INSTRUCTIONS

There are three questions.

Answer all questions.
Each question has a total of 15 marks available.
Please read through the instructions on the online platform (e.g. Canvas) before your exam.
You have 1 hour 30 minutes to complete this examination and submit online. This includes time
to scan/take pictures and upload.
You are responsible for finding a suitable venue that is quiet and conducive to assessment.
You need to ensure that you have access to a reliable internet connection and an electronic device
that can open a web browser (tablet or computer).
All answers submitted need to be your own, but it is an open book exam so you can consult other
material.
Whilst you will have access to the internet and other sources of information, please remember that
questions assess your ability to integrate and synthesise information. A quick Google search will
not be helpful.
We will be analysing all your answers after the exam as part of our quality assurance process.
If there is evidence that you have conferred with anybody during the remote exam, this will be
treated as an exam irregularity, and where relevant a professionalism issue.
Once you have completed the exam please submit online.
1. (i) The probability of an event B given A is given by the Bayes’ Theorem:

P (A|B) P (B)
P (B|A) =
P (A)
If you know that P (B|Ā) = 0.3, P (A) = 0.1, and P (B) = 0.35. Use these values work
out the probability of A given B. [3 MARKS]
(ii) In a randomised control trial, breast cancer patients are assigned to group receiving a
new treatment with probability 0.35 and to the old treatment group with probability
0.65.
(a) The old treatment, based on a previous trial, has a 74% 5 year disease free survival.
It has been determined that the total probability of 5 year survival across both arms
of the trial 82%. [4 MARKS]
(b) What is the probability that a patient received the new treatment given that they
have been disease free for 5 year. [3 MARKS]
(iii) Patients in this trail are recruited from multiple cities. In one city 15 patients are
recruited.
(a) What is the probability that at least 3 of the patients in the city won’t have a
five-year survival. [3 MARKS]
(b) In the city there are only three specialised hospital beds to deal with patient with
recurrence. What is the probability that there will be more hospital beds required?
[2 MARKS]
2. (i) (a) A survey is conducted to look at the BMI of patients in a trial for a disease with
50 participants. Explain why any statistic computed from the participants has a
sampling distribution? [3 MARKS]
(b) After the survey a estimate for the population mean is calculated and a confidence
interval is derived. What is a confidence interval, and why is it a random quantity?
[3 MARKS]
(ii) (a) Using this survey the sample mean for the BMI is calculated to be 22.4 with a
variance of 4.3. Construct a 90% confidence interval for the population mean BMI.
State any assumptions you make. [4 MARKS]
(b) In a subsequent study the BMI is measure again. In this study the sample size is
45 and the sample mean is calculated as 22.52 and the variance is 5.21.
Conduct a hypothesis test at 95% significance to determine if the BMI is different
in the new study. Show clear working and state any assumptions you have made.
[5 MARKS]

The following information may be used in answering this question.

For normal populations with known variance σ 2 , the confidence interval for the population
mean can be computed using the following expression:
√ √
(x − ∆σ/ n, x + ∆σ/ n)

where x is the sample mean, n is the sample size and ∆ is a suitably chosen interval width
for the desired confidence level.
If Z is a Normally distributed random variable with zero mean and unit variance then for a
one sided test we have:

z 1.28 1.65 1.96 2.33

P (Z < z) 0.90 0.95 0.975 0.99

For two samples with sample sizes n1 and n2 with observed means x1 and x2 , the Z-statistic
for a two-sample comparison of means test is given by:

(x1 − x2 ) − (µ1 − µ2 )
Z= p
σ12 /n1 + σ22 /n2

where µ1 , µ2 are the population means and σ12 , σ22 are the population variances.
3. (i) (a) A random variable X has an expected value 2 and variance 0.25. What is the
expected value and variance of 4 X − 10? [2 MARKS]
(b) A second random variable Y has an expected value of 1.5 and variance of 0.5. What
is the expected value and variance of 2.5 X − 0.1 Y ? [3 MARKS]
(ii) What is meant by a two-sample location test if we want to measure the difference between
populations A and B. [3 MARKS]
(iii) In a preclinical trial for a new drug, toxicity was measured on 90 cell batches. After
the drug is administered the number of dead cells are counted per batch. These are the
observations:
Dead Cells 0 1 2 3 4 5+
Frequency 6 26 23 19 12 4
Determine if the observed frequencies of dead cells per batch are consistent with a Poisson
distribution with rate parameter 1.9 at a 95% significance.
State your assumptions and working throughout. [7 MARKS]

The following information may be used in answering this question.

The Poisson distribution has a probability mass function given by:

λx exp −λ
P (X = x) =
x!
for x ≥ 0.
The chi-square statistic is given by:

K
2
X (Ok − Ek )2
χ = ,
Ek
k=1

where Ok and Ek are the observed and expected counts for the k-th category (out of K
categories).
If χ2 is a random variable with a chi-squared distribution with ν degrees of freedom then the
probability P (χ2 > k):

k .05 .025 .01 .005

ν=2 5.99 7.38 9.21 10.60
ν=3 7.81 9.35 11.34 12.84
ν=4 9.49 11.14 13.28 14.86

END OF PAPER