Statistics Final Exam

07/01/2023
= Time Allowed: 2h30m =

Name/Nome:______________________________________________________________________

N.º___________________ Pages Delivered/ Folhas entregues: ___________________________

Part I [5 points]

(ENG) Write in the response grid the option you consider valid. If you want to change an answer, cross
out the initial option and mark the new option, clearly and unambiguously. Each correct answer earns you
0.5 points. Each incorrect answer sets you back 0.125 points.

1. (ENG) The time taken to deliver a pizza (in minutes) is a random variable, denoted by 𝑋, with the following
probability density function:
𝑘, 𝑖𝑓 20 ≤ 𝑥 ≤ 70
𝑓(𝑥) = {
0, 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒
Given that the delivery time is at least 30 minutes, what is the probability that the time to deliver a pizza is
less than 40 minutes?

a. 0.35
b. 0.20
c. 0.25
d. 0.30
e. None of the above.

40 − 30
𝑃(30 ≤ 𝑋 < 40) 70 − 20 0.2
𝑃(𝑋 < 40|𝑋 ≥ 30) = = = = 0.25
𝑃(𝑋 ≥ 30) 70 − 30 0.8
70 − 20

1
 2. (ENG) A public health researcher examines the medical records of a group of 1000 men who died in 2019
and finds that 220 of the men died from causes related to heart disease. Moreover, 200 of the 1000 men
had at least one parent who suffered from heart disease, and, of these 200 men, 100 died from causes
related to heart disease. Calculate the probability that a man randomly selected from this group died of
causes related to heart disease, given that neither of his parents suffered from heart disease.

a. 0.150
b. 0.125
c. 0.225
d. 0.170
e. None of the above.

A: man died from causes related to heart disease

B: man had at least one parent who suffered from heart disease

220 200
𝑃(𝐴) = = 0.22; 𝑃(𝐵) = = 0.2
1000 1000
120
𝑃(𝐴|𝐵̅) = = 0.15
800

3. (ENG) Which of the following is true regarding hypothesis testing?

a. A Type II error occurs when a true null hypothesis is rejected.

b. The significance level of a test is the probability that the null hypothesis is false.
c. The power of a test is the probability of rejecting a null hypothesis that is false.
d. The p-value of a test is the probability that the null hypothesis is false.
e. None of the above.

4. (ENG) A contractor that plans to bid on a project is interested in estimating the total cost of a project that
he wants to bid on. The cost of material for the project will be 4,000 EUR and his labor cost will be 800 EUR
per day. The probability distribution for the number of days to complete the project, which is denoted by
𝑋, is given below.

𝑥 4 5 6 7
𝑃(𝑥) 0.1 0.3 ? 0.2

2
 What is the mean of the total cost of the project?

a. 9,130 EUR
b. 8,560 EUR
c. 8,845 EUR
d. 9,415 EUR
e. None of the above.

𝑃(𝑋 = 6) = 1 − 0.1 − 0.3 − 0.2 = 0.4

𝐶 – Total cost of the project, 𝐶 = 4000 + 800𝑋
𝐸(𝐶) = 4000 + 800𝐸(𝑋) = 4000 + 800 × (4 × 0.1 + 5 × 0.3 + 6 × 0.4 + 7 × 0.2) = 4000 + 800 × 5.7
= 8,560 𝐸𝑈𝑅

5. (ENG) The lifetime of a printer costing 200 EUR is exponentially distributed with mean 2 years. The
manufacturer agrees to pay a full refund to a buyer if the printer fails during the first year following its
purchase, a one-half refund if it fails during the second year, and no refund for failure after the second year.
The expected total amount of refunds from the sale of 100 printers is approximately:

a. 6,321 EUR
b. 7,358 EUR
c. 10,257 EUR
d. 12,642 EUR
e. None of the above.

𝑋 – the lifetime of a printer, 𝑋~𝐸𝑥𝑝(2)

1
𝑃(𝑋 ≤ 1) = 1 − 𝑒 −2 = 0.3935
1 2
𝑃(1 < 𝑋 ≤ 2) = 𝑒 −2 − 𝑒 −2 = 0.2387

𝑌 – amount of refund for a printer

𝐸(𝑌) = 200 × 𝑃(𝑋 ≤ 1) + 100 × 𝑃(1 < 𝑋 ≤ 2) = 200 × 0.3935 + 100 × 0.2387 = 102.57 𝐸𝑈𝑅
For 100 printers: 100 × 102.57 = 10,257 𝐸𝑈𝑅

6. (ENG) A company has two assembly lines. The number of accidents that occur on each of the assembly lines
has a Poisson distribution with an average of 3 accidents per month. Assume that the performances of these
assembly lines are independent of one another. The probability that a total of 2 accidents occur in the
company in any given month is approximately:

3
 a. 0.045
b. 0.022
c. 0.011
d. 0.033
e. None of the above.

X – number of accidents that occur in line 1, 𝑋~𝑃𝑜𝑖𝑠𝑠𝑜𝑛(3)

Y – number of accidents that occur in line 2, 𝑌~𝑃𝑜𝑖𝑠𝑠𝑜𝑛(3)

𝑃(𝑡𝑜𝑡𝑎𝑙 𝑜𝑓 2 𝑎𝑐𝑐𝑖𝑑𝑒𝑛𝑡𝑠) = 𝑃(𝑋 = 2, 𝑌 = 0) + 𝑃(𝑋 = 1, 𝑌 = 1) + 𝑃(𝑋 = 0, 𝑌 = 2)

= (0.423 − 0.199) × 0.05 + (0.199 − 0.05)2 + 0.05 × (0.423 − 0.199) = 0.0446

7. (ENG) The minimum number of times a fair coin must be tossed so that the probability of getting at least
one head is more than 80% is:

a. 3
b. 4
c. 5
d. 6
e. None of the above.

X – number of heads obtained in n tosses, 𝑋~𝐵𝑖𝑛(𝑛, 0.5)

𝑃(𝑋 ≥ 1) > 0.8
1
1 − 𝑃(𝑋 = 0) > 0.8 ; ( )𝑛 < 0.2
2
𝑛=3

8. (ENG) A researcher wants to test a claim that more than 10% of individuals are left-handed in Portugal. He
randomly selects 1000 individuals and reports that the observed value of the test statistic is 2.31. Suppose
that using the researcher’s sample data, you are interested in testing a null hypothesis that the proportion
of left-handed individuals in Portugal is 0.10 against the alternative hypothesis that the proportion is
different from 0.10. What is the p-value of your test?

a. 0.9896
b. 0.0104
c. 0.0208
4
 d. 0.4896
e. None of the above.

𝐻0 : 𝑝 = 0.1
𝐻𝑎 : 𝑝 ≠ 0.1

𝑍 = 2.31
𝑝 − 𝑣𝑎𝑙𝑢𝑒
= 𝑃(𝑍 ≥ 2.31) = 0.5 − 𝑃(0 ≤ 𝑍 ≤ 2.31) = 0.0104
2
𝑝 − 𝑣𝑎𝑙𝑢𝑒 = 0.0208

9. (ENG) Let 𝑋1 , 𝑋2 , 𝑋3 and 𝑋4 be a random sample of four observations from a population with mean µ and
variance 𝜎 2 . Consider the following three point estimators of µ:

𝜃̂ = 0.10 X1 + 0.40 X2 + 0.40 X3 + 0.10 X4

𝛽̂ = 0.20 X1 + 0.30 X2 + 0.30 X3 + 0.20 X4
𝛼̂ = 0.20 X1 + 0.20 X 2 + 0.20 X3 + 0.20 X4

Which of the following statements is true?

a. 𝛽̂ is an unbiased estimator of µ but 𝜃̂ is a biased estimator of µ.
b. Both 𝛽̂ and 𝛼̂ are unbiased estimators of µ.
c. 𝛼̂ is an unbiased estimator of µ and has the smallest variance among these three estimators.
̂ is an unbiased estimator of µ and has the largest variance among these three estimators.
d. 𝜽
e. None of the above.

𝐸(𝜃̂) = 0.1𝐸(𝑋1 ) + 0.4𝐸(𝑋2 ) + 0.4𝐸(𝑋3 ) + 0.1𝐸(𝑋4 ) = 0.1𝜇 + 0.4𝜇 + 0.4𝜇 + 0.1𝜇 = 𝜇

𝐸(𝛽̂ ) = 0.2𝜇 + 0.3𝜇 + 0.3𝜇 + 0.2𝜇 = 𝜇

𝐸(𝛼̂) = 0.2𝜇 + 0.2𝜇 + 0.2𝜇 + 0.2𝜇 = 0.8𝜇
𝑉(𝜃̂) = 0.12 𝜎 2 + 0.42 𝜎 2 + 0.42 𝜎 2 + 0.12 𝜎 2 = 0.34𝜎 2

𝑉(𝛽̂ ) = 0.22 𝜎 2 + 0.32 𝜎 2 + 0.32 𝜎 2 + 0.22 𝜎 2 = 0.26𝜎 2

𝑉(𝛼̂̂ ) = 0.22 𝜎 2 + 0.22 𝜎 2 + 0.22 𝜎 2 + 0.22 𝜎 2 = 0.16𝜎 2

10. (ENG) The ages of students at Católica Porto Business School are normally distributed with a mean of 22
years and standard deviation of 4 years. A random sample of 64 Católica students is selected. What is the
probability that the sample mean will be between 21 and 23 years?

5
 a. 0.0987
b. 0.9544
c. 0.4772
d. 0.1974
e. None of the above.

Let X denote the age of a student at the Católica Porto Business School
𝑋~𝑁(22,4)
4
𝑛 = 64: 𝑋̅~𝑁 (22, )
√64
21 − 22 23 − 22
𝑃(21 < 𝑋̅ < 23) = 𝑃 ( <𝑍< ) = 𝑃(−2 < 𝑍 < 2) = 2 × 𝑃(0 < 𝑍 < 2) = 2 × 0.4772
0.5 0.5
= 0.9544

6
(ENG) Please Answer Part II and Part III in Separate Answer Sheets. Justify your answers as clearly as
possible. Use FOUR decimal places to round (final or intermediate) results.

Part II [8 points]

1. [4 points] (ENG) Consider two boxes A and B filled with yellow and red balls. Box A has 4 yellow and 6 red
balls. Box B has 3 yellow and 3 red balls. First you choose one of the two boxes. Box A has a probability of
0.4 to be chosen and box B has a probability of 0.6 to be chosen. Let us assume that you randomly pick,
without replacement, 3 balls out of the chosen box.
A: box A was chosen
B: box B was chosen

(𝐴) =0.4, P(B)=0.6

a) What is the probability that you have picked 2 yellow and 1 red balls?

Let X denote the number of yellow balls picked from box A

𝑋~𝐻𝑦𝑝𝑒𝑟𝑔𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐 (10,4,3)
𝐶24 𝐶16
𝑃(𝑋 = 2) = 10 = 0.3
𝐶3

Ley Y denote the number of yellow balls picked from box B

𝑌~𝐻𝑦𝑝𝑒𝑟𝑔𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐 (6,3,3)
𝐶23 𝐶13
𝑃(𝑌 = 2) = 6 = 0.45
𝐶3

C: the 2 yellow and 1 red balls were picked

𝑃(𝐶) = 𝑃(𝐴 ∩ 𝐶) + 𝑃(𝐵 ∩ 𝐶) = 𝑃(𝐶|𝐴)𝑃(𝐴) + 𝑃(𝐶|𝐵)𝑃(𝐵) = 0.3 × 0.4 + 0.45 × 0.6 = 0.39

The probability of picking 2 yellow and 1 red balls is 39%.

b) Given that you have picked 2 yellow and 1 red balls, what is the probability that you picked them
from box B?

𝑃(𝐵∩𝐶) 0.45×0.6 0.27

𝑃(𝐵|𝐶) = = = = 0.6923
𝑃(𝐶) 0.39 0.39

The probability of choosing box B, given that 2 yellow and 1 red balls were picked, is approximately 69%.

7
 c) Consider the event “box B was chosen” and the event “2 yellow and 1 red balls were picked”. Are
these two events independent? Explain.

𝑃(𝐵|𝐶) = 0.6923
𝑃(𝐵) = 0.6

Since 𝑃(𝐵|𝐶) ≠ 𝑃(𝐵), the events “box B was chosen” and “2 yellow and 1 red balls were picked” are not
independent.

OR

𝑃(𝐶|𝐵) = 0.45 ≠ 0.39 = 𝑃(𝐶)

OR

𝑃(𝐶 ∩ 𝐵) = 0.27 ≠ 0.234 = 𝑃(𝐶)𝑃(𝐵)

2. [4 points] (ENG) The annual profits (in billions of dollars) of multinational companies A, B, C and D are
assumed to be normally distributed.

a) Company A’s annual profit is normally distributed with a mean of 2 billion $ and a standard
deviation of 1 billion $. Calculate the probability that company A’s profit in a year is at most 2.5
billion $ given that the profit in the year is positive.

Let 𝑋𝐴 be the annual profit of company A, 𝑋𝐴 ~𝑁(2,1)

0−2 2.5 − 2
𝑃(0 < 𝑋𝐴 ≤ 2.5) 𝑃 ( 1 < 𝑍 ≤ 1 ) = 𝑃(−2 < 𝑍 ≤ 0.5)
𝑃(𝑋𝐴 ≤ 2.5|𝑋𝐴 > 0) = =
𝑃(𝑋𝐴 > 0) 0−2 𝑃(𝑍 > −2)
𝑃 (𝑍 > )
1
𝑃(0 < 𝑍 < 2) + 𝑃(0 < 𝑍 ≤ 0.5) 0.4772 + 0.1915
= = = 0.6843
0.5 + 𝑃(0 < 𝑍 < 2) 0.5 + 0.4772

The probability that company A’s profit in a year is at most 2.5 billion $ given that the profit in the year is
positive is approximately 68%.

b) Company B’s annual profit has a normal distribution with variance equal to its mean. In a certain
year, the probability that company B makes a loss (i.e., a negative annual profit) is 1%. Calculate
company B’s expected annual profit.

8
 Let 𝑋𝐵 be the annual profit of company B, 𝑋𝐵 ~𝑁(𝜇, 𝜎), where 𝜎 = √𝜇
−𝜇 𝜇
𝑃(𝑋𝐵 < 0) = 0.01 ⇔ 𝑃 (𝑍 < ) = 0.01 ⇔ 𝑃 (0 < 𝑍 < ) = 0.49
√𝜇 √𝜇
𝜇
From the table, = 2.33 ⇔ √𝜇 = 2.33 ⇔ 𝜇 = 5.4289
√𝜇

The expected annual profit for company B is 5.4289 billion $.

c) The annual profits of companies C and D are normally distributed with the same mean. The
variance of company D’s annual profit is 2.25 times the variance of company C’s annual profit. The
14th percentile of company C’s annual profit is the same as the pth percentile of company D’s
annual profit. Calculate p.

𝑋𝐶 – annual profit of company C

𝑋𝐷 – annual profit of company D

𝑋𝐶 ~𝑁(𝜇, 𝜎), and 𝑋𝐷 ~𝑁(𝜇, 1.5𝜎)

𝑐−𝜇
𝑃(𝑋𝐶 < 𝑐) = 0.14 ⇔ 𝑃 (𝑍 < ) = 0.14
𝜎
𝑐−𝜇
From the table, = −1.08 ⇔ 𝑐 = 𝜇 − 1.08𝜎
𝜎
𝜇 − 1.08𝜎 − 𝜇
𝑃(𝑋𝐷 < 𝜇 − 1.08𝜎) = 𝑃 (𝑍 < ) = 𝑃(𝑍 < −0.72) = 0.2358
1.5𝜎

The 14th percentile of company C’s annual profit corresponds to the 23.58th percentile of company D’s
annual profit.

Part III [7 points]

3. [4 points] (ENG) A drug company claims that its new drug used to treat symptoms of allergies causes a
headache in 5% of individuals who take it. Assume that the drug company’s claim is correct in answering
parts (a) and (b).

a) A medical researcher randomly selects 5 individuals who take the drug to treat symptoms of
allergies. What is the probability that none of the individuals will get a headache?

Let X be the number of individuals who get a headache after taking the drug in a random sample of 5
individuals.

9
𝑋~𝐵𝑖𝑛(5,0.05)

𝑃(𝑋 = 0) = 𝐶05 × 0.050 × 0.955 = 0.7738

The probability that none of the 5 individuals will get a headache after taking the drug is approximately
77%.

b) Another medical researcher randomly selects 400 individuals who take the drug to treat symptoms
of allergies. The probability is 0.10 that the sample proportion of individuals who get a headache
after taking the drug exceeds the population proportion by how much?

𝑛 = 400 and 𝑝 = 0.05

Since 𝑛𝑝 = 400 × 0.05 = 20 > 15 and 𝑛𝑞 = 400 × 0.95 = 380 > 15 , the CLT applies.
0.05 × 0.95
𝑝̂ ~𝑁 (0.05, √ )
400

0.05 + 𝑐 − 0.05 𝑐
𝑃(𝑝̂ > 0.05 + 𝑐) = 0.1 ⇔ 𝑃 (𝑍 > ) = 0.1 ⇔ 𝑃 (𝑍 > ) = 0.1
0.0109 0.0109
𝑐
From the table, = 1.28 ⇔ 𝑐 = 0.0140
0.0109

There is a 0.10 probability that the sample proportion of individuals who get a headache after taking the
drug exceeds the population proportion by 0.014.

c) A doctor believes that the drug company’s claim is incorrect and more than 5% of those who take
the drug get a headache. In a random sample of 400 individuals who took the drug, 32 individuals
were reported to get a headache. Using the sample data, test the doctor’s belief at the 1% level of
significance.

𝐻0 : 𝑝 = 0.05
𝐻𝑎 : 𝑝 > 0.05 where p denotes the proportion of individuals who get a headache after taking the drug.

32
𝑛 = 400, 𝑝̂ = = 0.08
400

Since 𝑛𝑝0 = 400 ∗ 0.05 = 20 > 15 and 𝑛𝑞0 = 400 × 0.95 = 380 > 15, the sample size is sufficiently
large to apply the CLT.

𝑝̂ − 𝑝0 0.08 − 0.05 0.03

𝑍= = = = 2.75
𝑝0 𝑞0 0.0109
√ √0.05 × 0.95
𝑛 400

10
 𝛼 = 1%. Under the null hypothesis, the test statistic follows standard normal distribution.

From the table: 𝑅𝑅 = {𝑧 > 2.33}

The observed value of the test statistic falls into the rejection region. Therefore, at the 1% significance
level, we reject the null hypothesis, suggesting that the sample data provide sufficient evidence that the
proportion of individuals who get a headache after taking the drug exceeds 0.05.

4. [3 points] (ENG) A battery manufacturer advertises that the mean lifetime for its batteries has increased
because of improvements in the production process. To evaluate the credibility of the manufacturer’s
advertising campaign, an independent consumer group tested a random sample of 64 batteries. The sample
mean lifetime and standard deviation were found to be 47.5 hours and 4.2 hours, respectively.

X – lifetime of a battery
𝜇 is unknown
𝑛 = 64, 𝑥̅ = 47.5, 𝑠 = 4.2

a) Using the consumer group’s sample data, construct a 98% confidence interval for the population
mean lifetime. Interpret your answer.
𝑛 > 30, so, the CLT applies.
𝑠 4.2
𝐶𝐼98% = 𝑥̅ ± 𝑧0.01 = 47.5 ± 2.33 × = 47.5 ± 1.2233 = [46.2767,48.7233]
√𝑛 √64

We are 98% confident that the population mean lifetime of the batteries is between 46.2767 and 48.7233
hours. In other words, at the 98% confidence level, we estimate the population mean lifetime of the
batteries to be between 46.2767 and 48.7233 hours.

b) Assuming that the confidence level remains constant, what sample size is needed to reduce by half
the margin of error of the confidence interval in part (a).
(SE in part (a) was 1.2233. Half of the SE in part (a) is considered to be SE here.)

𝑆𝐸 = 0.6116
𝑠 4.2
𝑧0.01 = 𝑆𝐸 ⇔ 2.33 × = 0.6116 ⇔ √𝑛 = 16.0007 ⇔ 𝑛 = 256.0224
√𝑛 √𝑛

The sample size needed to reduce by half the margin of error of the confidence interval in part (a) is 257
batteries.

c) Using a random sample, the research group tests the following null hypothesis 𝐻0 : 𝜇 = 𝜇0 , against
the alternative hypothesis 𝐻𝑎 : 𝜇 > 𝜇0 where 𝜇 denotes the population mean lifetime.

11
 The null hypothesis is not rejected at the 5% significance level. Does this necessarily imply that 𝜇0
is contained in the 90% confidence interval for 𝜇 if the observed sample mean is larger than 𝜇0 ?
Explain your answer.
𝐻0 : 𝜇 = 𝜇0
𝐻𝑎 : 𝜇 > 𝜇0

The null hypothesis is not rejected at the 5% significance level. Therefore, the observed value of test
statistic does not fall into the rejection region.

𝑅𝑅 = {𝑧 > 𝑧0.05 } (From the table 𝑧0.05 = 1.645)

𝑥̅ − 𝜇0
𝑠 ≤ 𝑧0.05
√𝑛

If the observed sample mean is larger than 𝜇0 (i.e., 𝑥̅ > 𝜇0 ), then the observed value of the test statistic
is between zero and 𝑧0.05

𝑥̅ − 𝜇0 𝑠
0< 𝑠 ≤ 𝑧0.05 ⇔ 𝑥̅ − 𝑧0.05 ≤ 𝜇0 < 𝑥̅
√𝑛
√𝑛

The endpoints of the 90% confidence interval for 𝜇 :

𝑠
𝐶𝐼90% = 𝑥̅ ± 𝑧0.05
√𝑛

Therefore, 𝜇0 is contained in the the 90% confidence interval for 𝜇.

12