Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
Final 2021SM1 with solution
Quiz Instructions
18 ptsQuestion 1
A newly opened driving school is running a promotion to advertise their lessons to
new drivers with learner permits. For each of the next 120 new drivers who receive
their learner permit, a representative of the driving school invites them to flip a coin
:
"heads" means the new driver can take driving lessons for free,
"tails" means the new driver will not receive any driving lessons.
Let N be a random variable denoting the number of new drivers who take the
driving lessons.
1. (3 marks) What is the probability distribution of N?
2. (2 marks) What is the expected value of N?
The probability that a new driver will progress to passing their driving test without
driving lessons is 0.8, while those who have taken the lessons would pass their
driving test with probability 0.9.
Let P be a random variable denoting the number out of all of the new drivers with
learner permits who pass their driving test.
3. (6 marks) Derive an expression for E( P | N ).
(Hint: it may help to write P = P + P , where P is the number who passed the
driving test out of those who flipped "heads" (and hence received the lessons),
and P is the number who passed the driving test out of those who flipped
"tails" (and missed out on the lessons).
4. (3 marks) Hence calculate E( P ).
5. (4 marks) After the promotion is completed you meet a new driver who has
passed their driving test. What is the probability that they took the driving
lessons?
H T H
T
Edit View Insert Format Tools Table
12pt Paragraph
09/06/2022, 00:21 Quiz: Final 2021SM1 with solution
https://canvas.lms.unimelb.edu.au/courses/124333/quizzes/157982/take 2/10
p 0 words
Ginger is now ac
Start writing! Your mistakes wi
10 ptsQuestion 2
There is growing international interest is comparing hospitalisation
rates (proportions of patients entering hospital) for older male and female patients
with Covid-19. In particular there is a hypothesis that the male hospitalisation rate
is higher than the female hospitalisation rate.
In order to carry out a study on this, independent simple random samples were
obtained of 80 male and 80 female patients over the age of 70 who tested positive
to Covid-19. Each of these patients was monitored until they either recovered at
home or were hospitalised, and the numbers of each for males and females
recorded.
Define the notation
p : the population hospitalisation rate for older male patients with Covid-19
p : the population hospitalisation rate for older female patients with Covid-19
(i) To test the theory, the null hypothesis would be
a. H : p = p
b. H : p > p
c. H : p < p
d. H : p ≠ p
(ii) To test the theory, the alternative hypothesis would be
a. H : p = p
b. H : p > p
c. H : p < p
d. H : p ≠ p
M
F
0 M F
0 M F
0 M F
0 M F
A M F
A M F
A M F
A M F
09/06/2022, 00:21 Quiz: Final 2021SM1 with solution
https://canvas.lms.unimelb.edu.au/courses/124333/quizzes/157982/take 3/10
(iii) Suppose it is genuinely true that hospitalisation rates are higher for males than
for females. If we carried out a hypothesis test and did not reject the null
hypothesis at the 5% level of significance, this would be:
a. a correct decision
b. a Type I error
c. a Type II error
(iv) Suppose it is genuinely true that hospitalisation rates are equal for males and
females. If we carried out a hypothesis test and did not reject the null hypothesis at
the 5% level of significance, this would be:
a. a correct decision
b. a Type I error
c. a Type II error
(v) An independent statistician evaluating the study plan is concerned the sample
size is not large enough to reliably detect the hypothesised difference in
hospitalisation rates if it is genuinely present. A larger sample would address this
concern because it would
a. improve the significance level of the hypothesis test
b. improve the power of the hypothesis test
c. reduce the probability of a Type I error
d. imply that the Central Limit Theorem is not necessary
e. none of these
(vi) The sample design for the study can be best described as
a. matched pairs of male and female patients
b. a single simple random sample of 160 patients
c. independent samples of male and female patients
d. all of these
(vii) When the study was carried out, there were 14 recorded hospitalisations
amongst the male patients and 10 amongst the females. The sample
hospitalisation rates are therefore
a. p = 0.175, p = 0.125
b. p = 0.125, p = 0.175
c. p̂ = 0.175, p̂ = 0.125
d. p̂ = 0.125, p̂ = 0.175
(viii) The standard error of the difference between the male and female sample
hospitalisation rates is closest to
a. 0.042
b. 0.056
c. 0.037
d. 0.888
M F
M F
M F
M F
09/06/2022, 00:21 Quiz: Final 2021SM1 with solution
https://canvas.lms.unimelb.edu.au/courses/124333/quizzes/157982/take 4/10
p 0 words
(ix) The p value for the hypothesis test is closest to
a. 0.376
b. 0.05
c. 0.188
d. 0.812
(x) At the 5% level of significance, the conclusion of this test is
1. Reject H and conclude that there is evidence that the male hospitalisation rate
is higher than the female rate
2. Reject H and conclude that there is no evidence that the male hospitalisation
rate is higher than the female rate
3. Do not reject H and conclude that there is evidence that the male
hospitalisation rate is higher than the female rate
4. Do not reject H and conclude that there is no evidence that the male
hospitalisation rate is higher than the female rate
0
0
0
0
Looks like you got si
In order to continue receiving th
Ginger, please sign b
SIGN BACK IN
Don't show this ag
12 ptsQuestion 3
The 24/7 chain of small convenience stores has been having problems with
shoplifting, so they installed some CCTV cameras in visible places in each of their
Edit View Insert Format Tools Table
12pt Paragraph
09/06/2022, 00:21 Quiz: Final 2021SM1 with solution
https://canvas.lms.unimelb.edu.au/courses/124333/quizzes/157982/take 5/10
p 0 words
stores to try to deter offenders. In order to assess the effect of these cameras, they
chose a random sample of 35 of their stores and gathered some data on the
value of thefts over a four week period prior to the installation of the cameras. For
these stores the mean of the losses to theft was $1,336 with standard deviation of
$437. Data were then collected for the value of thefts in the same stores over the
four week period following the installation of the cameras. The mean losses were
$1,299 with standard deviation of $312. The standard deviation of the changes in
losses for each store was $266.
1. Carry out a hypothesis test to evaluate whether the presence of the CCTV
cameras made any difference to the mean losses because of theft.
Use the 5% level of significance and a p value decision rule.
Include
(a) statement of hypotheses (including definition of any notation)
(b) brief justification of testing approach chosen
(c) test statistics and decision rule
(d) decision of the test and brief interpretation
2. Compute a 90% confidence interval for the change in the mean losses to theft
as a result of the CCTV installation.
Looks like you got si
In order to continue receiving th
Ginger, please sign b
SIGN BACK IN
Don't show this ag
12 ptsQuestion 4
Edit View Insert Format Tools Table
12pt Paragraph
09/06/2022, 00:21 Quiz: Final 2021SM1 with solution
https://canvas.lms.unimelb.edu.au/courses/124333/quizzes/157982/take 6/10
To evaluate the statistical relationship between the profitability of firms and the
educational qualifications of their CEOs, a simple random sample of 119 firms was
collected and data on the following two variables obtained:
Profit = Annual profit of firm i in millions of dollars
Grad = 1 if the CEO of firm i had a post-graduate degree, = 0 otherwise.
The sample mean of the observations for Profit was calculated to be 0.238
($million), with standard deviation of 0.455. There were 68 CEOs in the sample
who had a post-graduate degree. The sample covariance
between Profit and Grad was 0.013.
1. (1 mark) What is the sample mean of Grad ?
2. (2 marks) The sample variance of Grad was calculated to be 0.247. Briefly
explain how this was calculated using only the information given in this
question.
3. (3 marks) Calculate a 95% confidence interval for the proportion of firms whose
CEOs have a post-graduate degree. (Report your standard error and critical
value as well as the confidence interval itself.)
4. (2 marks) Consider a regression of the form
E ( Profit | Grad ) = β + β Grad
Use the provided sample statistics to calculate β̂ and β̂ .
5. (2 marks) Use your regression results to calculate the sample mean of profits
for firms whose CEOs do not have a post-graduate degrees. Explain how you
did the calculation.
6. (2 marks) Use your regression results to calculate the sample mean of profits
for firms whose CEOs do have a post-graduate degree. Explain how you did
the calculation.
i
i
i
i i
i
i
i i 0 1 i
0 1
Looks like you got si
In order to continue receiving th
Ginger, please sign b
SIGN BACK IN
Don't show this ag
Edit View Insert Format Tools Table
12pt Paragraph
09/06/2022, 00:21 Quiz: Final 2021SM1 with solution
https://canvas.lms.unimelb.edu.au/courses/124333/quizzes/157982/take 7/10
p 0 words
15 ptsQuestion 5
As an analyst of firms characteristics and profitability, you are considering the
relationship between firm profitability and the amount paid to their CEOs. A simple
random sample of 117 firms was collected and data on the following two variables
obtained:
Profit = Annual profit of firm i in millions of dollars
Salary = Salary of the CEO in millions of dollars
A regression of Profit on Salary is estimated, given the following equation, (with
standard errors in brackets):
Ê(Profit |Salary ) = -0.197 + 0.33 Salary
(0.071) (0.048)
Use these results to answer the following questions.
1. (2 marks) Give an interpretation of the coefficient on Salary
2. (2 marks) Construct a 95% confidence interval for the population coefficient
on Salary in this equation. Give the critical value as well as the confidence
interval.
3. (3 marks) Your boss claims there should be some statistical relationship
between firm profitability and CEO salary. Use your previous answer to
evaluate this claim.
4. (3 marks) Your boss further claims that paying a CEO a higher salary will
cause the CEO to perform better and hence produce higher firm profitability. Is
this claim confirmed by the regression results? Explain.
5. (2 marks) The R for the regression is 0.289. What does this measure and
what are its implications for this regression?
6. (3 marks) Consider the scatter plot below of Profit and Salary . What do you
observe about this plot that might be a problem for statistical inference in the
regression? Explain.