CLARIFYING THE CONCEPTS
1. What is the difference between and ? (p. 543)
9.5.1
is the sample proportion andis population proportion.
2. What are the conditions for the test for ? (p. 545)
3. Explain the essential idea about hypothesis testing for the proportion. (p. 543)
9.5.3
Answers will vary.
4. Explain what refers to. (p. 543)
5. What possible values can take? (p. 543)
9.5.5
Between 0 and 1 inclusive:
6. What is the difference between and a -value? (p. 546)
553
PRACTICING THE TECHNIQUES
CHECK IT OUT!
To do | Check out | Topic |
---|---|---|
Exercises 7–14 | Example 25 | Calculating for the test for proportion |
Exercises 15–18 | Example 26 |
test for using the critical-value method |
Exercises 19–22 | Example 27 |
test for using the -value method |
Exercises 23–26 | Example 29 | Using a confidence interval for to perform two-tailed hypothesis tests about |
Exercises 27–30 | Example 30 | Interpreting software output |
For Exercises 7–9, find the value of the test statistic for a right-tailed test with .
7. A sample of size 50 yields 30 successes.
9.5.7
2.8868
8. A sample of size 50 yields 40 successes.
9. A sample of size 50 yields 45 successes.
9.5.9
7.2169
10. What kind of pattern do we observe in the value of for a right-tailed test as the number of successes becomes more extreme?
For Exercises 11–13, find the value of the test statistic for a two-tailed test with .
11. A sample of size 80 yields 20 successes.
9.5.11
–4.47
12. A sample of size 80 yields 30 successes.
13. A sample of size 80 yields 40 successes.
9.5.13
0
14. What kind of pattern do we observe in the value of as the sample proportion approaches ?
For Exercises 15–18, do the following:
15. Test whether the population proportion is less than 0.5. A random sample of size 225 yields 100 successes. Let level of significance .
9.5.15
(a) We have and , so we can use thetest for proportions. (b) (c) . Reject if . (d) −1.67 (e) Since , we reject . There is evidence that the population proportion is less than 0.5.
16. Test whether the population proportion differs from 0.3. A random sample of size 100 yields 25 successes. Let level of significance .
17. Test whether the population proportion exceeds 0.6. A random sample of size 400 yields 260 successes. Let level of significance .
9.5.17
(a) We have and , so we can use thetest for proportions. (b) (c) . Reject if . (d) 2.04 (e) Since , we reject . There is evidence that the population proportion is greater than 0.6.
18. Test whether differs from 0.4. A random sample of size 900 yields 400 successes. Let level of significance .
For Exercises 19–22, do the following:
19. Test whether the population proportion exceeds 0.4. A random sample of size 100 yields 44 successes.
9.5.19
(a) We have and , so we can use the test for proportions. (b) . Reject if the (c) 0.82 (d) (e) Since the is not , we do not reject . There is insufficient evidence that the population proportion is greater than 0.4.
20. Test whether the population proportion is less than 0.2. A random sample of size 400 yields 75 successes.
21. Test whether the population proportion differs from 0.5. A random sample of size 900 yields 475 successes.
9.5.21
(a) We have and , so we may use the test for proportions. (b) Reject if the . (c) 1.67 (d) (e) Since the is not , we do not reject . There is insufficient evidence that the population proportion is not equal to 0.5.
22. Test whether the population proportion exceeds 0.9. A random sample of size 1000 yields 925 successes.
For Exercises 23–26, a confidence interval for is given. Use the confidence interval to test, using level of significance , whether differs from each of the indicated hypothesized values.
23. A 95% confidence interval for is (0.1, 0.9). Hypothesized values are
9.5.23
Value of | Form of hypothesis test, with | Where lies in relation to 95% confidence interval (0.1, 0.9) | Conclusion of hypothesis test | |
---|---|---|---|---|
(a) | 0 | Outside | Reject | |
(b) | 1 | Outside | Reject | |
(c) | 0.5 | Inside | Do not reject |
24. A 99% confidence interval for is (0.51, 0.52). Hypothesized values are
25. A 90% confidence interval for is (0.1, 0.2). Hypothesized values are
9.5.25
Value of | Form of hypothesis test, with | Where lies in relation to 90% confidence interval (0.1, 0.2) | Conclusion of hypothesis test | |
---|---|---|---|---|
(a) | 0.09 | Inside | Reject | |
(b) | 0.9 | Outside | Reject | |
(c) | 0.19 | Inside | Do not reject |
26. A 95% confidence interval for is (0.05, 0.95). Hypothesized values are
For Exercises 27–30, software output from a test for is provided. For each, examine the indicated software output, and provide the following steps:
Use level of significance for each hypothesis test.
27. TI-83/84 output
9.5.27
Step 1
We reject if the .
Step 2
Step 3
Step 4 The is , so we reject . There is evidence at level of significance that the population proportion is less than 0.5.
554
28. TI-83/84 output
29. Minitab output
9.5.29
Step 1
We reject if the .
Step 2
Step 3
Step 4 The is not , so we do not reject . There is insufficient evidence at level of significance that the population proportion is not equal to 0.05.
30. Minitab output
APPLYING THE CONCEPTS
31. Facebook Not Cool Anymore? In 2014, Facebook
reported that 23% of its users were ages 18–24, a decrease from the 2011 level of 30.9%. (Though this is partly due to more older people joining Facebook, there was nevertheless a loss of over 3 million users in this age group.) A survey of 500 randomly selected Facebook users showed that 100 of them were ages 18–24. If appropriate, test, using level of significance , whether the population proportion of Facebook users who are ages 18–24 has decreased from 23%.
9.5.31
Both and , so the conditions for using the test for proportions are met.
Step 1
We reject if the .
Step 2
Step 3
Step 4 The is , so we reject . There is evidence at level of significance that the population proportion of Facebook users ages 18–24 has decreased from 23%.
32. Twenty-Somethings. According to the U.S. Census Bureau, 7.1% of Americans living in 2004 were between the ages of 20 and 24. Suppose that a random sample of 400 Americans taken this year yields 35 between the ages of 20 and 24. If appropriate, test whether the population proportion of Americans ages 20–24 is different from 7.1%. Use level of significance .
33. Nonmedical Pain Reliever use. The National Survey on Drug Use and Health reported that 4.8% of persons ages 12 or older used a prescription pain reliever nonmedically.14 Suppose that a random sample of 900 persons ages 12 or older found 54 who had used a prescription pain reliever nonmedically. If appropriate, test whether the population proportion has increased, using level of significance .
9.5.33
Both and , so the conditions for using the test for proportions are met.
Step 1
We reject if the .
Step 2
Step 3
Step 4 The is not , so we do not reject . There is insufficient evidence at level of significance that the population proportion of persons ages 12 or older who used a prescription pain reliever nonmedically has increased from 4.8%.
34. Is This a Date, or What? In 2014, USA Today reported that 57% of 18- to 24-year-olds agreed that texting has made it more difficult to determine whether an outing is an actual date or not. Suppose that a recent random sample of 1000 18- to 24-year-olds found 500 who agreed with that sentiment. Test whether the population proportion of 18- to 24-year-olds who agree that texting has made it more difficult to determine whether an outing is an actual date has decreased, using level of significance .
35. Mutual Fund Performance. Business Insider reported that 58% of mutual funds underperformed in 2013, against the Standard and Poor 500 benchmark. Suppose a random sample of 100 mutual funds showed 67 that had underperformed. Test whether the population proportion has increased, using level of significance .
9.5.35
Both and , so the conditions for using the test for proportions are met.
Step 1
We reject if the .
Step 2
Step 3
Step 4 The is , so we reject . There is evidence at level of significance that the population proportion of mutual funds that underperform has increased from 58%.
36. Affective Disorders Among Women. What do you think is the most common nonobstetric (not related to pregnancy) reason for hospitalization among 18- to 44-year-old American women? According to the U.S. Agency for Healthcare Research and Quality (www.ahrq.gov), this is the category of affective disorders, such as depression. In 2002, of hospitalizations among 18- to 44-year-old American women, 7% were for affective disorders. Suppose that a random sample taken this year of 1000 hospitalizations of 18- to 44-year-old women showed 80 admitted for affective disorders. We are interested in whether the population proportion of hospitalizations for affective disorders has changed since 2002. Test, using level of significance .
37. Latino Household Income. The U.S. Census Bureau reported that 15.3% of Latino families had household incomes of at least $75,000. We are interested in whether the population proportion has changed, using the critical-value method and level of significance . Suppose that a random sample of 100 Latino families reported 23 with household incomes of at least $75,000.
9.5.37
(a) Yes. We have and . (b) Reject if . . . Since the is not , we do not reject . There is insufficient evidence that the population proportion of Hispanic families that had a household income of at least $75,000 is not equal to 0.153.
38. Massively Online Dropouts? TechCrunch.com reported in 2014 that the completion rate for massively open online courses (MOOCs) is only 6.5%. Suppose a new study of 1000 randomly chosen students taking a MOOC found that 75 completed the course. Test, using level of significance , whether the population proportion of students completing MOOCs has increased.
39. Living with the Parents. The National Association of Home Builders reported that 57% of young people ages 18–24 were living with their parents. Suppose a sample of 100 young people ages 18–24 showed 60 who were living with their parents. Test, using level of significance , whether the population proportion has changed.
9.5.39
Both and , so the conditions for using the test for proportions are met.
Step 1 We reject if the .
Step 2
Step 3
Step 4 The is not , so we do not reject . There is insufficient evidence at level of significance that the population proportion of young people ages 18–24 who are living with their parents differs from 57%.
40. Living with the Parents. Refer to Exercise 39.
BRINGING IT ALL TOGETHER
Children and Environmental Tobacco Smoke at Home. Use the following information for Exercises 41–44. The Environmental Protection Agency reported that 11% of children age 6 and under were exposed to environmental tobacco smoke (ETS) at home on a regular basis (at least four times per week).15 A random sample of 100 children age 6 and under showed that 6% of these children had been exposed to ETS at home on a regular basis.
555
41. Answer the following:
9.5.41
(a) We have and , so we can use thetest for proportions. (b) . Reject if . . Since is not , we do not reject . There is insufficient evidence that the population proportion of children age 6 and under exposed to ETS at home on a regular basis is less than 0.11.
42. Refer to Exercise 41.
43. Refer to your work in Exercise 41.
9.5.43
(a) Since the , we reject . There is evidence that the population proportion of children age 6 and under exposed to environmental tobacco smoke at home on a regular basis is less than 0.11. (b) The difference is because we changed the value of and not because we used different methods for the two different hypothesis tests. (c) Since , there is mild evidence against the null hypothesis that the population proportion of children age 6 and under exposed to environmental tobacco smoke at home on a regular basis is greater than or equal to 0.11.
44. Refer to Exercise 41. What if the sample proportion decreased, but everything else stayed the same? Describe what would happen to the following, and why.
Car Accidents Among Young Drivers. Use the following information for Exercises 45–47. The National Transportation Safety Board publishes statistics on the number of automobile crashes that people in various age groups have. Young people ages 18–24 have an accident rate of 12%, meaning that on average 12 out of every 100 young drivers per year had an accident. A researcher claims that the population proportion of young drivers having accidents is greater than 12%. Her study examined 1000 young drivers ages 19–24 and found that 134 had an accident this year.
45. Perform the appropriate hypothesis test with level of significance .
9.5.45
Both and , so the conditions for using the test for proportions are met.
Step 1 We reject if the .
Step 2
Step 3
Step 4 The is not , so we do not reject . There is insufficient evidence at level of significance that the population proportion of young people ages 18–24 who have had an accident has increased from 12%.
46. What if our sample size and the number of successes are doubled, so that remains the same? Otherwise, everything else is the same. Describe how this change would affect the following.
47. What if the hypothesized proportion was no longer 0.12? Instead, takes some value between 0.12 and 0.134. Otherwise, everything else is the same as in the original example. Describe how this change would affect the following.
9.5.47
(a) Increases. (b) Decreases (c) Increases (d) Stays the same (e) Stays the same
WORKING WITH LARGE DATA SETS
Fast Food versus Full Service Restaurants. Open the data set, Restaurants. Here, we will look at the fast food restaurants, but in a different way. We will explore the proportion of counties that have no fast food restaurants at all. Use technology to do the following:
restaurants
48. Obtain a random sample of size 100 from the data set.
restaurants
49. Find the sample proportion of counties with zero fast food restaurants.
restaurants
50. Using your sample, test whether the population proportion of counties with no fast food restaurants is less than 0.10, using level of significance .
restaurants
51. Find the actual value of the population proportion of counties with no fast food restaurants. Did your hypothesis test in Exercise 50 make the right decision? Explain.
WORKING WITH LARGE DATA SETS
Chapter 9 Case Study: Clothing Store Sales. Open the Chapter 9 Case Study data set, Clothing Store. Here, we will compare the proportions of buyers of blouses and buyers of suits. Use technology to do the following:
clothingstore
52. Obtain a random sample of size 100 from the data set.
clothingstore
53. Find the sample proportion of customers who have bought a blouse (value = 1 for variable = blouse).
clothingstore
54. Find the sample proportion of customers who have bought a suit (value = 1 for variable = suit).
clothingstore
55. Using your sample, test whether the population proportion of customers who have bought a blouse exceeds 0.5, using level of significance .
clothingstore
56. Explain why we need not perform a hypothesis test to determine whether the population proportion of customers who have bought a suit exceeds 0.5.
clothingstore
57. Using your sample, test whether the population proportion of customers who have bought a suit exceeds 0.1, using level of significance .