STAT1302 – Statistical Analysis II

Assignment #2 – Solutions
Winter 2018

Q1. The management at New Century Bank claims that the mean waiting time for all customers at its branches is
less than that at the Public Bank, which is its main competitor. A business consulting firm took a sample of 200
customers from the New Century Bank and found that they waited an average of 4.5 minutes before being served.
Another sample of 300 customers taken from the Public Bank showed that these customers waited an average of
4.75 minutes before being served. Assume that the standard deviations for the two populations are 1.2 and 1.5
minutes, respectively.
a) Make a 97% confidence interval for the difference between the two population means.
• Let 𝜇1 and 𝜇 2 be the population means waiting time for all customers at New Century Bank and at the
Public Bank branches, respectively.
• From the given information, 𝑛1 = 200, 𝑥̅1 = 4.5, 𝜎1 = 1.2, 𝑛2 = 300, 𝑥̅ 2 = 4.75, 𝜎2 = 1.5
• The two samples are independent; The standard deviations 𝜎1 = 1.2 and 𝜎2 = 1.5 are known; Both sample
sizes are large; so a 97% confidence interval for 𝜇1 − 𝜇 2 is
𝜎12 𝜎22
(𝑥̅1 − 𝑥̅2 ) ± 𝐸, where 𝐸 = 𝑧𝛼/2 √ + .
𝑛1 𝑛2
• Point estimate of 𝜇1 − 𝜇 2 is 𝑥̅1 − 𝑥̅ 2 = 4.5 − 4.75 = −0.25 min
• 1 − 𝛼 = 0.97 → 𝛼 = 0.03 → 𝛼⁄2 = 0.015 → 𝑧0.015 = 2.17
𝜎12 𝜎22 1.22 1.52
• 𝐸 = 𝑧𝛼/2 √ + = 2.17 × √ + = 2.17 × 0.12124 = 0.26
𝑛1 𝑛2 200 300
• Therefore, a 97% confidence interval for 𝜇1 − 𝜇 2 is −0.25 ± 0.26 → (−0.51, 0.01)
• We are 97% confident that the difference between the two population means waiting time for all customers
at New Century Bank and the Public Bank branches is between −0.51 and 0.01 min.

b) Test at a 2.5% significance level whether the claim of the management of the New Century Bank is true. Use
the critical value approach.
1) Identify the model parameters of interest. The parameters of interest are 𝜇1 and 𝜇 2: 𝜇1 the population
means waiting time for all customers at New Century Bank and at the Public Bank branches, respectively.

𝐻 : 𝜇 − 𝜇2 = 0
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 . { 0 1
𝐻1 : 𝜇1 − 𝜇 2 < 0
3) Select the distribution to use.
The two samples are independent; The standard deviations 𝜎1 = 1.2 and 𝜎2 = 1.5 are known; Both sample
sizes are large; Therefore, we use the normal distribution to make the test.

4) Calculate the value of the test statistic and determine the rejection and non-rejection regions.
( 𝑥̅1 − 𝑥̅ 2 ) − (𝜇1 − 𝜇 2 ) −0.25 − 0
𝑧= = = −2.06
0.12124
𝜎12 𝜎22
√ +
𝑛1 𝑛2
Left-tailed test: We reject 𝐻0 if 𝑧 ≤ −𝑧𝛼 = −𝑧0.025 = −1.96

5) Make a decision. Under the critical value approach: 𝑧 = −2.06 < −1.96 → Reject 𝐻0

6) Conclusion: At a 2.5% level of significance, we have sufficient statistical evidence to conclude that the
mean waiting time for all customers at New Century Bank is less than that at the Public, and the bank’s
claim is true.
1
 c) Calculate the 𝑝-value for the test of part (b). Based on this 𝑝-value, would you reject the null hypothesis if 𝛼 =
0.01? What if 𝛼 = 0.05?
• Left-tailed test → 𝑃(𝑍 < −2.06) = 0.0197
• For 𝛼 = 0.01, do not reject 𝐻0 since 0.0197 > 0.01.
• For 𝛼 = 0.05, reject 𝐻0 since 0.0197 < 0.05.

Q2. A company sent seven of its employees to attend a course in building self-confidence. These employees were
evaluated for their self-confidence before and after attending this course. The following table gives the scores (on a
scale of 1 to 15, 1 being the lowest and 15 being the highest score) of these employees before and after they
attended the course. Construct a 95% confidence interval for the difference between the mean of the scores of these
employees before and the mean after they attended the course. State the underlying assumption(s).

Before 8 5 4 9 6 9 5

After 10 8 5 11 6 7 9

𝒅 −2 −3 −1 −2 0 2 −4

• 𝑡0.025 = 2.447 using 𝑑𝑓 = 7 − 1 = 6

1
• 𝑑̅ = −1.43, ∑𝑛𝑖=1 𝑑𝑖2 = 38, 𝑠𝑑 = √ (∑𝑛𝑖=1 𝑑𝑖2 − 𝑛𝑑̅2 ) = 1.9880
𝑛−1
𝑠𝑑
• A 95% CI for 𝜇 𝑑: 𝑑̅ ± 𝑡0.025 = −1.43 ± 1.84 ↔ (−3.27, 0.41)
√𝑛
• We are 95% confident that the mean paired difference, the score of an employee before attending the
course minus the score of the same employee after attending is between −3.27 and 0.41.

Q3. Quadro Corporation has two supermarket stores in a city. The company’s quality control department wanted to
check if the customers are equally satisfied with the service provided at these two stores. A sample of 380
customers selected from Supermarket I produced a mean satisfaction index of 7.6 (on a scale of 1 to 10, 1 being the
lowest and 10 being the highest) with a variance of 0.5625. Another sample of 370 customers selected from
Supermarket II produced a mean satisfaction index of 8.1 with a variance of 0.3481. Assume that the customer
satisfaction index for each supermarket has unknown but same population standard deviation.

a) Construct a 98% confidence interval for the difference between the mean satisfaction indexes for all customers
for the two supermarkets.
• Population 1: customers of Supermarket I; Population 2: customers of Supermarket II
𝛼
• 𝑛1 = 380, 𝑥̅1 = 7.6, 𝑠1 = 0.5625, 𝑛2 = 370, 𝑥̅ 2 = 8.1, 𝑠22 = 0.3481, 𝛼 = 0.02 ↔ = 0.01
2
2
• The two samples are independent; The standard deviations of the two populations are unknown, but
assumed to be equal; Both sample sizes are large; so, we use the 𝑡 distribution.
• The 98% confidence interval for 𝜇1 − 𝜇 2 is (𝑥̅1 − 𝑥̅ 2 ) ± 𝐸 where 𝐸 = 𝑡0.01 𝑠𝑥̅ 1−𝑥̅ 2
• 𝑑𝑓 = 𝑛1 + 𝑛2 − 2 = 380 + 370 − 2 = 748 → 𝑡0.01 ≈ 𝑧0.01 = 2.33
( 𝑛1−1) 𝑠21+( 𝑛2−1) 𝑠22 ( 380−1) ×0.5625+(370−1)×0.3481 1 1
• 𝑠𝑝 = √ =√ = 0.6758 → 𝑠𝑥̅ 1−𝑥̅ 2 = 𝑠𝑝 √ + = 0.0493
𝑛1+𝑛2−2 748 𝑛1 𝑛2
• 𝐸 = 𝑡0.01 𝑠𝑥̅ 1−𝑥̅ 2 = 2.33 × 0.0493 = 0.12
• The 98% confidence interval for 𝜇1 − 𝜇 2 is (𝑥̅1 − 𝑥̅2 ) ± 𝐸 ↔ −0.5 ± 0.12 ↔ (−0.62, − 0.38)
• We are 98% confident that the difference in the mean satisfaction indexes for all customers for the two
supermarkets lies between − 0.62 and −0.38.

2
b) Test at a 1% significance level whether the mean satisfaction indexes for all customers for the two
supermarkets are different. . Use either the 𝑝-value approach or the critical value approach.

1) Identify the model parameters of interest. 𝜇1 and 𝜇 2: the means satisfaction index for Supermarket I and
Supermarket II, respectively.
𝐻 : 𝜇 − 𝜇2 = 0
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 . { 0 1
𝐻1 : 𝜇1 − 𝜇 2 ≠ 0
3) Select the distribution to use. The two samples are independent; The standard deviations of the two
populations are unknown, but assumed to be equal; Both sample sizes are large; so, we use the 𝑡
distribution.

4) Calculate the value of the test statistic

(𝑥̅1 − 𝑥̅2 ) − (𝜇1 − 𝜇 2) −0.5 − 0
𝑡= = = −10.13
1 1 0.0493
𝑠𝑝 √ +
𝑛1 𝑛2
4a) Calculate the 𝑝-value: Two-tailed test → 𝑝-value = 2𝑃(𝑇 < −10.13) < 0.002 for 𝑑𝑓 = ∞ (𝑑𝑓 = 1148)
4b) Determine the Rejection Regions: Two-tailed test: Reject 𝐻0 if 𝑡 ≥ 𝑡0.005 = 2.576 or 𝑡 ≤ −2.576

5) Make a decision.
Under the 𝑝-value approach: 𝑝-value < 0.002 < 0.01 = 𝛼 → Reject 𝐻0
Under the critical value approach: 𝑡 = −10.13 < 2.576 → Reject 𝐻0

6) Conclusion:
At a 1% level of significance, we have sufficient statistical evidence to conclude that the mean satisfaction
index for Supermarket I is different from the mean for Supermarket II.

Q4. In a random sample of 800 men aged 25 to 35 years, 24% said they live with one or both parents. In another
sample of 850 women of the same age group, 18% said that they live with one or both parents. Construct a 95%
confidence interval for the difference between the proportions of all men and all women aged 25 to 35 years who
live with one or both parents.

• Let 𝑝1 and 𝑝1 be the proportions of all men and all women aged 25 to 35 years who live with one or both
parents, respectively.
• 𝑝̂1 = 0.24 and 𝑝̂ 2 = 0.18
• 𝑛1 = 800 ≥ 30, 𝑛2 = 850 ≥ 30, 𝑛1 𝑝1 = 800 × 0.24 = 192 > 5, 𝑛1 𝑞1 = 800 × 0.76 = 608 > 5,
𝑛2 𝑝2 = 850 × 0.18 = 153 > 5 and 𝑛2 𝑞2 = 850 × 0.82 = 697 > 5 → So, we use the normal distribution.
• The point estimate of 𝑝1 − 𝑝2 = 𝑝̂1 − 𝑝̂ 2 = 0.24 − 0.18 = 0.06
𝑝̂1𝑞̂ 1 𝑝̂2𝑞̂ 2
• Therefore, a 95% confidence interval 𝑝1 − 𝑝2 is (𝑝̂ 1 − 𝑝̂ 2 ) ± 𝑧𝛼/2 √ +
𝑛1 𝑛2
• 𝑧𝛼/2 = 𝑧0.025 = 1.96
𝑝̂1𝑞̂ 1 𝑝̂2𝑞̂ 2 0.24×0.76 0.18×0.82
• 𝑠𝑝̂1−𝑝̂2 = √ + =√ + = 0.02
𝑛1 𝑛2 800 850
𝑝̂1𝑞̂ 1 𝑝̂2𝑞̂ 2
• 𝐸 = 𝑧𝛼/2 √ + = 1.96 × 0.02 = 0.04
𝑛1 𝑛2
• The 95% confidence interval 𝑝1 − 𝑝2 : 0.06 ± 0.04 ↔ (0.02, 0.1)
• We are 95% confident that the difference between the two population proportions is between 0.02 and 0.1.

3
Q5. The owner of a mosquito-infested fishing camp in Alaska wants to test the effectiveness of two rival brands of
mosquito repellents, X and Y. During the first month of the season, eight people are chosen at random from those
guests who agree to take part in the experiment. For each of these guests, Brand X is randomly applied to one arm
and Brand Y is applied to the other arm. These guests fish for 4 hours, then the owner counts the number of bites on
each arm. The table below shows the number of bites on the arm with Brand X and those on the arm with Brand Y
for each guest.

a) Test at a 5% significance level whether the effectiveness of Brand X is different from the effectiveness of
Brand Y. State the underlying assumption(s). Use the critical value approach.

Guest A B C D E F G H
Brand X 12 23 18 36 8 27 22 32
Brand Y 9 20 21 27 6 18 15 25
𝒅 3 3 −3 9 2 9 7 7
37 1
• 𝑑̅ = = 4.63, ∑𝑛𝑖=1 𝑑𝑖2 = 291, 𝑠𝑑 = √ (∑𝑛𝑖=1 𝑑𝑖2 − 𝑛𝑑̅ 2 ) = 4.1382
8 𝑛−1

1) Identify the model parameters of interest.

Let 𝜇1 and 𝜇 2 be the means number of bites on the arm with Brand X and Brand Y, respectively. Then 𝜇 𝑑 =
𝜇1 − 𝜇 2 is the mean of the differences between the number of bites on the arm with Brand X and Brand Y.

𝐻0 : 𝜇 𝑑 = 0
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 . {
𝐻1 : 𝜇 𝑑 ≠ 0

3) Select the distribution to use.

i. The standard deviation 𝜎𝑑 of the population of paired differences is unknown.
ii. The sample size is small (i.e., 𝑛 = 8 < 30), and we need to assume that the population of paired
differences is approximately normally distributed .
so, we use the 𝑡 distribution to make the test about 𝜇 𝑑.

4) Calculate the value of the test statistic and determine the Rejection and Non-rejection Regions
𝑑̅−𝜇𝑑 4.63−0
The observed 𝑡 is: 𝑡 = 𝑠𝑑 = 4.1383 = 3.165
⁄ ⁄
√𝑛 √8
Two-tailed test: We reject 𝐻0 if 𝑡 ≥ 𝑡𝛼/2 = 𝑡0.025 = 2.365 or 𝑡 ≤ −𝑡𝛼/2 = −2.365

5) Make a decision.
Under the critical value approach: 𝑡 = 3.165 > 2.365 → Reject 𝐻0

6) Conclusion:
At a 5% level of significance, we have sufficient statistical evidence to conclude that the mean number of bites
on the arm with Brand X and the mean number of bites on the arm with Brand Y are different for all such
guests, so the effectiveness of Brand X is different from the effectiveness of Brand Y.

b) What would your decision in part (a) be if the probability of making a type I error were zero? Explain.
Type I error = 𝛼 = 0 → There is no rejection region → We fail to reject for 𝛼 = 0, so we cannot conclude that the
mean number of bites on the arm with Brand X and the mean number of bites on the arm with Brand Y are different
for all such guests.

4
Q6. A mail-order company has two warehouses, one on the East Coast and the second on the West Coast. The
company’s policy is to mail all orders placed with it within 72 hours. The company’s quality control department
checks quite often whether or not this policy is maintained at the two warehouses. A recently taken sample of 300
orders placed with the warehouse on the East Coast showed that 279 of them were mailed within 72 hours. Another
sample of 400 orders placed with the warehouse on the West Coast showed that 364 of them were mailed within
72 hours.
a) Using a 2.5% significance level, can you conclude that the proportion of all orders placed at the warehouse on
the West Coast that are mailed within 72 hours is lower than the corresponding proportion for the warehouse
on the East Coast? Use the critical value approach.
• Population 1: warehouse on the East Coast and Population 2: warehouse on the West Coast
279 364
• 𝑝̂1 = = 0.93 and 𝑝̂ 2 = = 0.91
300 400
𝑥 1+𝑥 2 279+364
• 𝑝̅ = = = 0.919 and 𝑞̅ = 1 − 𝑝̅ = 1 − 0.919 = 0.081
𝑛1 +𝑛2 300+400
1 1 1 1
• 𝑠𝑝̂1−𝑝̂2 = √𝑝̅ 𝑞̅ ( + ) = √0.919 × 0.081 ( + ) = 0.0208
𝑛1 𝑛2 300 400

1) Identify the model parameters of interest. 𝑝1 and 𝑝2: the proportion of all orders placed at the warehouse
on the East Coast and the West Coast that are mailed within 72 hours, respectively.
𝐻 : 𝑝 − 𝑝2 = 0 (or 𝑝1 = 𝑝2 )
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 . { 0 1
𝐻1 : 𝑝1 − 𝑝2 > 0 (or 𝑝1 > 𝑝2 )
3) Select the distribution to use.
𝑛1 = 300 ≥ 30, 𝑛2 = 400 ≥ 30, 𝑛1 𝑝1 = 300 × 0.93 = 279 > 5, 𝑛1 𝑞1 = 300 × 0.07 = 21 > 5, 𝑛2 𝑝2 =
400 × 0.91 = 364 > 5 and 𝑛2 𝑞2 = 400 × 0.09 = 36 > 5 → So, we use the normal distribution.

4) Calculate the value of the test statistic and determine the rejection and non-rejection regions
(𝑝̂ 1 − 𝑝̂ 2 ) − (𝑝1 − 𝑝2 ) (0.93 − 0.91) − 0
𝑧= = = 0.96
1 1 0.0208
√𝑝̅ 𝑞̅ ( + )
𝑛1 𝑛2
Right-tailed test: We reject 𝐻0 if 𝑧 ≥ 1.96

5) Make a decision. Under the critical value approach: 0.96 < 1.96 → Fail to reject 𝐻0

6) Conclusion. At a 2.5% level of significance, we do not have sufficient statistical evidence to conclude that
the proportion of all orders placed at the warehouse on the East Coast that are mailed within 72 hours is
higher than the corresponding proportion for the warehouse on the West Coast (or we do not have sufficient
statistical evidence to conclude that the proportion of all orders placed at the warehouse on the West Coast
that are mailed within 72 hours is lower than the proportion for the warehouse on the East Coast).

b) What is the type I error in (a)? Explain. What is the probability of making such an error
• The probability of making such an error occurs when a true null hypothesis is rejected. In this question, A
Type I error occurs when the proportions of all orders placed at the warehouse on the East Coast and on
the West Coast that are mailed within 72 hours are the same, but we conclude that the proportion of all
orders placed at the warehouse on the East Coast that are mailed within 72 hours is higher than the
corresponding proportion for the warehouse on the West Coast.
• The probability of making a Type I error is 𝛼 = 0.025.

c) Find the 𝑝-value for the test in part (a). 𝑝-value = 𝑃 (𝑍 > 0.96) = 0.1685
Q7.
5
 a) Determine the value of 𝜒 2 for 23 degrees of freedom and an area of 0.990 in the left tail of the chi-square
distribution curve. 41.638
b) Find the value of 𝜒 2 for 4 degrees of freedom and 0.005 area in the right tail of the chi-square distribution
curve. 14.860
c) Find the value of 𝜒 2 for 𝑑𝑓 = 6 and 0.05 area in the left tail of the chi-square distribution curve. 1.635

Q8. Chance Corporation produces beauty products. Two years ago the quality control department at the company
conducted a survey of users of one of the company’s products. The survey revealed that 53% of the users said the
product was excellent, 31% said it was satisfactory, 7% said it was unsatisfactory, and 9% had no opinion. Assume
that these percentages were true for the population of all users of this product at that time. After this survey was
conducted, the company redesigned this product. A recent survey of 800 users of the redesigned product conducted
by the quality control department at the company showed that 495 of the users think the product is excellent, 255
think it is satisfactory, 35 think it is unsatisfactory, and 15 have no opinion. Is the percentage distribution of the
opinions of users of the redesigned product different from the percentage distribution of users of this product before
it was redesigned? Use 𝛼 = 0.025.

1) Identify the parameters of interest: The proportions of users, 𝑝1 , 𝑝2 , 𝑝3 and 𝑝4 , who think the product is
excellent, satisfactory, unsatisfactory and who have no opinion, respectively.
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 .
𝐻0 : The percentage distribution of users' opinions is unchanged since the product was redesigned
{ (or 𝑝1 = 0.53, 𝑝2 = 0.31, 𝑝3 =0.07 and 𝑝4 =0.09)
𝐻1 : The percentage distribution of users' opinions has changed since the product was redesigned.

3) Select the distribution to use.

Since this is a multinomial experiment and 𝑛 = 800 is large (𝐸𝑖 > 5), use the chi-square distribution.

4) Calculate the value of the test statistic and determine the rejection and non-rejection regions

Opinion 𝑶 𝒑 𝑬 = 𝒏𝒑 𝑶𝟐 𝑶𝟐 /𝑬

Excellent 495 0.53 424 245025 577.8892

Satisfactory 255 0.31 248 65025 262.2

Unsatisfactory 35 0.07 56 1225 21.875

No opinion 15 0.09 72 225 3.125

Total 𝑛 = 800 1 𝑛 = 800 865.087

𝑂2
𝜒2 = ∑ − 𝑛 = 865.087 − 800 = 65.087
𝐸
𝑘 = 4, 𝑑𝑓 = 4 – 1 = 3 For 𝛼 = 0.025, the critical value of 𝜒 2 is 9.348. We reject 𝐻0 if 𝜒 2 > 9.348

5) Make a decision. 65.087 > 9.348 → Reject 𝐻0

6) Conclusion
At a 2.5% level of significance, we have sufficient statistical evidence to conclude that the percentage
distribution of users' opinions has changed since the product was redesigned.

6
Q9. Four hundred people were selected from each of the four geographic regions (Midwest, Northeast, South,
West) of the United States, and they were asked which form of camping they prefer. The choices were pop-up
camper/trailer, family style (tenting with sanitary facilities), rustic (tenting, no sanitary facilities), or none. The
results of the survey are shown in the following table. Based on the evidence from these samples, can you conclude
that the distributions of favorite forms of camping are different for at least two of the regions? Use 𝛼 = 0.01.

1) Identify the characteristic of interest: The choices camper/trailer, family style, rustic, or none for four
different geographic regions.
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 .
𝐻 : The distributions of favorite forms of camping are the same for all four regions.
{ 0
𝐻1 : The distributions of favorite forms of camping are different for at least two regions .

3) Select the distribution to use.

Since this is a test of homogeneity and the sample size is large, 𝑛 = 1600, use the chi-square distribution.
4) Calculate the value of the test statistic and determine the rejection and non-rejection regions

Midwest Northeast South West Total

Camper/trailer 132 129 129 135 525
(131.25) (131.25) (131.25) (131.25)
Family style 180 175 168 146 669
(167.25) (167.25) (167.25) (167.25)
Rustic 46 50 59 68 223
(55.75) (55.75) (55.75) (55.75)
None 42 46 44 51 183
(45.75) (45.75) (45.75) (45.75)
Total 400 400 400 400 1600
𝑂 2 132 2 1292 512
𝜒2 = ∑ − 𝑛 = + +⋯+ − 1600 = 10.379
𝐸 131.25 131.25 45.75
𝑑𝑓 = (𝑅 − 1)(𝐶 − 1) = 3 × 3 = 9 For 𝛼 = 0.01, the critical value of 𝜒 2 is 21.666. We reject 𝐻0 if 𝜒 2 >
21.666

5) Make a decision. 10.739 < 21.666 → Fail to reject 𝐻0

6) Conclusion
At a 1% level of significance, we do not have sufficient statistical evidence to conclude that the distributions
of favorite forms of camping are different for at least two of the regions.

Q10. Many students graduate from college deeply in debt from student loans, credit card debts, and so on. A
sociologist took a random sample of 401 single persons, classified them by gender, and asked, “Would you
consider marrying someone who was $25,000 or more in debt?” The results of this survey are shown in the
following table. Test at a 1% significance level whether gender and response are related.

1) Identify the characteristics of interests: There are two characteristics of interest: gender and responses
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 .
𝐻 : Gender and responses are independent.
{ 0
𝐻1 : Gender and responses are dependent.
3) Select the distribution to use.
Since this is a test of independence and 𝑛 = 401 is large, use the chi-square distribution.
7
 4) Calculate the value of the test statistic and determine the rejection and non-rejection regions

Yes No Uncertain Total

Women 125 59 21 205
(115.54) (70.55) (18.92)
Men 101 79 16 196
(110.46) (67.45) (18.08)
Total 226 138 37 401
𝑂 2 125 2 59 2 16 2
𝜒2 = ∑ − 𝑛 = + +⋯+ − 401 = 406.922 − 401 = 5.922
𝐸 115.54 70.55 18.08
𝑑𝑓 = (𝑅 − 1)(𝐶 − 1) = 1 × 2 = 2 For 𝛼 = 0.01, we reject 𝐻0 if 𝜒 2 > 9.210

5) Make a decision. 5.922 < 9.210 → Fail to reject 𝐻0

6) Conclusion. At a 1% level of significance, we do not have sufficient statistical evidence to conclude that
gender and response are related.

Q11. In 2014, the variance of the ages of all workers at a large company that has more than 30,000 workers was
133. A recent random sample of 25 workers selected from this company showed that the variance of their ages is
112.
a) Using a 2.5% significance level, can you conclude that the current variance of the ages of workers at this
company is lower than 133? Assume that the ages of all current workers at this company are (approximately)
normally distributed.
1) Identify the model parameters of interest. 𝜎 2 : the variance of the ages of all workers at a large company.

𝐻 : 𝜎 2 = 133
2) State the null and alternative hypotheses; 𝐻0 and 𝐻1 . { 0 2
𝐻1 : 𝜎 < 133
3) Select the distribution to use. Since this is a test about a population variance and the ages of all current
workers at this company are (approximately) normally distributed, use the chi-square dist with 𝑑𝑓 = 24.

4) Calculate the value of the test statistic and determine the rejection and non-rejection regions
(𝑛 − 1)𝑠 2 (25 − 1)112
𝜒2 = = = 20.211
𝜎2 133
We reject 𝐻0 if 0 < 𝜒 2 ≤ 12.401 𝑑𝑓 = 25 − 1 = 24

5) Make a decision. Since 20.211 > 12.401 → Fail to reject 𝐻0

6) Conclusion
At a 2.5% level of significance, we do not have sufficient statistical evidence to conclude that the current
variance of the ages of workers at this company is lower than 133.

b) Construct a 98% confidence intervals for the variance of the ages of all current workers at this company.
𝜒 2 for 24 𝑑𝑓 and 0.01 area in the right tail = 42.980
𝜒 2 for 24 𝑑𝑓 and 0.99 area in the right tail = 10.856
( 𝑛−1) 𝑠2 ( 𝑛−1) 𝑠2 ( 25−1) 112 ( 25−1) 112
The 98% confidence interval for 𝜎 2 is to → to →
𝜒2𝛼⁄2 𝜒21−𝛼⁄2 42.980 10.856
(62.5407,247.6050 )