Please Print Your Name: ___________________________

THE UNIVERSITY OF HONG KONG

FACULTY OF BUSINESS AND ECONOMICS
2020 - 2021 (SEMESTER TWO) MIDTERM EXAMINATION

Course: IIMT3636DEF Decision and Risk Analysis I

Instructors: Dr. Hu, Xing & Prof. Wan, Zhixi (Subclasses F, G, and H)
Dr. Zhang, Wei (Subclass E)

Date: March 6, 2021 (Saturday)

Time: 10:00 a.m. - 12:00 p.m.

Please Enter Your Student UID: __________________________

Please Enter the Section You Have Registered: _____________

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.
Calculator Name:______________________________ Model:_____________________________

IMPORTANT NOTE
This paper is NOT to be taken away and NOT for circulation. You may not copy, reproduce, distribute, publish, display,
perform, modify, create derivative works, transmit, or in any way exploit any such content, nor may you distribute any
part of this content over any network. Copying or storing any content contained in this paper is expressly prohibited.

Failing to comply with the above will be regarded as having engaged in an act of misconduct. The Faculty views any act
of this nature as of the utmost seriousness. It is a form of misconduct that the University and the Faculty do not tolerate.
Such a case would be brought before the University’s Disciplinary Committee for a formal hearing, which could possibly
result in a penalty inclusive of suspension of study or expulsion from the University.

Attempt all questions. Points are as indicated.

Section Points
I. True/False 10
II. Multiple Choice 20
III. Short Answers 70
Total 100

Find the equations and the z-table in the Appendix.

This paper has 10 pages (including this cover page).

Please Print Your Name: ___________________________

SECTION I: TRUE/FALSE. Choose 'A' if the statement is true and 'B' if the statement is false. (1
point for each question.)

1) Given two statistically independent events (A,B), the joint probability of 1) _______
P(AB) = P(A) + P(B).

2) To estimate the probability of an event, we must conduct repeated 2) _______

experiments.

3) If a bucket has three black balls and seven green balls, and we draw balls 3) _______
without putting them back, the probability of drawing a green ball is
independent of the number of green balls previously drawn.

4) To measure the quality of a consulting firm’s market survey, we can use the 4) _______
historical data to calculate the probability of a favorable market given a
positive prediction from the survey.

5) The following two events, A and B, are independent. 5) _______

Venn Diagram: B
A

6) A decision tree can deal with decision problems in which the states of nature 6) _______
depend on all the decisions made previously.

7) If a person is a risk-seeker, then s/he would always pick the option that has a 7) _______
lower expected payoff.

8) The decision processes of maximizing expected monetary value (EMV) and 8) _______
minimizing expected opportunity loss (EOL) should lead us to choose the same
alternatives.

9) If a simple linear regression between Y and X has a p-value that is bigger than 9) _______
5%, we can conclude that there is no relationship between the two.

10) The SST measures the total variability in the dependent variable about the 10) _______
regression line.

SECTION II: MULTIPLE CHOICE. Choose the one alternative that best completes the statement
or answers the question. (2 points for each question.)

11) What is the probability of drawing two cards with the same color from a full 11) _______
deck of 52 cards?
A) 25/51 B) 26/51 C) 1/13 D) 3/13 E) None of the above
Please Print Your Name: ___________________________

12) At a university with 1,000 business majors, there are 200 business students 12) _______
enrolled in an introductory statistics course. Of these 200 students, 50 are also
enrolled in an introductory accounting course. There are an additional 250
business students enrolled in accounting but not enrolled in statistics. If a
business student is selected at random, what is the probability that the student
is either enrolled in accounting or statistics, but not both?
A) 0.05
B) 0.50
C) 0.45
D) 0.40
E) None of the above

13) A company is considering producing some new Gameboy electronic games. 13) _______
Based on past records, management believes that there is a 70 percent chance
that each of these will be successful and a 30 percent chance of failure. Market
research may be used to revise these probabilities. In the past, the successful
products were predicted to be successful based on market research 90 percent
of the time. However, for products that failed, the market research predicted
these would be successes 20 percent of the time. If market research is
performed for a new product, what is the probability that the results indicate
an unsuccessful market for the product and the product is actually successful?
A) 0.21 B) 0.07 C) 0.06 D) 0.14 E) 0.63

14) Decision trees are particularly useful when 14) _______

A) the opportunity loss table is available.
B) a sequence of decisions must be made.
C) perfect information is available.
D) formulating a conditional values table.
E) all possible outcomes and alternatives are not known.

15) A market research survey is available for $8,000. Using a decision tree analysis, 15) _______
it is found that the expected monetary value with the survey is $75,000. The
expected monetary value with no survey is $65,000. What is the expected value
of sample information?
A) $10,000
B) $2,000
C) $6,000
D) $18,000
E) None of the above

16) Which of the following is not an assumption for linear regression? 16) _______
A) Errors are independent (random sampling)
B) Errors are normally distributed
C) Errors have a zero mean
D) Errors have a constant variance
E) The independent variables are independent of each other
Please Print Your Name: ___________________________

17) The following is an opportunity loss table. What decision should be made 17) _______
based on the minimax regret criterion?

States of Nature
Alternatives A B C
1 0 90 85
2 50 0 110
3 75 80 0

A) Alternative 1
B) Alternative 2
C) Alternative 3
D) State of Nature A
E) Does not matter

18) A healthcare executive is using regression to predict total revenues. She has 18) _______
decided to include both patient length of stay and insurance type in her model.
Insurance type can be grouped into the following categories: Medicare, Self-
Pay, and Charity. Her model is
A) 𝑌𝑌 = 𝑏𝑏0 + 𝑏𝑏1 𝑋𝑋1 + 𝑏𝑏2 𝑋𝑋2 + 𝑏𝑏3 𝑋𝑋3.
B) 𝑌𝑌 = 𝑏𝑏0 + 𝑏𝑏1 𝑋𝑋1 + 𝑏𝑏2 𝑋𝑋2 + 𝑏𝑏3 𝑋𝑋3 + 𝑏𝑏4 𝑋𝑋4.
C) 𝑌𝑌 = 𝑏𝑏0 + 𝑏𝑏1 𝑋𝑋1.
D) 𝑌𝑌 = 𝑏𝑏0 + 𝑏𝑏1 𝑋𝑋1 + 𝑏𝑏2 𝑋𝑋2.
E) None of the above.

19) Which of the following is true regarding a regression model with 19) _______
multicollinearity, a high r2 value, and a low F-test significance level?
A) The significance level for the F-test is not valid.
B) The interpretation of the coefficients is valuable.
C) The significance level tests for the coefficients are not valid.
D) The high r2 value is due to the multicollinearity.
E) The model is not a good prediction model.

20) The F-test for a simple linear regression gives a P-value of 0.0525 and 𝑏𝑏1 , the 20) _______
estimated coefficient for the independent variable, is 3.5265. Which of the
following is true?
A) The F-statistic is greater than the F-score corresponding to α = 0.05.
B) If the independent variable is 5, the best estimate for the dependent variable
is 5*3.5265.
C) The dependent and independent variables are significantly correlated at the
level of α = 0.05.
D) The hypothesis test for 𝛽𝛽1 > 0 is significant at the level of 0.05.
E) None of the above.
Please Print Your Name: ___________________________

SECTION III: SHORT ANSWERS. Show all necessary steps to get full credit. (See detailed point
allocation below.)

21) William has long desired to be a YouTuber making fancy videos he loves. Now he has a full-
time job in a bank, so he only uses his spare time to create videos. Currently, he only produces
one video per week, and he has 800 subscribers. Suppose each of his videos will be viewed by
a subscriber with probability p = 0.6. Hence, we know that X, the number of subscribers who
will view his next video, follows a Binomial distribution B(n,p), with n = 800.
Define
𝑋𝑋 − 𝑛𝑛𝑛𝑛
𝑍𝑍 = .
�𝑛𝑛𝑛𝑛(1 − 𝑝𝑝)
We can prove that Z approximately follows the Standard Normal distribution.
(a) Suppose a video must have at least 500 views before YouTube starts to recommend the
video to nonsubscribers. What is the probability that his next video will be viewed by at least
500 of his subscribers? (10 points)

(b) Suppose William decides to resign from the bank and starts to be a full-time YouTuber. He
estimates that, by tripling his video production per week, the number of subscribers would go
to 2,000. If this is true and the probability of viewing p does not change, then what is the
probability that a new video will have more than 1,000 views from the subscribers? (10 points)
Please Print Your Name: ___________________________

22) The ABC Co. is considering a new consumer product. They believe there is a probability of 0.4
that the XYZ Co. will come out with a competitive product. If ABC adds an assembly line for
the product and XYZ does not follow with a competitive product, their expected profit is
$40,000; if they add an assembly line and XYZ does follow, they will expect a loss of $10,000. If
ABC adds a new plant and XYZ does not produce a competitive product, they expect a profit
of $600,000; if XYZ does compete for this market, ABC expects a loss of $100,000.

(a) Develop a decision tree. (5 points)

(b) Determine the Expected Monetary Value of each decision. (5 points)

(c) Determine the Expected Opportunity Loss of each decision. (5 points)

(d) Calculate the Expected Value of Perfect Information (EVPI). (5 points)

Please Print Your Name: ___________________________

23) Bob White is conducting research on individual adult’s monthly expenses for medical care,
including over the counter medicine. His dependent variable is monthly expenses for medical
care while his independent variables are number of family members (𝑋𝑋1 ) and insurance type
(government funded, private insurance and no insurance). He has coded insurance type as the
following: 𝑋𝑋2 = 1 if government funded; 𝑋𝑋3 = 1 if private insurance. Below is his excel output:

(a) According to the F test, is this model significantly effective overall? (4 points)

(b) What is the prediction equation? (4 points)

(c) What is the predicted monthly expense for a family of four with private insurance? (4
points)
Please Print Your Name: ___________________________

(d) Between the government funded and private insurances, which one is predicting lower
medical expenses on average? Which one is predicting significantly lower medical expenses in
contrast to no insurance? (4 points)

(e) To increase the adjusted r squared, what other factors may Bob include in this model?
Suggest at least two important factors. (4 points)

24) Explain why a nonlinear relationship may exist between Y and X, while a linear relationship
does not exist between them. (10 points; Word limit: 50)
Please Print Your Name: ___________________________

Appendix: Equations and the z-Table

(1) 𝑃𝑃(𝐴𝐴 ∪ 𝐵𝐵) + 𝑃𝑃(𝐴𝐴 ∩ 𝐵𝐵) = 𝑃𝑃(𝐴𝐴) + 𝑃𝑃(𝐵𝐵).

(2) Conditional probability: P(A|B) = P(AB)/P(B).

𝑃𝑃(𝐵𝐵|𝐴𝐴)𝑃𝑃(𝐴𝐴)
(3) Bayes’ formula: 𝑃𝑃(𝐴𝐴|𝐵𝐵) = , where A and A’ are mutually exclusive and
𝑃𝑃(𝐵𝐵|𝐴𝐴)𝑃𝑃(𝐴𝐴)+𝑃𝑃(𝐵𝐵|𝐴𝐴′ )𝑃𝑃(𝐴𝐴′ )

collectively exhaustive.

(4) The expected value of a discrete probability distribution: 𝐸𝐸(𝑋𝑋) = ∑ 𝑋𝑋𝑖𝑖 ∙ 𝑃𝑃(𝑋𝑋𝑖𝑖 ).
𝑋𝑋−𝜇𝜇
(5) For a normal random variable 𝑋𝑋~𝑁𝑁(𝜇𝜇, 𝜎𝜎 2 ), the transformed variable 𝑍𝑍 = is standard normal.
𝜎𝜎
2 2
(6) 𝑆𝑆𝑆𝑆𝑆𝑆 = ∑(𝑌𝑌 − 𝑌𝑌�)2 , 𝑆𝑆𝑆𝑆𝑆𝑆 = ∑�𝑌𝑌 − 𝑌𝑌�� , 𝑆𝑆𝑆𝑆𝑆𝑆 = ∑�𝑌𝑌� − 𝑌𝑌�� , and 𝑆𝑆𝑆𝑆𝑆𝑆 = 𝑆𝑆𝑆𝑆𝑆𝑆 + 𝑆𝑆𝑆𝑆𝑆𝑆, wherein 𝑌𝑌� is the

mean value and 𝑌𝑌� is the predicted value.

𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆/(𝑛𝑛−𝑘𝑘−1)
(7) 𝑟𝑟 2 = =1− , 𝑟𝑟 = ±√𝑟𝑟 2 , and Adj. 𝑟𝑟 2 = 1 − , wherein n is the sample size and k
𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆/(𝑛𝑛−1)

is the number of independent variables.

(8) The F-statistic = MSR/MSE = (SSR/k)/(SSE/(n – k – 1)).

(9) The expected value of perfect information EVPI = EVwPI (before paying for the information) –

EMV (given by the best option without the information);

The expected value of sample information EVSI = EVwSI (before paying for the information) –

EMV (given by the best option without the information).

Please Print Your Name: ___________________________

END OF PAPER