Type equation here.

Statistics Final Exam

13/06/2023
= 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.

(PT) Escreva na grelha de respostas a opção que considera válida. Caso queira alterar uma resposta,
risque a opção inicial e escreva a nova opção, de forma clara e inequívoca. Cada resposta correta
equivale a 0,5 pontos. Cada resposta incorreta desconta 0,125 pontos.

1 2 3 4 5 6 7 8 9 10

1. (ENG) Which of the following statements about hypothesis testing is false?

a. A type II error occurs when we do not reject a false null hypothesis.

b. The power of a test is the probability of not rejecting a false null hypothesis.
c. If a null hypothesis is rejected for a significance level of 2%, then using the same data, it also must
be rejected for a significance level of 4%.
d. A type I error occurs when we reject a true null hypothesis.
e. None of the above.

1
 2. (ENG) An insurance policy pays a total medical benefit consisting of two parts: one paid to the surgeon (X)
and one paid to the hospital (Y). The total benefit is X + Y.
It is known that Var(X) = 5,000; Var(Y) = 10,000, and Var (X + Y) = 17,000. Due to increasing medical costs,
the company has decided to increase Y by 10% per claim. Calculate the variance of the total benefit after
this revision has been made.

a. 18, 800
b. 19, 300
c. 19, 520
d. 20, 670
e. None of the above.
Var(X + 1.1Y) = Var(X) + Var(1.1Y) + 2Cov(X, 1.1Y)
= Var(X) + 1.12 Var(Y) + 2 * (1.1) Cov(X,Y)
Var(X,Y) = Var(X) + Var(Y) + 2Cov(X,Y)
17,000 = 5,000 + 10,000 + 2Cov(X,Y)
Cov(X,Y) = 1,000
Var(X + 1.1Y) = 5,000 + (1.1)2 * 10,000 + 2 * 1.1 * 1,000
Var(X + 1.1Y) = 19,300

3. (ENG) Suppose that a three-person board of directors will be chosen at random from a group of 8
candidates. Five of these candidates are economists and the rest are lawyers. What is the probability that
there will be at least 2 directors who are economists on the board?

a. 0.8214
b. 0.7143
c. 0.5357
d. 0.2857
e. None of the above.
X = the number of economists on the board
X ~ Hypergeometric (N, n, r)
where N=8; n=3; and r=5
P (X ≥ 2) = P(X=2) + P(X=3)
= 0.7143

2
 4. (ENG) A company administers a typing test to screen applicants for a secretarial position. In order to pass
the test, an applicant must complete the test in 50 minutes with no more than one error. Historical data
reveals the following about the population of applicants:

i. The number of test errors follows a Poisson distribution with a mean of 3 errors.
ii. The time required to complete the test follows a normal distribution with a mean of 45 minutes
and a standard deviation of 10 minutes.
iii. The number of errors and the time required to complete the test are independent.

What is the probability that an applicant chosen at random will pass the test?

a. 0.0381
b. 0.1030
c. 0.1376
d. 0.1915
e. None of the above.
X = the number of errors on the test
Y = the amount of time required to complete the test
X ~ Poisson with λ = 3
Y ~ N(µ, σ), where µ = 45 and σ = 10
From the Poisson Table:
P (X ≤ 1) = 0.199
Y− μ 50 − μ 5
P(Y ≤ 50) = P ( ≤ ) = P (Z ≤ )
σ σ 10
P(Y ≤ 50) = P(Z < 0.5) = 0.6915
P (X ≤ 1) ∗ P(Y ≤ 50) = 0.199 ∗ 0.6915 = 0.1376

5. (ENG) Let 𝑋1 , 𝑋2 and 𝑋3 be a random sample of three observations from a population with mean μ and
variance 𝜎 2 . Consider the following three-point estimators of μ:
1 1 1 1 2 1 1 2
𝛼
̂1 = 3 𝑋1 + 3 𝑋2 + 3 𝑋3 ; 𝛼
̂2 = 4 𝑋1 + 4 𝑋3 ; 𝛼
̂3 = 4 𝑋1 + 4 𝑋2 + 4 𝑋3

Which of the following statement(s) about the point estimators is (are) correct?

i. Expected value of 𝛼
̂2 is equal to μ.
ii. 𝛼
̂1 and 𝛼
̂3 are unbiased estimators of μ.

3
 iii. Among the unbiased estimators, 𝛼
̂3 has the smallest variance.

a. Only ii.
b. i. and iii.
c. ii. and iii.
d. i., ii., and iii.
e. None of the above.

6. A researcher generates a number y from a uniform distribution on the interval [0, β]. He decides to test the
null hypothesis H0 : β = 2 against the alternative hypothesis Ha : β ≠ 2 by rejecting H0 if y ≤ 0.1 or y ≥ 1.9.
What is the probability of making a Type-1 error?

a. 0.10
b. 0.05
c. 0.01
d. 0.02
e. None of the above.
P (y ≤ 0.1 or y ≥ 1.9 | β=2) =?
y ~ Uniform (0,2)
P (y ≤ 0.1 or y ≥ 1.9 | β=2) = 2 * (0.1 / 2) = 0.1
f(y) = 1/2

7. (ENG) A company establishes a fund of 1200 EUR to pay an amount, C, to any of its 20 employees who
achieve a high performance level during the coming year. Each employee has a 1% chance of achieving a
high performance level during the coming year, independent of any other employee.

Determine the maximum value of C such that the probability that the fund will be inadequate to cover all
payments for high performance is less than 0.01.

a. 240
b. 300
c. 400
d. 600
e. None of the above.

4
Let X denote the number of employees who achieve a high performance level during the coming year.
X ~ Binomial (n, p) where n=20 and p=0.01
P (X > 1,200/C) < 0.01
P(X>a) < 0.01
1-P (X≤ a) < 0.01
P (X≤ a) > 0.99
From the binomial table where n=20 and p=0.01, a=2 is the minimal integer such that
P (X≤ a) > 0.99
a = 1,200/C ➔ C=600

8. (ENG) Let X denote the amount of time (in minutes) a postal clerk spends with a customer. X is an
exponential random variable with the following density function:
1 −𝑥
𝑓(𝑥) = 𝑒 𝑘
𝑘

Suppose a customer has spent four minutes with a postal clerk. What is the probability that he or she
will spend at least an additional three minutes with the postal clerk?

−13
a. 𝑒 𝑘

−11
b. 𝑒 𝑘

−4
c. 𝑒 𝑘
−𝟑
d. 𝒆 𝒌
e. None of the above.
P(X≥7 | X≥4) = ?
−𝟕
𝑃(𝑋 ≥7) 𝒆
𝒌
P (X≥7 | X≥4) = 𝑃(𝑋 ≥4) = −𝟒
𝒆
𝒌
−𝟑
=𝒆𝒌

9. (ENG) To estimate the mean GPA of all students at a large university, a professor of economics randomly
selects 9 students from the university and finds that his sample has a mean GPA of 2.5 with a standard
deviation of 0.6. Assuming that the GPAs of the students are normally distributed, find a 99% confidence
interval for the mean GPA of all students at the university.

5
 a. 2.5  0.5642
b. 2.5  0.6710
c. 2.5  0.5140
d. 2.5  0.6500
e. None of the above.

n = 9, 𝑋̅ = 2.5, s = 0.6 and α = 0.01

𝑠
𝑋̅ ± 𝑡𝛼/2 ∗
√𝑛
From the t- table: t0.005 = 3.355
2.5  0.6710

10. (ENG) Let X denote a random variable that follows a normal distribution. The probability density function
of X is given below:

1 1 𝑥−4 2
− ( )
𝑓(𝑥) = 𝑒 2 2
2√2𝜋

Which of the following statements is correct?

a. The probability that random variable X takes the value of 4 is 0.5.

b. Approximately 95% of x-values lie between 2 and 6.
c. E (X2) is equal to 20.
d. The approximate value of the probability density function of X evaluated at point 5 means that the
probability that the random variable X takes a value greater or equal to 5 is approximately 0.824. F

X ~ N (4,2)

V(X) = E(X2) – [E(X)]2

4 = E(X2) – 16

E(X2) = 20

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.

(PT) Por favor responda às partes II e III em folhas de resposta separadas. Justifique todas as respostas o
mais claramente possível. Arredonde os resultados (finais e intermédios) a QUATRO casas decimais.

Part II [8 points]

1. [4 points] (ENG) A statistician, who attempts to predict a firm’s earnings next year, believes that the firm’s
business is very sensitive to the level of interest rates. He believes that if average interest rates next year
are more than 1% higher than this year, the probability of significant earnings growth is 0.1. If average
interest rates next year are more than 1% lower than this year, the probability of significant earnings is
estimated to be 0.8. Finally, if average interest rates next year are within 1% of this year’s rates, the
probability of significant earnings growth is estimated to be 0.5. The statistician estimates that the
probability is 0.25 that average interest rates next year will be more than 1% higher than this year, and that
the probability is 0.15 that they will be more than 1% lower than this year.
Let us define the following four events:

H = Average interest rates next year are more than 1% higher than this year.
L = Average interest rates next year are more than 1% lower than this year.
M= Average interest rates next year are within 1% of this year’s rates.
G = Firm experiences significant earnings growth.

P(H) = 0.25, P(L) = 0.15, P(M) = 0.6

P(G|H) = 0.1, P(G|L) = 0.8, and P(G|M) = 0.5

a. What is the estimated probability that both average interest rates next year will be more than 1% higher
than this year and the firm will experience significant earnings growth.
P(H∩G) = P(G|H) * P(H)
= 0.1 * 0.25
= 0.025

Answer: The probability that both average interest rates next year will be more than 1% higher than this year
and the firm will experience significant earnings growth is 2.5%.

7
 b. What is the probability that this firm will not experience significant earnings growth?

H, L and M are mutually exclusive and exhaustive events.

G∩S = G∩(HULUM)
G = (G∩H) U (G∩L) U (G∩M)
P(G) = P(G∩H) + P(G∩L) + (G∩M)
P(G) = P(G|H) *P(H) + P(G|L)*P(L) + (G|M)*P(M)
P(G) = 0.1 * 0.25 + 0.8 * 0.15 + 0.5 * 0.6
P(G) = 0.445
1-P(G) = 0.555

Answer: the probability that this firm will not experience significant earnings growth is 55.5%

c. Are the events “the firm experiences significant earnings growth” and “average interest rates next year
are more than 1% higher than this year” statistically independent? Explain.

These two events are said to be statistically independent if and only if one of the following conditions is met:
i. P(H∩G) = P(H) * P(G)
ii. P(G|H) = P(G)
iii. P(H|G) = P(H)

P(H∩G) = 0.025, P(H) * P(G) = 0.25 * 0.445 = 0.113

Since P(H∩G) ≠ P(H) * P(G), they are dependent events.

d. If the firm exhibits significant earnings growth, what is the probability that average interest rates will
have been more than 1% lower than the current year?

We can use Bayes’ rule to find the probability that average interest rates will have been more than 1% lower
than the current year given that the firm exhibits significant earnings growth.

𝑃(𝐿 ∩ G)
𝑃(𝐿|𝐺) =
𝑃(𝐺)
𝑃(𝐺|𝐿) ∗ 𝑃(𝐿)
𝑃(𝐿|𝐺) =
𝑃(𝐺|𝐻) ∗ 𝑃(𝐻) + 𝑃(𝐺|𝐿) ∗ 𝑃(𝐿) + 𝑃(𝐺|𝑀) ∗ 𝑃(𝑀)
0.8 ∗ 0.15
𝑃(𝐺|𝐿) = = 0.2697
0.445

Answer: The probability that average interest rates will have been more than 1% lower than the current year
given that the firm exhibits significant earnings growth is 26,97%

8
 2. [4 points] (ENG) A machine that produces a special type of transistor used in computers has a 10% defective
rate. The production is considered a random process where each transistor is independent of
the others.
a. A random sample of n transistors produced by the machine is taken. If the standard deviation of the
number of defective transistors in this sample is 3, what is the value of n?

Let X denote the number of defective transistors in a random sample of n transistors.

X ~ Binomial (n, p) where p=0.10

V(X) = npq → √V(X) = √np(1 − p)

3 = √np(1 − p)
32 = n * 0.09
n = 9 / 0.09 → n=100

b. Suppose that an engineer responsible for the machine randomly selects 6 transistors. What is the
probability that at least one of these transistors are defective?

Let Y denote the number of defective transistors in a random sample of 6 transistors.

Y ~ Binomial (n, p) where n=6 and p=0.10

P (Y ≥ 1) = 1 – P (Y = 0)
n
P(Y=y) = (y) ∗ py ∗ qn−y
6
P(Y = 0) = ( ) ∗ 0.100 ∗ 0.906 = 0.5314
0
P (Y ≥ 1) = 1 – 0.5314 = 0.4686

Answer: The probability that at least one of these transistors are defective is 46.86%

c. Suppose that Company XYZ receives a very large shipment of transistors produced by the machine. The
manager of company XYZ randomly selects 250 transistors from the shipment. He accepts the shipment
if fewer than 28 of these transistors are defective. What is the probability that the manager will accept
the shipment?

Let X denote the number of defective transistors in a random sample of 250 transistors from the shipment.

X ~ Binomial (n,p) where n = 250 and p = 0.10

µ = E(X) = np = 250 * 0.10 = 25
σ = √V(X) = √npq = √250 ∗ 0.10 ∗ 0.90 = 4.7434

The interval µ±3σ should lie within the range of binomial random variable X in order for the normal
approximation to be adequate, we should check whether µ±3σ ∈ [0, n]
µ ± 3σ = 25 ± 3 * 4.7434 = (8.2698, 36.7302)
9
(8.2698, 36.7302) ∈ [0, 250]

P (X < 28) = P (X ≤ 27)

P (X ≤ 27) ≈ P (X ≤ 27.5) → Continuity correction is applied.
X− μ 27.5 − μ
P(X ≤ 27.5) = P ( ≤ ) = P(Z ≤ 0.53)
σ σ

From the Z-table: P (0 < Z < 0.53) = 0.2019

P (Z ≤ 0.53) = 0.5 + 0.2019 = 0.7019

P (X < 28) ≈0.7019

Answer: The probability that the manager will accept the shipment is approximately 70.19%.

d. If the shipment in part (c) is accepted, the transportation cost to the producer of the transistors is 10,000
EUR. If the shipment is not accepted, the transportation cost is 15,000 EUR. Calculate the expected
transportation cost of the shipment for the producer.

Let T denote the transportation cost of the shipment to the producer.

E(T) = 10,000 EUR * P(X<28) + 15,000EUR * (1 - P(X<28))

E(T) = 10,000 * 0.7019 + 15,000 * (1-0.7019)
E(T) = 11,490.50 EUR

Answer: the expected transportation cost of the shipment for the producer is 11,490.50 EUR.

Part III [7 points]

3. [4 points] (ENG) Suppose that a study is conducted to examine the stress levels of students at University A.
The stress scores of students are found to follow a uniform distribution with the lowest stress score equal
to 1 and the highest stress score equal to 5.

a. What is the probability that a randomly selected student from University A has a stress score less than
3.5? Find the median stress score of the students at University A.

Let X denote the stress score of a student from University A

X ~ Uniform (1,5)
1
f(x) = = 0.25
5−1
P (1 < X < 3.5) = 0.25 * (3.5 – 1) = 0.625

10
 Answer: The probability that a randomly selected student from University A has a stress score less than 3.5
is 62.5%

Let m denote the median of the stress score distribution.

P (1 ≤ X ≤ m) = 0.5
P (1 ≤ X ≤ m) = 0.25 * (m-1) = 0.5
0.5
m−1 = →m=3
0.25

b. A random sample of 64 students from University A was taken. Describe the sampling distribution of the
sample mean stress score. Find the 90th percentile for the sampling distribution of the sample mean stress
score.
Let ̅
X denote the mean stress score of randomly selected 64 students from University A.
Since the sample size is 64, which is greater than 30, we can use the Central Limit Theorem to describe the
sampling distribution of the sample mean stress score.
According to the CLT, the sampling distribution of 𝑋̅ is approximately normal with mean 𝜇𝑥̅ = E(X) = μ and
σ
standard deviation is σx̅ = n
√
5+1
E(X) = =3
2

5−1
√V(X) = σ = = 1.1547
√12

𝜎 1.1547
𝜎𝑥̅ = = = 0.1443
√𝑛 8

̅
X ~ N(3, 0.1443)

Let a denote the 90th percentile of the sampling distribution of ̅

X.

̅ ≤ a) = 0.90
P(X

̅
X − 𝜇𝑥̅ 𝑎 − 𝜇𝑥̅
𝑃( ≤ ) = 0.90
𝜎𝑥̅ 𝜎𝑥̅

P(Z ≤ z0 ) = 0.90

From the Z-table: P (0 < Z < 1.28) = 0.3997

𝑎−3
z0 = 1.28 → 1.28 = 0.1443

a = 3.1847

Answer: The 90th percentile for the sampling distribution of the sample mean stress score is 3.1847.

11
 c. What is the probability that the mean stress score for the randomly selected 64 students from University
A is greater than 3.1?
̅
X − μx̅ 3.1 − μx̅
̅ > 3.1) = P (
P(X ≤ )
σx̅ σx̅

3.1 − 3
P (Z ≤ ) = P(Z > 0.69)
0.1443

From the Z-table:

P(Z > 0.69) = 0.5 − P(0 < Z < 0.69) = 0.5 − 0.2549 = 0.2451

Answer: The probability that the mean stress score for the randomly selected 64 students from University A is
greater than 3.1 is 24.51%.

d. The mean stress score of students at University B is claimed to be no more than 3. An educational
psychologist randomly selects 100 students from University B to check this claim and conduct a hypothesis
test. The psychologist finds that the standard deviation of the stress scores in her sample is 1.2. Using a
1% significance level, determine what the sample mean stress score needs to be to reject the claim.
Let µ denote the mean stress score of students at University B.

Ho : µ = 3
Ha : µ > 3

̅ denotes the sample mean stress score of 100 randomly selected students from University B
X
n = 100, s= 1.2

̅ − 𝜇𝑜
X
𝑍= 𝑠
⁄ 𝑛
√

α = 0.01, under the null hypothesis, Z ~ N (0,1)

From the Z-table:

P(Z > z∝ ) = 0.01 → z∝ = 2.33

The rejection region→ RR: {z > 2.33}

If the observed value of the test statistic falls into the rejection region, then the null hypothesis will be rejected.

̅
X− 3
̅ − 3 > 2.33 ∗ 0.12
> 2.33 → X
0.12
12
̅ > 3 + 0.2796 → X
X ̅ > 3.2796

If the sample mean of stress score is greater than 3.2796, then we can reject the null hypothesis and conclude
that there is evidence to support the claim that the population mean stress score of students at University B is
greater than 3 at a 1% significance level.

4. [3 points] (ENG) Three researchers, namely researcher A, researcher B and researcher C, are interested in
estimating the proportion of students at a university who smoke.

a. Out of a random sample of 400 students from this university, researcher A finds that 80 students smoke.
Using researcher A’s sample data, construct a 95% confidence interval for the proportion of students at
this university who smoke. Interpret your answer. Suppose that a doctor who works at the health center
of this university conjectures that 25% of the students at the university smoke. Is this claim supported
by the confidence interval you have constructed?

p: proportion of students at the university who smoke.

X 80
p̂ = = = 0.2
n 400

The sample size is large enough if both np̂ ≥ 15 and nq̂ ≥ 15

np̂ = 80 and nq̂ = 320

1 − α = 0.95; α = 0.05 and α⁄2 = 0.025

From the Z-table, Z0.025 = 1.96

̂q
p ̂ 0.2∗0.8
The confidence interval is: p̂ ± Z0.025 √ n = 0.2 ± 1.96√ 400
= 0.2 ± 0.0392 → (0.1608, 0,2392)

We are 95% confident that the proportion of students at this university who smoke is between 0.1608 and 0.2392.

The value of 25% lies outside of the interval, so we have convincing evidence that the doctor’s conjecture is wrong.

b. If researcher B wants to construct an 80% confidence interval for the proportion of students who smoke
that has a sampling error smaller than 2%, how big of a sample does she need?

p (1 − p)
SE = 𝑍𝛼/2 √
n

1 − α = 0.8; α = 0.2 and α⁄2 = 0.1

13
From the Z-table, Z0.1 = 1.28

0.5 ∗ (1 − 0.5)
1.28 ∗ √ < 0.02
𝑛

0.5 ∗ 0.5
1.282 ∗ < 0.022
𝑛

1.282 ∗ 0.25
<𝑛
0.022

1024 < 𝑛

Researcher B needs to randomly select 1025 or more students to ensure that the sampling error of the 80%
confidence interval is smaller than 2%.

c. Researcher C’s colleague claims that the proportion of students at the university who smoke is less than
20%. To test his colleague’s claim, researcher C randomly selects 400 students. If the p-value of
researcher C’s test is 0.025, what is the sample proportion of students who smoke?

Ho : p = 0.20
Ha : p < 0.20

Since np0 = 400 * 0.2 > 15 and nq0 = 400 * 0.8 > 15, the sample is considered large.

If the p-value of the test is 0.025, this means that the value of the test statistic would be -1.96 (i.e., Z = -1.96)
From the Z-table: P (Z < z0) = 0.025 → z0 = - 1.96

𝑝̂ − 𝑝0
𝑍=
𝑝 𝑞
√ 0 0
𝑛
p̂ − 0.2
−1.96 =
√0.2 ∗ 0.8
400
p̂ − 0.2
−1.96 = → −1.96 ∗ 0.02 = p̂ − 0.2
0.02
−0.0392 = p̂ − 0.2 → p̂ = 0.1608

Answer: The sample proportion of students who smoke is 16.08%.

14