Martin Luther University Halle-Wittenberg Statistics II

Chair of Econometrics

Exercise Set 9 Solutions

Winter Semester 2023/24

1. A filling machine for wine bottles is constructed in such a way, that the fill quantity x is
normally distributed with the standard deviation σ = 3 cm3 . It shall be verified within a
sample, that the machine fills in 1000 cm3 on average. The filled quantities are assumed to
be independently distributed.

a) How shall the customer formulate the hypothesis to prove that less than 1000 cm3 are
filled in on average?

H0 : µ ≥ 1000 cm3
H1 : µ < 1000 cm3

b) How shall the company formulate the hypothesis if it wants to prove that not less than
1000 cm3 are normally filled in?

H0 : µ ≤ 1000 cm3
H1 : µ > 1000 cm3

c) A customer measured the quantity in nine bottles:

997 1000 1003 996 998 1001 995 1002 990
Perform a test on the basis of the hypotheses in a). Give the rejection rule, empirical
value of the test statistic and your decision.

• mean: x̄ = 91 ∑ xi = 91 (997 + 1000 + 1003 + 996 + 998 + 1001 + 995 + 1002 + 990) = 998
• Define the hypotheses: H0 : µ ≥ 1000 cm3 H1 : µ < 1000 cm3
• Select test statistic:
x̄ − µ0
Z= ∼ N(0, 1)
√σ
n

• Rejection rule: reject H0 if Z < −zα = −z0.05 = −1.64

• Calculate the empirical value of test statistic:
x̄ − µ0 998 − 1000 2
Z= = = − = −2
√σ √3 1
n 9

• Decision: Because the test statistic is lower than the critical value,
the consumer can reject the null hypothesis that the producer fills in
more than 1000 cm3 per bottle in favor of the alternative hypothesis at
the 5% significance level.
(Note: The question did not specify the level of α, therefore the
researcher has the freedom to set this level. In this above solution α
was set equal to 0.05)

1
2. The local transport operator advertises its timeliness. It claims, at minimum 95% of the
trains run without delay.

a) Formulate the hypothesis supporting the local transport operator’s claim.

H0 : p ≤ 0.95
H1 : p > 0.95

b) The operator has observed 16 delays out of 400 tours. Carry out a test corresponding
to the hypotheses in a). Use the significance level α = 0.10.

• p̂(on time) = 400−16

400 = 0.96
• Define the hypotheses: H0 : p ≤ 0.95 vs. H1 : p > 0.95
• Select the test statistic
p̂ − p0
Z=q ∼ N(0, 1)
1
n p0 (1 − p0 )

• Rejection rule: reject H0 if Z > zα = z0.10 = 1.28

• Calculate the empirical value of test statistic:
p̂ − p0 0.96 − 0.95
Z=q =q = 0.9177
1 0.95·(1−0.95)
n p0 (1 − p0 ) 400

• Decision: Because the test statistic is smaller than the critical value,
one cannot reject the null hypothesis that the local transport operator
is on time in less than 95% at 10% significance level.

3. It is claimed that an average child spends 15 hours per week watching television. A survey of
50 children finds an average of 12.75 hours per week, with a standard deviation of 8 hours.
Is the claim justified (α = 0.05)? Give the hypotheses, degrees of freedom, rejection rule,
empirical value of the test statistic, and your decision.

• Define the hypotheses: H0 : µ = µ0 = 15 vs. H1 : µ ̸= µ0 = 15

• Select the test statistic:
x̄ − µ0
T= ∼ tn−1
√s
n

• degrees of freedom: n − 1 = 50 − 1 = 49
(Since the statistical table does not report the critical values for degrees
of freedom 49, I have used 40)
• Rejection rule: reject H0 if |T | > tn−1; α2 = t40;0.025 = 2.021
• Calculate the empirical value of test statistic:
x̄ − µ0 12.75 − 15 2.25
t= = =− = −1.9887
√s √8 0.8
n 50

• Decision: Because the test statistic in absolute values is smaller than

the critical value, we cannot reject the null hypothesis that the the mean

2
 television time is equal to 15 hours per week at 5% significance level (or
95% confidence level).

4. After years of production, a manufacturer of batteries for automobiles finds that on average,
their batteries last 42.3 months with a standard deviation of σ = 4. Now, a new manufacturing
process is being introduced. In a sample of ten batteries produced by the new method the
average life is found to be 43.4 months. Test whether the batteries produced by the new
method have a longer life on average. State the hypotheses, rejection rule, empirical value
of the test statistic, and your decision (α = 0.05).

• Define hypotheses: H0 : µ ≤ 42.3 vs. H1 : µ > 42.3

• Select the test statistic:
x̄ − µ0
Z= ∼ N(0, 1)
√σ
n

• Rejection rule: reject H0 if Z > zα = z0.05 = 1.64

• (Step 4) Calculate the empirical value of test statistic:
x̄ − µ0 43.4 − 42.3 1.1
Z= = = = 0.8696
√σ √4 1.2649
n 10

• Decision: Because the test statistic is smaller than the critical value, we
cannot reject the null hypothesis at the 5% significance level. We do not
have sufficient empirical evidence to claim that batteries produced by the
new method have a higher life on average.

5. A review session is administered to a sample of students (n = 15) in an attempt to deter-

mine if this has an effect on exam performance. Based on information gathered in previous
semesters, the researcher knows that the population mean for a given exam is 65 points (out
of 100 points). The sample mean is 69, with a standard deviation of 10.
Test whether review sessions positively affect the exam performance. Formulate the hy-
potheses, give the degrees of freedom, rejection rule, empirical value of the test statistic, and
your conclusion (α = 0.05).

• Define the hypotheses: H0 : µ ≤ 65 vs. H1 : µ > 65

• Select the test statistic:
x̄ − µ0
T= ∼ tn−1
√s
n

• degrees of freedom: n − 1 = 15 − 1 = 14
• Rejection rule: reject H0 if T > tn−1;α = t14;0.05 = 1.761
• Calculate the empirical value of test statistic:
x̄ − µ0 69 − 65 4
T= = = = 1.5492
√s √10 2.5820
n 15

3
• Decision: Because the test statistic is smaller than the critical value,
we cannot reject the null hypothesis at the 5% significance level. The
review sessions do not seem to have a significant positive effect on exam
performance.