent such that P (A) > 0 and P (B) > 0, then:

P (A ∪ B) < P (A) + P (B).

ii. If A and B are independent, Ac and B are independent.

iii. If P (B) > 0, P (A) < 1 and B ⊂ A, Ac and B are independent.
iv. If A and B are mutually exclusive, then:

P (Ac ∩ B c ) = 1 − P (A) − P (B).

v. If P (A | B) > P (B | A), then P (A) > P (B).

(10 marks)

(c) Using the axioms of probability, show that P (∅) = 0.

(5 marks)

©LSE ST 2012/ST107 Page 4 of 12

 5. (a) The random variable X has a binomial distribution with n = 4 and π = 0.25.
Determine the probability distribution of Y = (X − 1)2 . Calculate E(Y ) and Var(Y ).
(6 marks)

(b) The local fair has a game for children. Children pay $0.25 to draw a marble at random
from a covered jar which contains 100 marbles, of which 4 are silver, 1 is gold, and
the remaining marbles are white. If a silver marble is drawn, the child wins $1. If a
gold marble is drawn, the child wins $10. If a white marble is drawn, the child wins
nothing. What is the expected amount of winnings for this game, per play?
(4 marks)

(c) If X ∼ N (3, 2), Y ∼ N (2, 3) and X and Y are independent, calculate P (X < Y ).
(4 marks)

(d) In the general population 38% have type O+ blood. Use a suitable normal
approximation to find the probability of finding more than 100 people with type O+
blood in a random sample of 300 people.
(6 marks)

©LSE ST 2012/ST107 Page 5 of 12

 6. (a) Let T1 be an unbiased estimator of the parameter θ, and T2 be an unbiased
estimator of the parameter φ. Is T1 T2 an unbiased estimator of θφ?
(4 marks)

(b) A random variable X can take the values 0, 1 and 2. We know that:
3α α α
P (X = 0) = 1 − , P (X = 1) = and P (X = 2) = .
4 2 4
One observation is taken and we want to estimate α, such that 0 < α < 4/3.
Consider the estimators T1 = X and T2 = 2X(X − 1).
i. Show that they are both unbiased.
ii. Would you prefer estimator T1 or T2 ? Justify your choice.
(10 marks)

(c) Briefly explain the difference between an estimator and a point estimate.
(3 marks)

2 ) and N (µ , σ 2 ),
(d) Independent samples of size m and n are drawn from N (µX , σX Y Y
respectively. Give the sampling distribution of X̄ − Ȳ .
(3 marks)

©LSE ST 2012/ST107 Page 6 of 12

 7. (a) A pet food supplier is studying the difference between two of its stores. It is
particularly interested in the time it takes before customers receive the products
they have ordered. Using standard notation, the data of delivery times from the two
stores are as follows:

Store A Store B
x̄ 34.3 days 38.6 days
s 2.5 days 3.4 days
n 41 31

i. Use an appropriate hypothesis test to see if there is a difference in the average

delivery times for the two stores. Test at two appropriate levels and comment on
your findings.
ii. Give the assumptions you have made.
iii. Give the 98% confidence interval for the mean delivery time for Store B.
(13 marks)

(b) A UK government agency carries out a large-scale random survey of public attitudes
towards the recession. 83 of the 500 workers surveyed indicated they were worried
about losing their job. Newspaper reports claim 20% of workers fear losing their
job. Is such a high percentage claim justified? State and carry out an appropriate
hypothesis test at two levels and explain your results.
(7 marks)

©LSE ST 2012/ST107 Page 7 of 12

 8. (a) The ministry of education is considering funding pre-school education. Before
making their recommendations, administrators take a random sample of 100
students from various areas to compare the performance of students in algebra
between those who attended pre-school and those who did not. The results are
summarised in the table below:

Below Grade Level At Grade Level Advanced

Pre-school 12 29 16
No Pre-school 18 11 14

i. Test for association between performance in algebra and pre-school attendance

at two appropriate significance levels. State the null and alternative hypotheses
clearly.
ii. Comment on your results describing potential associations in detail. Also,
discuss any potential differences in the algebra marks between students who
did and students who did not attend pre-school.
(11 marks)

(b) It is assumed that there is a linear relationship between the yield obtained from apple
trees and the amount of fertiliser supplied to them. In order to test this assumption,
nine apple trees of the same type were randomly selected and supplied weekly with
a fixed quantity (x grams) of fertiliser. The yield of each apple tree (y kilograms) was
recorded.

Tree 1 2 3 4 5 6 7 8 9
x 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
y 3.9 4.3 5.5 6.4 6.9 7.1 7.3 7.7 8.0

The summary statistics for these data are:

Sum of x data: 27 Sum of the squares of x data 96

Sum of y data: 57.1 Sum of the squares of y data: 379.51
Sum of the products of x and y data: 186.75

i. Calculate the least squares line of y on x.

ii. What prediction will you give for a tree that is treated weekly with 3.2 grams of
fertiliser?
iii. Would you use the equation to predict the yield of an apple tree that is treated
weekly with 15 grams of fertiliser? Justify your answer.
(9 marks)

©LSE ST 2012/ST107 Page 8 of 12

 ST107 Formula Sheet

1. Sample statistics

• Corrected sum of squares:

n n n n
!2
X X X 1 X
Sxx = (xi − x̄)2 = x2i − nx̄2 = x2i − xi .
i=1 i=1 i=1
n i=1

• Corrected sum of cross-products:

n n n n
X X 1X X
Sxy = xi yi − nx̄ȳ = xi yi − xi yi .
i=1 i=1
n i=1 i=1

• Sample variance:
n
2 1 X Sxx
s = (xi − x̄)2 = .
n−1 i=1
n−1

2. Probability theory

• Additive law: P (A ∪ B) = P (A) + P (B) − P (A ∩ B).

• Conditional law (Bayes’ formula):

P (A ∩ B)
P (A | B) =
P (B)
or:
P (A | B) P (B) = P (A ∩ B) = P (B | A) P (A).

• Total probability formula:

P (A) = P (A | B) P (B) + P (A | B c ) P (B c )

or:

P (A) = P (A | B1 ) P (B1 ) + P (A | B2 ) P (B2 ) + · · · + P (A | Bn ) P (Bn ).

©LSE ST 2012/ST107 Page 9 of 12

 3. Discrete probability distributions

Distribution Probability function, P (X = x) Mean Variance

 
n x
Binomial π (1 − π)n−x for x = 0, 1, 2, . . . , n nπ nπ(1 − π)
x

Poisson e−λ λx /x! for x = 0, 1, 2, . . . λ λ

Geometric (1 − π)x−1 π for x = 1, 2, . . . 1/π (1 − π)/π 2

x−1
 
Negative binomial π r (1 − π)x−r for x = r, r + 1, r + 2, . . . r/π r(1 − π)/π 2
r−1

4. Continuous probability distributions

Distribution Probability density function, f (x) Mean Variance

Exponential λe−λx for x ≥ 0 1/λ 1/λ2

(x − µ)2
 
1
Normal √ exp − for −∞ < x < ∞ µ σ2
2πσ 2 2σ 2

©LSE ST 2012/ST107 Page 10 of 12

 5. Statistical inference – point estimates and (estimated) standard errors

Single parameters Point estimate (Estimated) standard error

n
P √
µ (σ known) x̄ = xi /n σ/ n
i=1

n
P √
µ (σ unknown) x̄ = xi /n s/ n
i=1
p
π p = r/n C.I.: p(1 − p)/n
p
H.T.: π0 (1 − π0 )/n

Differences between Point estimate (Estimated) standard error

parameters
p
µ1 − µ2 x̄1 − x̄2 σ12 /n1 + σ22 /n2
(known variances)
p
µ1 − µ2 x̄1 − x̄2 s21 /n1 + s22 /n2
(σ12 , σ22 unknown & unequal)
q
µ1 − µ2 x̄1 − x̄2 s2p (1/n1 + 1/n2 )
(σ12 , σ22 unknown but equal)

(n1 −1)s2 2
1 +(n2 −1)s2
σ12 = σ22 s2p = n1 +n2 −2
—
p
π1 − π2 p1 − p2 = r1 /n1 − r2 /n2 C.I.: p1 (1 − p1 )/n1 + p2 (1 − p2 )/n2
p
H.T.: p(1 − p)(1/n1 + 1/n2 )

p (common (pooled) p = (r1 + r2 )/(n1 + n2 ) —

proportion)

©LSE ST 2012/ST107 Page 11 of 12

 6. Correlation and linear regression

Parameter Sample value Population counterpart

Covariance c = Sxy /(n − 1) γ = E(XY ) − E(X) E(Y )

p
Correlation r = Sxy / Sxx Syy ρ = γ/(σX σY )

1
Residual variance s2 = 2
(Syy − Sxy /Sxx ) σ2
n−2

1
or s2 = (Syy − βb12 Sxx )
n−2

√ √
Under H0 : ρ = 0, the test statistic r n − 2/ 1 − r 2 ∼ tn−2 .

Parameter Point estimate Standard error

√
β1 (regression slope) βb1 = Sxy /Sxx s/ Sxx
p
β0 (regression intercept) βb0 = ȳ − βb1 x̄ s 1/n + x̄2 /Sxx
p
yb(x0 ) (y prediction at x0 ) βb0 + βb1 x0 s 1 + 1/n + (x0 − x̄)2 /Sxx

Sampling distribution of Sxx (corrected sum of squares) for a single sample:

Sxx /σ 2 ∼ χ2n−1 E(Sxx ) = (n − 1)σ 2 Var(Sxx ) = 2(n − 1)σ 4 .

[END OF PAPER]

©LSE ST 2012/ST107 Page 12 of 12