Exercise Sheet 2

Statistical Inference: Hypothesis Testing

1. A researcher thinks that a proportion of families in a certain area who owns more than
two cars is less than 15%. In order to test this, a random sample of n = 200 families is
selected and it turns out that 22 families in that sample own more than two cars.

a) Is the researcher right? (α = 0.05)

b) What is the required sample proportion for the researcher to be right at α = 0.01?

c) Using the same sample proportion as in the original formulation of the exercise, what is
the required sample size in order to start rejecting H0 if α = 0.05?

2. A hotel manager thinks than the average daily expenditure is less than 250 euros per
tourist. In order to test this, a random sample of n = 101 tourists is selected, where the
sample average is 242 euros and sample variance is 23 euros2 . For α = 0.05:

a) Using the information above, can we say that the hotel manager is right?

b) What is the required sample mean in order to start rejecting the H0 ?

c) What is the required variance of the sample mean in order to start rejecting the H0 ?

3. Consider the survey on the average daily expenditure per tourist from the previous
exercise, whose sample size is n = 101, sample average expenditure is 242 euros, and sample
variance is 23 euros2 . Suppose that the hotel manager now believes that the variance of
tourist expenditure is less than 30 euros2 . Considering a significance level of α = 0.05:

a) Using the information above, can we say that the hotel manager is right?

b) What is the required sample variance in order to start rejecting the H0 ?

4. A bookstore chain hires a manager to estimate its mean weekly sales and its standard
deviation, where both are unknown. The manager believes that the average weekly sales
are less than 12,000e and that the standard deviation of the sales is greater than 3,000e.
The manager gathers information on a sample of 31 bookstores and observes an average
weekly sales of 10,500e with a standard deviation of 5,000e. Is the manager right about the
average weekly sales? Is the manager right about the standard deviation of the sales? Use
the significance level α = 5%.

1
 5. A multinational beer producer is interested in the possibility to switch from non-reusable
containers to the reusable ones. The change can happen only if at least 65% of their regular
clients are in favor. In order to test this, the producer performs a market study using a
random sample of 1000 people, out of which 570 are in favor of the change. Using this
information, should the producer decide to switch to the reusable containers? Use the
significance level α = 5%.

6. (Exam of February 2019) A restaurant owner believes that the standard deviation of
daily expenditure per customer is lower than 9 euros. He conducted a survey of size n =
28, and he obtained a daily average expenditure of 52.5 euros with a sample standard
deviation of 7.2 euros. Given a significance level of α = 0.01:

a) The owner of the restaurant wants to investigate whether the standard deviation of the
daily expenditure per customer is lower than 9 euros. Can we say that the standard
deviation of the daily expenditure per customer is lower than 9 euros?

b) From what value of the sample standard deviation can one start rejecting the H0 of part
(a)?

7. An agricultural company states that 7% of the harvest is not usable and has to be
thrown away. The organization of consumers thinks that the actual proportion is larger. In
order to test this, the organization selects a random sample of 700 products and finds that
72 of them are not usable and have to be thrown away. Can the statement made by the
agricultural company be rejected? Is there evidence that the organization of consumers is
right? Use the significance level α = 5%.

8. For all given situations state the null and the alternative hypothesis and explain how to
carry out the test:

a) The CEO of a certain chain of hotels thinks that an adult person has an average expen-
diture different from 10e per day at the bar of the hotel.

b) He also thinks that less than half of the adult guests go to the bar of the hotel.

9. (Exam of February 2015) A pharmaceutical company decides to start producing new

drug called XXX and the average weight of a pill has to be 5 grams. After the first day of
production a random sample of 12 pills is selected and after weighting each of them (xi ) the
following results are obtained:
12
X 12
X
xi = 59.94; (xi − x̄)2 = 0.0107.
i=1 i=1

Use the significance level α = 5% to answer to the questions below.

a) Can we say that the production process of working correctly?

2
b) One of the scientists in the laboratory calculates that in the production process the vari-
ance of the weights of the pills is 0.000625 grams2 . Having this information and consid-
ering the same random sample of 12 pills, does the conclusion concerning the hypothesis
in the first part changes?

c) Test if the population variance in this production process is 0.000625 (at 5% significance
level).

10. (Exam of July 2015) In order to study the average length of stay of tourists in a certain
area, a random sample of n = 200 tourists is selected. The average stay of the people in the
sample is 9.45 days.

Table 1

Rounded values of the test statistic p-value

-1 0.15866
-2 0.02275
-3 0.00135
-4 0.00003
-5 0.00000

a) Assume that the length of stay is normally distributed N (µ, 100). Test if the average
length of stay is equal to 10 days for α = 0.05. Justify your answer using the information
given in Table 1.

b) You know that 35% of the tourists in the sample stay more than 7 days. Test whether
60% of the tourists in the population stay more than 7 days. Justify your answer using
the information given in Table 1.

11. (Exam of February 2016) A certain machine produces on average 678 units per hour.
After repairing the machine, it was observed for 500 hours an average production of 674 units
per hour with a standard deviation of 40. Given this information, a manager thinks that the
repairs have reduced the productivity of the machine. State the null and the alternative
hypothesis and test which one is right in the following cases:

a) The type I error is 0.05.

b) Significance level is 0.01 and the population standard deviation is 45.

12. The Marbella City Council announces in an advertisement that the average cost of a
residential plot is 125,000e per lot. Nevertheless, a real estate agency doubts this information
and believes that the average cost of a residential lot is less than 125,000e per lot. The real
estate agency collects a sample of 32 lots and finds a sample mean of 130,000e per lot with
a sample standard deviation of 12,500e. Is the real estate agency right? Use α = 0.05 as
the significance level.

3
 13. A certain part in the car has to be produced with very high precision in order to be
accepted by the client. The technical specifications indicate that the maximum variance of
the lengths of the parts has to be 0.0004. Consider a random sample of 30 parts and the
sample variance of s2 = 0.0005. Use a 0.05 confidence level to test whether the technical
specifications are being violated.

14. (Exam of November 2020) According to data obtained from the National Statistics
Institute (INE), 15% of the families living with small children in the Balearic Islands are
single parents. Nevertheless, a journalist from Última Hora believes that this proportion
is higher. To test this statement, the journalist obtains a sample of n = 400 families and
finds that 70 families living with small children in the Balearic Islands are single parents.
a) Considering a significance level of α = 10%, can we say that the journalist is right?
b) Using the same sample proportion as in the original formulation of the exercise and a
significance level of α = 10%, what is the required sample size to say that the journalist
is right?

Solutions
1. a) No; b) p̄ < 0.0912; c) n ≥ 216.
2. a) Yes, he is right at the 5% significance level; b) x̄ < 249.21; c) σx̄2 < 23.23;
3. a) Yes, at the 5% significance level; b) S 2 ≤ 23.37;
4. The manager is not right about the sales mean, but he is right about the standard
deviation of the sales at the 5% significance level.
5. No, because we do not reject H0 at the 5% significance level.
6. a) No, because we do not reject H0 at the 5% significance level; b) S < 6.216.
7. Yes, we reject H0 at the 5% significance level, and there is evidence that the organization
of consumers is right.
8. a) H0 : µ = 10e versus HA : µ 6= 10e; b) H0 : p ≥ 50% versus HA : p < 50%, where
p is the population proportion of adult clients that consume both in the bar and in the
cafeteria.
9. a) Yes since we do not reject H0 at the 5% significance level; b) Yes since we do not reject
H0 at the 5% significance level; c) We cannot reject H0 at the 5% significance level.
10. a)We cannot reject H0 at the 5% significance level because the p-value = 2 Pr(z <
−0.778) > Pr(z < −1) = 0.15866; b) We reject H0 at the 5% significance level because
the p-value = 2 Pr(z < −7.217) < Pr(z < −5) ≈ 0.000;
11. a) The manager is right since we reject H0 at the 5% significance level; b) The manager
is not right since we do not reject H0 at the 1% significance level.

4
12. The real state agency is wrong because we cannot reject H0 : µ ≥ 125000 at the 5%
significance level.

13. We do not reject H0 : σ 2 ≤ 0.004 at the 5% significance level.

14. a) The journalist is right because we reject H0 at the 10% significance level; b) n ≥ 335.