Chapter 8

Example 8.1:
Solution:
(a) Use binomial distribution with 𝑛 = 10, 𝑝 = 0.6 𝑎𝑛𝑑 𝑞 = 0.4. If X is the number of orders in the sample that arrived
late,
3 3
10
𝛼 = 𝑃(𝑋 ≤ 3 𝑤ℎ𝑒𝑛 𝑝 = 0.6) = ∑ ( ) (0.6𝑥 )(0.410−𝑥 ) = ∑(10𝐶𝑥) (0.6𝑥 )(0.410−𝑥 ) = 0.0548
𝑥
𝑥=0 𝑥=0

We say that the null hypothesis, p=0.6, is being tested at the 𝛼 = 0.0548 level of significance.

(b) If p = 0.3, this test procedure gives

3

𝛽 = 𝑃(𝑋 > 3 𝑤ℎ𝑒𝑛 𝑝 = 0.3) = 1 − ∑(10𝐶𝑥) (0.3𝑥 )(0.710−𝑥 ) = 0.3504

𝑥=0

If p = 0.4,

𝛽 = 𝑃(𝑋 > 3 𝑤ℎ𝑒𝑛 𝑝 = 0.4) = 1 − ∑(10𝐶𝑥) (0.4𝑥 )(0.610−𝑥 ) = 0.6177

𝑥=0

If p = 0.5,

𝛽 = 𝑃(𝑋 > 3 𝑤ℎ𝑒𝑛 𝑝 = 0.5) = 1 − ∑(10𝐶𝑥) (0.5𝑥 )(0.510−𝑥 ) = 0.8281

𝑥=0
_____________________________________________________

Example 8.2:
Solution:
(a) Use the normal distribution with 𝑛 = 50, 𝑝 = 0.6 𝑎𝑛𝑑 𝑞 = 0.4. Thus,

24.5 − (50)(0.6)
𝛼 = 𝑃(𝑋 ≤ 24 𝑤ℎ𝑒𝑛 𝑝 = 0.6) = 𝑃 (𝑍 ≤ ) = 𝑃(𝑍 ≤ −1.59).
√50(0.6)(0.4)
Using Table A.3,
𝛼 = 0.0559.

(b) If p = 0.3, this test procedure gives

24.5 − (50)(0.3)
𝛽 = 𝑃(𝑋 > 24 𝑤ℎ𝑒𝑛 𝑝 = 0.3) = 𝑃 (𝑍 > ) = 𝑃(𝑍 > 2.93) = 1 − 𝑃(𝑍 < 2.93).
√50(0.3)(0.7)

𝛽 = 1 − 0.9983 = 0.0017.

If p = 0.4,

24.5 − (50)(0.4)
𝛽 = 𝑃(𝑋 > 24 𝑤ℎ𝑒𝑛 𝑝 = 0.4) = 𝑃 (𝑍 > ) = 𝑃(𝑍 > 1.3) = 1 − 𝑃(𝑍 < 1.3).
√50(0.4)(0.6)

𝛽 = 1 − 0.9032 = 0.0968.

If p = 0.5,

24.5 − (50)(0.5)
𝛽 = 𝑃(𝑋 > 24 𝑤ℎ𝑒𝑛 𝑝 = 0.5) = 𝑃 (𝑍 > ) = 𝑃(𝑍 > −0.14) = 1 − 𝑃(𝑍 < −0.14).
√50(0.5)(0.5)

𝛽 = 1 − 0.4443 = 0.5557.

REZEL A. STO. TOMAS, ECE 1

 _____________________________________________________

Example 8.3:
Solution:
(a) We may find the probability of committing a type I error when 𝜇 = 200 milliliters as

α = 𝑃(𝑋̅ < 191 when 𝜇 = 200) + 𝑃(𝑋̅ > 209 when 𝜇 = 200)

The z-values that correspond to the critical values 63.5 and 66.5 are

191 − 200 209 − 200

𝑧1 = = −1.8 and 𝑧2 = = 1.8
15 15
√9 √9

Therefore
α = 𝑃(𝑍 < −1.8) + 𝑃(𝑍 > 1.8) = 0.0359 + (1 − 0.9641) = 0.0718.

(b) We may find the probability of committing a type II error when 𝜇 = 215 milliliters as

𝛽 = 𝑃(191 < 𝑋̅ < 209 𝑤ℎ𝑒𝑛 𝜇 = 215)

The z-values that correspond to the critical values 63.5 and 66.5 are

191 − 215 209 − 215

𝑧1 = = −4.8 and 𝑧2 = = −1.2
15 15
√9 √9

Therefore
𝛽 = 𝑃(−4.8 < 𝑍 < −1.2) = 𝑃(𝑍 < −1.2) − 𝑃(𝑍 < −4.8) = 0.1151 − 0 = 0.1151.
_____________________________________________________

Example 8.4:
Solution:
(a) If the critical region is 𝑥̅ > 185, the type I error probability is

α = 𝑃(𝑋̅ > 185 when 𝜇 = 175)

The z-value that correspond to the critical value 185 is

185 − 175
𝑧= = 1.58
20
√10

Therefore
α = 𝑃(𝑍 > 1.58) = 1 − 𝑃(𝑍 < 1.58) = 1 − 0.9429 = 0.0571.

(b) We may the probability of type II error if the true mean foam height is 195 millimeters as

β = 𝑃(𝑋̅ ≤ 185 when 𝜇 = 195)

The z-value that correspond to the critical value 185 is

185 − 195
𝑧= = −1.58
20
√10

Therefore
β = 𝑃(𝑍 ≤ −1.58) = 0.0571.

REZEL A. STO. TOMAS, ECE 2

(c) We may find the critical value in which the boundary of the critical region be placed if the type I error probability were
to remain equal to the value that it took on when n = 10 as
α = 𝑃(𝑋̅ > 𝑐 when 𝜇 = 175) = 0.0571

The z-value that correspond to the critical value, c, is

𝑐 − 175
𝑧= = 1.58
20
√16
Solving for c yields 182.9 (𝑐 = 182.9).

(d) Using n = 16 and the new critical region found in part (c), we may find the type II error probability if the true mean
foam height is 195 millimeters as
β = 𝑃(𝑋̅ ≤ 182.9 when 𝜇 = 195)

The z-value that correspond to the critical value 185 is

182.9 − 195
𝑧= = −2.42
20
√16

Therefore
β = 𝑃(𝑍 ≤ −2.42) = 0.0078.

(d) For the same level of α, with the increased sample size, β is reduced.
_____________________________________________________

Example 8.5:
Solution:
(a)
1. 𝐻0 ∶ 𝜇 = 800 hours
𝐻1 ∶ 𝜇 ≠ 800 hours

2. 𝛼 = 0.05

3. The test statistic is

𝑥̅ − 𝜇0
𝑧0 =
𝜎 ⁄√𝑛

4. Reject 𝐻0 if 𝑧0 > 1.96 or 𝑧0 < −1.96, and fail to reject 𝐻0 if −1.96 ≤ 𝑧0 ≤ 1.96. Note that this results from step 4,
where we specified 𝛼 = 0.05, and so the boundaries of the critical region are at 𝑧0.025 = 1.96 and −𝑧0.025 = −1.96.

5. Computations: Since 𝑥̅ = 788 and 𝜎 = 40,

788 − 800
𝑧0 = = −1.64
40⁄√30

6. Conclusion: Since −1.96 ≤ 𝑧0 ≤ 1.96, we fail to reject 𝐻0 ∶ 𝜇 = 65 at the 0.05 level of significance. Stated more
completely, we conclude that the mean is not significantly different from 800.

(b)
The computed value of the test statistic is 𝑧0 = −1.64 and since the alternative hypothesis is two-tailed, the P-value is

𝑃 − value = 2[1 − 𝑃(𝑍 ≤ |−1.64|)] = 2[1 − 0.9495] = 0.101

Hence, the mean is not significantly different from 800 for 𝛼 < 0.101.
_____________________________________________________

Example 8.6:
Solution:

REZEL A. STO. TOMAS, ECE 3

(a)
1. 𝐻0 ∶ 𝜇 = 40 hours
𝐻1 ∶ 𝜇 > 40 hours

2. 𝛼 = 0.05

3. The test statistic is

𝑥̅ − 𝜇0
𝑧0 =
𝜎 ⁄√𝑛

4. Reject 𝐻0 if 𝑧0 > 1.645. Note that this results from step 4, where we specified 𝛼 = 0.05.

5. Computations: Since 𝑥̅ = 40.5 and 𝜎 = 1.25,

40.5 − 40
𝑧0 = = 1.26
1.25⁄√10

6. Conclusion: Since, 𝑧0 < 1.645, we do not reject 𝐻0 ∶ 𝜇 = 40 at the 0.05 level of significance. Stated more completely,
we conclude that there is no significant evidence to support the claim that battery life exceeds 40 hours at a = 0.05.

(b)
The computed value of the test statistic is 𝑧0 = 1.26 and since the alternative hypothesis one-tailed (upper − tailed test),
the P-value is
𝑃 − value = 1 − 𝑃(𝑍 ≤ 1.26) = [1 − 0.8962] = 0.1038
_____________________________________________________
Example 8.7:
Solution:
1. 𝐻0 ∶ 𝜇 = 10 liters
𝐻1 ∶ 𝜇 ≠ 10 liters

2. 𝛼 = 0.01

3. The t-statistic is
𝑥̅ − 𝜇0
𝑡=
𝑠⁄√𝑛

4. Critical regions: 𝑡0.005 > 3.25 or 𝑡0.005 < −3.25 with 9 degrees of freedom (see Table A.4)

5. Computations:
Since
𝑋1 + 𝑋2 +. . . . . . . . +𝑋1
𝑥̅ = = 10.06
10
and

𝑠 = 0.2459,

10.06 − 10
𝑡= = 0.7716
0.2459 ⁄√10

6. Conclusion: Since −3.25 < 𝑡 < 3.25, we fail to reject 𝐻0 ∶ 𝜇 = 10 at the 0.01 level of significance and conclude that
the average content of containers of a particular lubricant is not significantly different from 10 liters.
_____________________________________________________

REZEL A. STO. TOMAS, ECE 4

Example 8.9:
Solution:
1. 𝐻0 ∶ 𝜇 = 35 minutes
𝐻1 ∶ 𝜇 < 35 minutes

2. 𝛼 = 0.05

3. The t-statistic is
𝑥̅ − 𝜇0
𝑡=
𝑠⁄√𝑛

4. Critical region: 𝑡0.05 < −1.729 with 19 degrees of freedom (see Table A.4)

5. Computations:
Since
𝑥̅ = 33.1
and
𝑠 = 4.3,

33.1 − 35
𝑡= = −1.976
4.3⁄√20

6. Conclusion: Since 𝑡0.05 < −1.729 , we reject 𝐻0 ∶ 𝜇 = 35 at the 0.05 level of significance and conclude it takes less
than 35 minutes, on the average, to take the test.
_____________________________________________________

Example 8.10:
Solution:
(a)
1. 𝐻0 : μ1 − μ2 = 0 or μ1 = μ2
𝐻1 : μ1 − μ2 < 0 or μ1 < μ2

2. 𝛼 = 0.05

3. The test statistic is

4. Reject 𝐻0 if z < −zα where zα = -1.645

5. Computations: Since,

(89.6 − 92.5) − 0
𝑧= = −7.25
2 2
√1.2247 + 1.0954
15 20

6. Decision: Since 𝑧 < −1.645, reject 𝐻0 . We conclude that the mean road octane number for formulation 2 exceeds that
of formulation 1 using 𝛼 = 0.05.

(b) 𝑃 − value = (𝑍 ≤ −7.25) ≅ 0

_____________________________________________________

Example 8.11:
Solution:
1. 𝐻0 : μ1 − μ2 = 0 or μ1 = μ2
𝐻1 : μ1 − μ2 ≠ 0 or μ1 ≠ μ2

REZEL A. STO. TOMAS, ECE 5

2. 𝛼 = 0.05

3. The t-statistic is

where

4. Reject 𝐻0 if 𝑡 > 2.042 or 𝑡 < −2.042. The degrees of freedom, 𝑣 = 15 + 17 − 2 = 30

5. Computations: Since,

𝑠𝑝 2 = 0.3767
𝑠𝑝 = 0.6138

(8.73 − 8.68) − 0
𝑡= = 0.23
1 1
0.6138 × √ +
15 17

6. Decision: Since −2.042 < 𝑡 < 2.042, fail to reject 𝐻0 . We conclude that the two machines do not produce rods with
significantly different mean diameters at 𝛼 = 0.05.
_____________________________________________________

Example 8.12:
Solution:

1. 𝐻0 : μ1 − μ2 = 4 kilometers per liter

𝐻1 : μ1 − μ2 > 4 kilometers per liter

2. 𝛼 = 0.10

3. The t-statistic is

where

4. Reject 𝐻0 if 𝑡 > 1.325. The degrees of freedom, 𝑣 = 12 + 10 − 2 = 20.

5. Computations:

𝑥̅1 = 16 𝑠1 = 1 𝑛1 = 12
𝑥̅2 = 11 𝑠2 = 0.8 𝑛2 = 10

Hence
𝑠𝑝 2 = 0.838
𝑠𝑝 = 0.9154

Solving for t yields

(16 − 11) − 4
𝑡= = 2.5513
1 1
0.9154 × √ +
12 15

REZEL A. STO. TOMAS, ECE 6

6. Decision: Since 𝑡 > 1.325, reject 𝐻0 . We conclude that the fuel economy of Volkswagen mini-trucks, on average,
exceeds that of similarly equipped Toyota mini-trucks by 4 kilometers per liter.
_____________________________________________________

Example 8.13;
Solution:
1. 𝐻0 : μ2 − μ1 = 10 minutes
𝐻1 : μ2 − μ1 > 10 minutes

2. 𝛼 = 0.10

3. The statistic is

which has an approximate t-distribution with approximate degrees of freedom

4. Reject 𝐻0 if 𝑡 ′ > 𝑡𝛼 . 𝑡𝛼 = 1.415 with 7 degrees of freedom.

5. Computations:

𝑥̅1 = 97.4 𝑠1 2 = 78.8 𝑛1 = 5

𝑥̅2 = 110 𝑠2 2 = 913.333 𝑛2 = 7

Hence
𝑣 = 7.38 ≈ 7 (degrees of freedom)

Solving for t yields

(110 − 97.4) − 10
𝑡′ = = 0.22
√78.8 + 913.333
5 7

6. Decision: Since 𝑡 ′ < 1.415, fail to reject 𝐻0 .

_____________________________________________________

Example 8.14:
Solution:
Let μ1 and μ2 be the average gasoline consumption, in kilometers per liter, of the use of radial tires and of regular belted
tires, respectively. We proceed as follows:

1. 𝐻0 : μ1 − μ2 = 0
𝐻1 : μ1 − μ2 > 0

2. 𝛼 = 0.05

3. The statistic is

which has 𝑛 − 1 = 11 degrees of freedom.

REZEL A. STO. TOMAS, ECE 7

4. Reject 𝐻0 if 𝑡 > 𝑡𝛼/2 . 𝑡𝛼/2 = 2.201 with 11 degrees of freedom.

5. Computations:

Hence
0.1417 − 0
𝑡= = 2.48
0.198/(√12)

6. Decision: Since 𝑡 > 2.201, reject 𝐻0 .

_____________________________________________________

Example 8.15:
Solution:
(a)
1. 𝐻0 : 𝜇 = 5.5
𝐻1 : 𝜇 < 5.5

2. 𝛼 = 0.05

3. The test statistic is

𝑥̅ − 𝜇0
𝑧0 =
𝑠⁄√𝑛

4. Reject 𝐻0 if 𝑧0 < −1.645.

5. Computations: Since 𝑥̅ = 5.23 and 𝜎 = 𝑠 = 0.24,

5.23 − 5.5
𝑧0 = = −9
0.24/√64

6. Conclusion: Since 𝑧0 < −1.645, we reject 𝐻0 ∶ 𝜇 = 5.5 at the 0.05 level of significance. Thus, the white cheddar
popcorn, on average, weighs less than 5.5 ounces.

REZEL A. STO. TOMAS, ECE 8

(b) Since 𝛼 = 0.05 and β = 0.10, we have zα = 1.645 and zβ = 1.28. For the alternative β = 5.2, we take δ = 5.2 − 5.5 =
−0.3 and then

(1.645 + 1.28)2 (0.242 )

𝑛= = 5.4756 ≈ 6
(−0.3)2
_____________________________________________________

Example 8.16:
Solution:
Since 𝛼 = 0.02 and β = 0.05, we have zα/2 = 2.33 and zβ = 1.645. δ =3.1, thus

(2.33 + 1.645)2 (6.92 )

𝑛= = 78.28 ≈ 79
(3.1)2
_____________________________________________________

Example 8.17:
Solution:
(a)
1. 𝐻0 : μ𝐴 − μ𝐵 = 12 kilograms
𝐻1 : μ𝐴 − μ𝐵 > 12 kilograms

2. 𝛼 = 0.05

3. The test statistic is

4. Reject 𝐻0 if z >zα where zα = 1.645.

5. Computations: Since,

(86.7 − 77.8) − 12
𝑧= = −2.60.
2 2
√6.28 + 5.61
50 50

6. Decision: Since 𝑧 < 1.645, fail to reject 𝐻0 and conclude that the average tensile strength of thread A does not exceed
the average tensile strength of thread B by at least s12 kilograms.

(b) Since 𝛼 = 𝛽 = 0.05, we have zα = zβ =1.645 and δ = 8, thus

(1.645 + 1.645)2 (6.282 + 5.612 )

𝑛= = 11.99 ≈ 12
(8)2
_____________________________________________________

REZEL A. STO. TOMAS, ECE 9

Example 8.18:
Solution:
1. 𝐻0 : 𝑝 = 0.4
𝐻1 : 𝑝 > 0.4

2. 𝛼 = 0.05

2. Test statistic: Binomial variable X with p = 0.4 and n = 20.

5. Computations: 𝑥 = 9. Therefore, Therefore, from Table A.1, the computed P-value is

𝑃 = 𝑃(𝑋 ≥ 𝑥 𝑤ℎ𝑒𝑛 𝑝 = 𝑝0 )
8 8

𝑃 = 1 − ∑ 𝑏(𝑥; 20, 0.4) = 1 − ∑(20𝐶𝑥)(0.4𝑥 )(0.620−𝑥 ) = 0.4044

𝑥=0 𝑥=0

6. Decision: Since P-value is greater than 𝛼 = 0.05, we fail to reject 𝐻0 .

_____________________________________________________

Example 8.19:
Solution:
1
1. 𝐻0 : 𝑝 = = 0.2
5
𝐻1 : 𝑝 < 0.2

2. Test statistic: Binomial variable X with p = 0.2 and n = 1000.

5. Computations: 𝑥 = 136. Therefore, the computed P-value is

𝑥 − 𝑛𝑝0
𝑃 ≈ 𝑃 (𝑍 < )
√𝑛𝑝0 𝑞0

136 − 1000(0.2)
𝑃 ≈ 𝑃 (𝑍 < )
√1000(0.20)(0.8)

𝑃 ≈ 𝑃(𝑍 < −5.06) ≈ 0.

6. Decision: Reject 𝐻0 and conclude that fewer than one-fifth of the homes are heated by oil.
_____________________________________________________

Example 8.20:
Solution:
1. 𝐻0 : 𝑝 = 0.80
𝐻1 : 𝑝 > 0.80

2. α = 0.04.

3. Test statistic:
𝑥 − 𝑛𝑝0
𝑧=
√𝑛𝑝0 𝑞0

4. Reject 𝐻0 if 𝑧 > zα where zα= 1.75.

5. Computations: 𝑥 = 250 and 𝑛𝑝0 = (300) (0.8) = 240. Thus

250 − 240
𝑧= = 1.44
√300(0.8)(0.2)

6. Decision: Fail to reject 𝐻0 .

REZEL A. STO. TOMAS, ECE 10

 _____________________________________________________
Example 8.21:
Solution: Let 𝑝1 and 𝑝2 be the true proportions of adult women in the urban community and rural community, respectively,
having breast cancer.

1. 𝐻0 : 𝑝1 = 𝑝2
𝐻1 : 𝑝1 > 𝑝2

2. α = 0.05.

4. Critical region: z > 1.645.

5. Computations:
𝑥1 20 𝑥2 10
𝑝̂1 = = = 0.1, 𝑝̂2 = = = 0.067,
𝑛1 200 𝑛2 150
and
𝑥1 + 𝑥2 20 + 10
𝑝̂ = = = 0.086.
𝑛1 + 𝑛2 200 + 150

Therefore,

0.1 − 0.067
𝑧= = 1.1
√(0.086)(1 − 0.086) ( 1 + 1 )
200 150

𝑃 = 𝑃(𝑍 > 1.1) = 1 − 𝑃(𝑍 < 1.1) = 1 − 0.8643 = 0.1357

6. Decision: Since z < 1.645 or P-value is greater than α = 0.05, fail to reject 𝐻0 and conclude that it can not be shown that
breast cancer is more prevalent in the urban community.

Example 8.22:
Solution: Let 𝑝1 and 𝑝2 be the true proportions of residents in a certain city and its suburbs, respectively, favoring the
construction of a nuclear power plant.
1. 𝐻0 : 𝑝1 = 𝑝2
𝐻1 : 𝑝1 ≠ 𝑝2

2. Computations:
𝑥1 63 𝑥2 59
𝑝̂1 = = = 0.63, 𝑝̂2 = = = 0.472,
𝑛1 100 𝑛2 125
and
𝑥1 + 𝑥2 63 + 59
𝑝̂ = = = 0.542.
𝑛1 + 𝑛2 100 + 125

Therefore,

0.63 − 0.472
𝑧= = 2.36
√(0.542)(1 − 0.542) ( 1 + 1 )
100 125

𝑃 = 2𝑃(𝑍 > 2.36) = 2( 1 − 𝑃(𝑍 < 2.36)) = 2(1 − 0.9909) = 0.0182

6. Conclusion: Yes, there a significant difference between the proportions of urban and suburban residents who favor
construction of the nuclear plant.

REZEL A. STO. TOMAS, ECE 11