Chapter 7 Introduction to Sampling Distributions Section 7.1 1. Answers vary. Students should identify the individuals (subjects) and variable involved. Answers may include: A population is a set of measurements or counts either existing or conceptual. For example, the population of all ages of all people in Colorado; the population of weights of all students in your school; the population count of all antelope in Wyoming. 2. See Section 1.2. Answer may include: A simple random sample of n measurements from a population is a subset of the population selected in a manner such that (a) every sample of size n from the population has an equal chance of being selected and (b) every member of the population has an equal chance of being included in the sample. 3. A population parameter is a numerical descriptive measure of a population, such as µ, the population mean; σ, the population standard deviation; σ2, the population variance; p, the population proportion; ρ (rho) the population correlation coefficient for those who have already studied linear regression from Chapter 10. 4. A sample statistic is a numerical descriptive measure of a sample such as x , the sample mean; s, the sample standard deviation; s2, the sample variance; pˆ , the sample proportion; r, the sample correlation coefficient for those who have already studied linear regression from Chapter 10. 5. A statistical inference is a conclusion about the value of a population parameter based on information about the corresponding sample statistic and probability. We will do both estimation and testing. 6. A sampling distribution is a probability distribution for a sample statistic. 7. They help us visualize the sampling distribution by using tables and graphs that approximately represent the sampling distribution. 8. Relative frequencies can be thought of as a measure or estimate of the likelihood of a certain statistic falling within the class bounds. 9. We studied the sampling distribution of mean trout lengths based on samples of size 5. Other such sampling distributions abound. Notice that the sample size remains the same for each sample in a sampling distribution. Copyright © Houghton Mifflin Company. All rights reserved. 421 422 Instructor’s Resource Guide Understandable Statistics, 8th Edition Section 7.2 Note: Answers may vary slightly depending on the number of digits carried in the standard deviation. 1. (a) µ x = µ = 15 σx = σ n = 14 49 = 2.0 Because n = 49 ≥ 30, by the central limit theorem, we can assume that the distribution of x is approximately normal. z= x −µ σx = x − 15 2.0 15 − 15 =0 2.0 17 − 15 x = 17 converts to z = =1 2.0 x = 15 converts to z = P (15 ≤ x ≤ 17 ) = P ( 0 ≤ z ≤ 1) = P ( z ≤ 1) − P ( z ≤ 0 ) = 0.8413 − 0.5000 = 0.3413 (b) µ x = µ = 15 σx = σ n = 14 64 = 1.75 Because n = 64 ≥ 30, by the central limit theorem, we can assume that the distribution of x is approximately normal. z= x −µ σx = x − 15 1.75 15 − 15 =0 1.75 17 − 15 x = 17 converts to z = = 1.14 1.75 x = 15 converts to z = P (15 ≤ x ≤ 17 ) = P ( 0 ≤ z ≤ 1.14 ) = P ( z ≤ 1.14 ) − P ( z ≤ 0 ) = 0.8729 − 0.5000 = 0.3729 (c) The standard deviation of part (b) is smaller because of the larger sample size. Therefore, the distribution about µ x is narrower in part (b). 2. (a) µ x = µ = 100 σ 48 σx = = = 5.33 n 81 Because n = 81 ≥ 30, by the central limit theorem, we can assume that the distribution of x is approximately normal. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 z= x −µ σx = 423 x − 100 5.33 92 − 100 = −1.50 5.33 100 − 100 x = 100 converts to z = =0 5.33 x = 92 converts to z = P ( 92 ≤ x ≤ 100 ) = P ( −1.50 ≤ z ≤ 0 ) = P ( z ≤ 0 ) − P ( z ≤ −1.50 ) = 0.5000 − 0.0668 = 0.4332 (b) µ x = µ = 100 σ 48 = = 4.36 σx = n 121 Because n = 121 ≥ 30, by the central limit theorem, we can assume that the distribution of x is approximately normal. z= x −µ σx = x − 100 4.36 92 − 100 = −1.83 4.36 100 − 100 x = 100 converts to z = =0 4.36 x = 92 converts to z = P ( 92 ≤ x ≤ 100 ) = P ( −1.83 ≤ z ≤ 0 ) = P ( z ≤ 0 ) − P ( z ≤ −1.83) = 0.5000 − 0.0336 = 0.4664 (c) The probability of part (b) is greater than that of part (a). The standard deviation of part (b) is smaller because of the larger sample size. Therefore, the distribution about µ x is narrower in part (b). 3. (a) No, we cannot say anything about the distribution of sample means because the sample size is only 9 and so it is too small to apply the central limit theorem. (b) Yes, now we can say that the x distribution will also be normal with µ x = µ = 25 and σ x = z= x −µ σx = σ n = 3.5 = 1.17. 9 x − 25 1.17 26 − 25 ⎞ ⎛ 23 − 25 P ( 23 ≤ x ≤ 26 ) = P ⎜ ≤z≤ ⎟ 1.17 ⎠ ⎝ 1.17 = P ( −1.71 ≤ z ≤ 0.86 ) = P ( z ≤ 0.86 ) − P ( z ≤ −1.71) = 0.8051 − 0.0436 = 0.7615 Copyright © Houghton Mifflin Company. All rights reserved. 424 Instructor’s Resource Guide Understandable Statistics, 8th Edition 4. (a) No, we cannot say anything about the distribution of sample means because the sample size is only 16 and so it is too small to apply the central limit theorem. (b) Yes, now we can say that the x distribution will also be normal with µ x = µ = 72 and σ x = z= x −µ σx = σ n = 8 = 2. 16 x − 72 2 73 − 72 ⎞ ⎛ 68 − 72 P ( 68 ≤ x ≤ 73) = P ⎜ ≤z≤ ⎟ 2 2 ⎠ ⎝ = ( −2 ≤ z ≤ 0.5 ) = P ( z ≤ 0.5 ) − P ( z ≤ −2 ) = 0.6915 − 0.0228 = 0.6687 5. (a) µ = 75, σ = 0.8 74.5 − 75 ⎞ ⎛ P ( x < 74.5 ) = P ⎜ z < ⎟ 0.8 ⎠ ⎝ = P ( z < −0.63) = 0.2643 (b) µ x = 75, σ x = σ n = 0.8 = 0.179 20 74.5 − 75 ⎞ ⎛ P ( x < 74.5 ) = P ⎜ z < ⎟ 0.179 ⎠ ⎝ = P ( z < −2.79 ) = 0.0026 (c) No. If the weight of only one car were less than 74.5 tons, we cannot conclude that the loader is out of adjustment. If the mean weight for a sample of 20 cars were less than 74.5 tons, we would suspect that the loader is malfunctioning. As we see in part (b), the probability of this happening is very low if the loader is correctly adjusted. 6. (a) µ = 68, σ = 3 69 − 68 ⎞ ⎛ 67 − 68 ≤z≤ P ( 67 ≤ x ≤ 69 ) = P ⎜ ⎟ 3 ⎠ ⎝ 3 = P ( −0.33 ≤ z ≤ 0.33) = P ( z ≤ 0.33) − P ( z ≤ −0.33) = 0.6293 − 0.3707 = 0.2586 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 (b) µ x = 68, σ x = σ n = 425 3 =1 9 69 − 68 ⎞ ⎛ 67 − 68 P ( 67 ≤ x ≤ 69 ) = P ⎜ ≤z≤ ⎟ 1 1 ⎠ ⎝ = P ( −1 ≤ z ≤ 1) = P ( z ≤ 1) − P ( z ≤ −1) = 0.8413 − 0.1587 = 0.6826 (c) The probability in part (b) is much higher because the standard deviation is smaller for the x distribution. 7. (a) µ = 85, σ = 25 40 − 85 ⎞ ⎛ P ( x < 40 ) = P ⎜ z < ⎟ 25 ⎠ ⎝ = P ( z < −1.8 ) = 0.0359 (b) The probability distribution of x is approximately normal with µ x = 85; σ x = σ n = 25 2 = 17.68. 40 − 85 ⎞ ⎛ P ( x < 40 ) = P ⎜ z < ⎟ 17.68 ⎠ ⎝ = P ( z < −2.55 ) = 0.0054 (c) µ x = 85, σ x = σ n = 25 = 14.43 3 40 − 85 ⎞ ⎛ P ( x < 40 ) = P ⎜ z < ⎟ 14.43 ⎠ ⎝ = P ( z < −3.12 ) = 0.0009 (d) µ x = 85, σ x = σ n = 25 = 11.2 5 40 − 85 ⎞ ⎛ P ( x < 40 ) = P ⎜ z < ⎟ 11.2 ⎠ ⎝ = P ( z < −4.02 ) < 0.0002 (e) Yes; If the average value based on five tests were less than 40, the patient is almost certain to have excess insulin. Copyright © Houghton Mifflin Company. All rights reserved. 426 Instructor’s Resource Guide Understandable Statistics, 8th Edition 8. µ = 7500, σ = 1750 3500 − 7500 ⎞ ⎛ (a) P ( x < 3500 ) = P ⎜ z < ⎟ 1750 ⎝ ⎠ = P ( z < −2.29 ) = 0.0110 (b) The probability distribution of x is approximately normal with µ x = 7500; σ x = σ n = 1750 2 = 1237.44. 3500 − 7500 ⎞ ⎛ P ( x < 3500 ) = P ⎜ z < ⎟ 1237.44 ⎠ ⎝ = P ( z < −3.23) = 0.0006 (c) µ x = 7500, σ x = σ = n 1750 = 1010.36 3 3500 − 7500 ⎞ ⎛ P ( x < 3500 ) = P ⎜ z < ⎟ 1010.36 ⎠ ⎝ = P ( z < −3.96 ) < 0.0002 (d) The probabilities decreased as n increased. It would be an extremely rare event for a person to have two or three tests below 3500 purely by chance; the person probably has leukopenia. 9. (a) µ = 63.0, σ = 7.1 54 − 63.0 ⎞ ⎛ P ( x < 54 ) = P ⎜ z < ⎟ 7.1 ⎠ ⎝ = P ( z < −1.27 ) = 0.1020 (b) The expected number undernourished is 2200(0.1020) = 224.4, or about 224. (c) µ x = 63.0, σ x = σ n = 7.1 = 1.004 50 60 − 63.0 ⎞ ⎛ P ( x < 60 ) = P ⎜ z < ⎟ 1.004 ⎠ ⎝ = P ( z < −2.99 ) = 0.0014 (d) µ x = 63.0, σ x = 1.004 64.2 − 63.0 ⎞ ⎛ P ( x < 64.2 ) = P ⎜ z < ⎟ 1.004 ⎠ ⎝ = P ( z < 1.20 ) = 0.8849 Since the sample average is above the mean, it is quite unlikely that the doe population is undernourished. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 427 10. (a) From the Central Limit Theorem, we expect the x distribution to be approximately normal with the σ 2 = = 0.3651. mean µ x = µ = 16 and standard deviation σ x = 30 n (b) µ x = 16, σ x = 0.3651 17 − 16 ⎞ ⎛ 16 − 16 P (16 ≤ x ≤ 17 ) = P ⎜ ≤z≤ ⎟ 0.3651 ⎠ ⎝ 0.3651 = P ( 0 ≤ z ≤ 2.74 ) = P ( z ≤ 2.74 ) − P ( z ≤ 0 ) = 0.9969 − 0.5000 = 0.4969 (c) µ x = 16, σ x = 0.3651 15 − 16 ⎞ ⎛ P ( x < 15 ) = P ⎜ z < ⎟ 0.3651 ⎠ ⎝ = P ( z < −2.74 ) = 0.0031 11. (a) The random variable x is itself an average based on the number of stocks or bonds in the fund. Since x itself represents a sample mean return based on a large (random) sample of stocks or bonds, x has a distribution that is approximately normal (Central Limit Theorem). (b) µ x = 1.6%, σ x = σ n = 0.9% = 0.367% 6 2% − 1.6% ⎞ ⎛ 1% − 1.6% P (1% ≤ x ≤ 2% ) = P ⎜ ≤z≤ ⎟ 0.367% ⎠ ⎝ 0.367% = P ( −1.63 ≤ z ≤ 1.09 ) = P ( z ≤ 1.09 ) − P ( z ≤ −1.63) = 0.8621 − 0.0516 = 0.8105 Note: It does not matter whether you solve the problem using percents or their decimal equivalents as long as you are consistent. (c) Note: 2 years = 24 months; x is monthly percentage return. σ 0.9% µ x = 1.6%, σ x = = = 0.1837% 24 n 2% − 1.6% ⎞ ⎛ 1% − 1.6% P (1% ≤ x ≤ 2% ) = P ⎜ ≤z≤ ⎟ 0.1837% 0.1837% ⎠ ⎝ = P ( −3.27 ≤ z ≤ 2.18 ) = P ( z ≤ 2.18 ) − P ( z ≤ −3.27 ) = 0.9854 − 0.0005 = 0.9849 (d) Yes. The probability increases as the standard deviation decreases. The standard deviation decreases as the sample size increases. Copyright © Houghton Mifflin Company. All rights reserved. 428 Instructor’s Resource Guide Understandable Statistics, 8th Edition (e) µ x = 1.6%, σ x = 0.1837% 1% − 1.6% ⎞ ⎛ P ( x < 1% ) = P ⎜ z < ⎟ 0.1837% ⎠ ⎝ = P ( z < −3.27 ) = 0.0005 This is very unlikely if µ = 1.6%. One would suspect that µ has slipped below 1.6%. 12. (a) The random variable x is itself an average based on the number of stocks in the fund. Since x itself represents a sample mean return based on a large (random) sample of stocks, x has a distribution that is approximately normal (Central Limit Theorem). (b) µ x = 1.4%, σ x = σ n = 0.8% = 0.2667% 9 2% − 1.4% ⎞ ⎛ 1% − 1.4% P (1% ≤ x ≤ 2% ) = P ⎜ ≤z≤ ⎟ 0.2667% ⎠ ⎝ 0.2667% = P ( −1.50 ≤ z ≤ 2.25 ) = P ( z ≤ 2.25 ) − P ( z ≤ −1.50 ) = 0.9878 − 0.0668 = 0.9210 Note: It does not matter whether you solve the problem using percents or their decimal equivalents as long as you are consistent. (c) µ x = 1.4%, σ x = σ n = 0.8% = 0.1886% 18 2% − 1.4% ⎞ ⎛ 1% − 1.4% P (1% ≤ x ≤ 2% ) = P ⎜ ≤z≤ ⎟ 0.1886% ⎠ ⎝ 0.1886% = P ( −2.12 ≤ z ≤ 3.18 ) = P ( z ≤ 3.18 ) − P ( z ≤ −2.12 ) = 0.9993 − 0.0170 = 0.9823 (d) Yes. The probability increases as the standard deviation decreases. The standard deviation decreases as the sample size increases. (e) µ x = 1.4%, σ x = 0.1886% 2% − 1.4% ⎞ ⎛ P ( x > 2% ) = P ⎜ z > ⎟ 0.1886% ⎠ ⎝ = P ( z > 3.18 ) = 1 − P ( z ≤ 3.18 ) = 1 − 0.9993 = 0.0007 This is very unlikely if µ = 1.4%. One would suspect that the European stock market may be heating up, i.e., µ is greater than 1.4%. 13. (a) Since x itself represents a sample mean from a large n ≈ 80 (random) sample of bonds, x is approximately normally distributed according to the Central Limit Theorem. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 (b) µ x = 10.8%, σ x = σ n = 429 4.9% 5 = 2.19% 6% − 10.8% ⎞ ⎛ P ( x < 6% ) = P ⎜ z < ⎟ 2.19% ⎠ ⎝ = P ( z < −2.19 ) = 0.0143 Yes. Since this probability is so small, it is very unlikely that x would be less than 6% if µ = 10.8%. The junk bond market appears to be weaker, i.e., µ is less than 10.8%. (c) µ x = 10.8%, σ x = 2.19% 16% − 10.8% ⎞ ⎛ P ( x > 16% ) = P ⎜ z > ⎟ 2.19% ⎠ ⎝ = P ( z > 2.37 ) = 1 − P ( z ≤ 2.37 ) = 1 − 0.9911 = 0.0089 Yes. Since this probability is so small, it is very unlikely that x would be greater than 16% if µ = 10.8%. The junk bond market may be heating up, i.e., µ is greater than 10.8%. 14. (a) µ x = 6.4, σ x = σ n = 1.5 = 0.2372 40 7 − 6.4 ⎞ ⎛ 6 − 6.4 P (6 ≤ x ≤ 7) = P ⎜ ≤z≤ ⎟ 0.2372 0.2372 ⎝ ⎠ = P ( −1.69 ≤ z ≤ 2.53) = P ( z ≤ 2.53) − P ( z ≤ −1.69 ) = 0.9943 − 0.0455 = 0.9488 (b) µ x = 6.4, σ x = σ n = 1.5 = 0.1677 80 7 − 6.4 ⎞ ⎛ 6 − 6.4 P (6 ≤ x ≤ 7) = P ⎜ ≤z≤ ⎟ 0.1677 0.1677 ⎝ ⎠ = P ( −2.39 ≤ z ≤ 3.58 ) = P ( z ≤ 3.58 ) − P ( z ≤ −2.39 ) ≈ 1 − 0.0084 = 0.9916 (c) Yes. Since this is such a large probability, the chances of x not being in this time interval is extremely unlikely. A second security guard should drop in for a look. 15. (a) The sample size should be 30 or more. (b) No. If the distribution of x is normal, the distribution of x is also normal, regardless of the sample size. Copyright © Houghton Mifflin Company. All rights reserved. 430 Instructor’s Resource Guide Understandable Statistics, 8th Edition 16. (a) By the Central Limit Theorem, the sampling distribution of x is approximately normal with σ $7 mean µ x = µ = $20 and standard error σ x = = = $0.70. It is not necessary to make any 100 n assumption about the x distribution because n is large. (b) µ x = $20, σ x = $0.70 $22 − $20 ⎞ ⎛ $18 − $20 P ( $18 ≤ x ≤ $22 ) = P ⎜ ≤z≤ ⎟ $0.70 ⎠ ⎝ $0.70 = P ( −2.86 ≤ z ≤ 2.86 ) = P ( z ≤ 2.86 ) − P ( z ≤ −2.86 ) = 0.9979 − 0.0021 = 0.9958 (c) µ x = $20, σ = $7 $22 − $20 ⎞ ⎛ $18 − $20 P ( $18 ≤ x ≤ $22 ) = P ⎜ ≤z≤ ⎟ $7 $7 ⎝ ⎠ = P ( −0.29 ≤ z ≤ 0.29 ) = 0.6141 − 0.3859 = 0.2282 (d) We expect the probability in part (b) to be much higher than the probability in part (c) because the standard deviation is smaller for the x distribution than it is for the x distribution. By the Central Limit Theorem, the sampling distribution of x will be approximately normal as n increases, and its standard deviation, σ n , will decrease as n increases. The standard deviation of x , a.k.a. the standard error of x , measures the spread of the x values; the smaller σ n is, the less variability there is in the x values. The less variability there is in the values of x , the more reliable x is as an estimate or predictor of µ. For large n, approximately 95% of the possible values of x are within 2σ n of µ. The amount x a typical customer spends on impulse buys also estimates µ (recall µ x = µ x = µ ), but approximately 95% of individual impulse buys x are within 2σ of µ (using either the Empirical Rule for somewhat mound-shaped data, or assuming x has a distribution that is approximately normal). For a fixed interval, such as $18 to $22, centered at the mean, $20 in this case, the proportion of the possible x values within the interval will be greater than the proportion of the possible x values within the same interval. 17. (a) The total checkout time for 30 customers is the sum of the checkout times for each individual customer. Thus, w = x1 + x2 + … + x30 and the probability that the total checkout time for the next 30 customers is less than 90 is P(w < 90). (b) If we divide both sides of w < 90 by 30, we get so w < 3. However, w is the sum of 30 waiting times, 30 w is x . Therefore, P ( w < 90 ) = P ( x < 3) . 30 (c) The probability distribution of x is approximately normal with mean µ x = µ = 2.7 and standard deviation σ x = σ n = 0.6 30 = 0.1095. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 3 − 2.7 ⎞ ⎛ (d) P ( x < 3) = P ⎜ z < ⎟ 0.1095 ⎝ ⎠ = P ( z < 2.74 ) = 0.9969 The probability that the total checkout time for the next 30 customers is less than 90 minutes is 0.9969, i.e., P(w < 90) = 0.9969. 18. Let w = x1 + x2 + … + x36. σ 2.5 (a) w < 320 is equivalent to w < 320 or x < 8.889. µ x = µ = 8.5, σ x = = = 0.4167 36 36 36 n P ( w < 320 ) = P ( x < 8.889 ) 8.889 − 8.5 ⎞ ⎛ = P⎜ z < ⎟ 0.4167 ⎠ ⎝ = P ( z < 0.93) = 0.8238 (b) w > 275 is equivalent to w > 275 or x > 7.639. µ x = 8.5, σ x = 0.4167 36 36 P ( w > 275 ) = P ( x > 7.639 ) 7.639 − 8.5 ⎞ ⎛ = P⎜z > ⎟ 0.4167 ⎠ ⎝ = P ( z > −2.07 ) = 1 − P ( z ≤ −2.07 ) = 1 − 0.0192 ≈ 0.9808 (c) P ( 275 < w < 320 ) = P ( 7.639 < x < 8.889 ) = P ( −2.07 < z < 0.93) = P ( z < 0.93) − P ( z < −2.07 ) = 0.8238 − 0.0192 = 0.8046 19. Let w = x1 + x2 + … + x45. σ 84 (a) w < 9500 is equivalent to w < 9500 or x < 211.111. µ x = 240, σ x = = = 12.522 45 45 45 n P ( w < 9500 ) = P ( x < 211.111) 211.111 − 240 ⎞ ⎛ = P⎜ z < ⎟ 12.522 ⎝ ⎠ = P ( z < −2.31) = 0.0104 Copyright © Houghton Mifflin Company. All rights reserved. 431 432 Instructor’s Resource Guide Understandable Statistics, 8th Edition 12, 000 (b) w < 12,000 is equivalent to w > or x > 266.667. µ x = 240, σ x = 12.522 45 45 P ( w > 12, 000 ) = P ( x > 266.667 ) 266.667 − 240 ⎞ ⎛ = P⎜z > ⎟ 12.522 ⎝ ⎠ = P ( z > 2.13) = 1 − P ( z ≤ 2.13) = 1 − 0.9834 = 0.0166 (c) P ( 9500 < w < 12, 000 ) = P ( 211.111 < x < 266.667 ) = P ( −2.31 < z < 2.13) = P ( z < 2.13) − P ( z < −2.31) = 0.9834 − 0.0104 = 0.9730 20. (a) Let w = x1 + x2 + … + x9. µ x = µ = 6.3, σ x = σ n = 1.2 9 = 0.4 ⎛ w 60 ⎞ P ( w < 60 ) = P ⎜ < ⎟ 9 ⎠ ⎝9 = P ( x < 6.667 ) 6.667 − 6.3 ⎞ ⎛ = P⎜ z < ⎟ 0.4 ⎝ ⎠ = P ( z < 0.92 ) = 0.8212 ⎛ w 65 ⎞ P ( w > 65 ) = P ⎜ > ⎟ 9 ⎠ ⎝9 = P ( x > 7.222 ) 7.222 − 6.3 ⎞ ⎛ = P⎜z > ⎟ 0.4 ⎝ ⎠ = P ( z > 2.31) = 1 − P ( z ≤ 2.31) = 1 − 0.9896 = 0.0104 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 433 (b) Let w = x1 + x2 + … + x50. µ x = µ = 6.3, σ x = σ n = 1.2 50 = 0.170 ⎛ w 342 ⎞ P ( w < 342 ) = P ⎜ < ⎟ ⎝ 50 50 ⎠ P ( x < 6.84 ) 6.84 − 6.3 ⎞ ⎛ = P⎜ z < ⎟ 0.170 ⎠ ⎝ = P ( z < 3.18 ) = 0.9993 No. By the Central Limit Theorem the sample size is large enough so the sampling distribution of x is approximately normal. 21. (a) Let w = x1 + x2 + + x5 . σ 3.3 = = 1.476 n 5 ⎛ w 90 ⎞ P ( w > 90) = P ⎜ > ⎟ 5 ⎠ ⎝5 = P ( x > 18) µ x = µ = 17, σ x = 18 − 17 ⎞ ⎛ = P⎜ z > ⎟ 1.476 ⎠ ⎝ = P ( z > 0.68) = 1 − 0.7517 = 0.2483 ⎛ w 80 ⎞ (b) P ( w < 80) = P ⎜ < ⎟ ⎝5 5 ⎠ = P ( x < 16) 16 − 17 ⎞ ⎛ = P⎜ z < ⎟ 1.476 ⎠ ⎝ = P ( z < −0.68) = 0.2483 (c) P (80 < w < 90) = P (16 < x < 18) = P (−0.68 < z < 0.68) = P ( z < 0.68) − P ( z < −0.68) = 0.7517 − 0.2483 = 0.5034 Section 7.3 1. (a) Answers vary. (b) The random variable p̂ can be approximated by a normal random variable when both np and nq exceed 5. µ pˆ = p, σ pˆ = pq n Copyright © Houghton Mifflin Company. All rights reserved. 434 Instructor’s Resource Guide Understandable Statistics, 8th Edition (c) np = 33 ( 0.21) = 6.93, nq = 33 ( 0.79 ) = 26.07 Yes, p̂ can be approximated by a normal random variable since both np and nq exceed 5. 0.21( 0.79 ) ≈ 0.071 33 µ pˆ = p = 0.21, σ pˆ = continuity correction = 0.5 0.5 = ≈ 0.015 n 33 P ( 0.15 ≤ pˆ ≤ 0.25 ) = P ( 0.15 − 0.015 ≤ x ≤ 0.25 + 0.015 ) = P ( 0.135 ≤ x ≤ 0.265 ) 0.265 − 0.21 ⎞ ⎛ 0.135 − 0.21 = P⎜ ≤z≤ ⎟ 0.071 0.071 ⎠ ⎝ = P ( −1.06 ≤ z ≤ 0.77 ) = P ( z ≤ 0.77 ) − P ( z ≤ −1.06 ) = 0.7794 − 0.1446 = 0.6348 (d) No; np = 25(0.15) = 3.75 which does not exceed 5. (e) np = 48 ( 0.15 ) = 7.2, nq = 48 ( 0.85 ) = 40.8 Yes, p̂ can be approximated by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.15, σ pˆ = continuity correction = 0.15 ( 0.85 ) ≈ 0.052 48 0.5 0.5 = = 0.010 n 45 P ( pˆ ≥ 0.22 ) = P ( x ≥ 0.22 − 0.010 ) = P ( x ≥ 0.21) 0.21 − 0.15 ⎞ ⎛ = P⎜z ≥ ⎟ 0.052 ⎠ ⎝ = P ( z ≥ 1.15 ) = 1 − P ( z < 1.15 ) = 1 − 0.8749 = 0.1251 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 435 2. (a) n = 50, p = 0.36 np = 50 ( 0.36 ) = 18, nq = 50 ( 0.64 ) = 32 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.36, σ pˆ = continuity correction = 0.36 ( 0.64 ) ≈ 0.068 50 0.5 0.5 = = 0.01 n 50 P ( 0.30 ≤ pˆ ≤ 0.45 ) ≈ P ( 0.30 − 0.01 ≤ x ≤ 0.45 + 0.01) = P ( 0.29 ≤ x ≤ 0.46 ) 0.46 − 0.36 ⎞ ⎛ 0.29 − 0.36 = P⎜ ≤z≤ ⎟ 0.068 0.068 ⎠ ⎝ = P ( −1.03 ≤ z ≤ 1.47 ) = P ( z ≤ 1.47 ) − P ( z ≤ −1.03) = 0.9292 − 0.1515 = 0.7777 (b) n = 38, p = 0.25 np = 38 ( 0.25 ) = 9.5, nq = 38 ( 0.75 ) = 28.5 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.25, σ pˆ = continuity correction = 0.25 ( 0.75 ) ≈ 0.070 38 0.5 0.5 = = 0.013 n 38 P ( pˆ > 0.35 ) = P ( x > 0.35 − 0.013) = P ( x > 0.337 ) 0.337 − 0.25 ⎞ ⎛ = P⎜z > ⎟ 0.070 ⎠ ⎝ = P ( z > 1.24 ) = 1 − P ( z ≤ 1.24 ) = 1 − 0.8925 = 0.1075 (c) n = 41, p = 0.09 np = 41( 0.09 ) = 3.69 We cannot approximate p̂ by a normal random variable since np < 5. Copyright © Houghton Mifflin Company. All rights reserved. 436 Instructor’s Resource Guide Understandable Statistics, 8th Edition 3. n = 30, p = 0.60 np = 30 ( 0.60 ) = 18, nq = 30 ( 0.40 ) = 12 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.6, σ pˆ = 0.6 ( 0.4 ) ≈ 0.089 30 continuity correction = 0.5 0.5 = = 0.017 n 30 (a) P ( pˆ ≥ 0.5 ) ≈ P ( x ≥ 0.5 − 0.017 ) = P ( x ≥ 0.483) 0.483 − 0.6 ⎞ ⎛ = P⎜z ≥ ⎟ 0.089 ⎠ ⎝ = P ( z ≥ −1.31) = 0.9049 (b) P ( pˆ ≥ 0.667 ) ≈ P ( x ≥ 0.667 − 0.017 ) = P ( x ≥ 0.65 ) 0.65 − 0.6 ⎞ ⎛ = P⎜z ≥ ⎟ 0.089 ⎠ ⎝ = P ( z ≥ 0.56 ) = 0.2877 (c) P ( pˆ ≤ 0.333) ≈ P ( x ≤ 0.333 + 0.017 ) = P ( x ≤ 0.35 ) 0.35 − 0.6 ⎞ ⎛ = P⎜z ≤ ⎟ 0.089 ⎠ ⎝ = P ( z ≤ −2.81) = 0.0025 (d) Yes, both np and nq exceed 5. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 437 4. (a) n = 38, p = 0.73 np = 38 ( 0.73) = 27.74, nq = 38 ( 0.27 ) = 10.26 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.73, σ pˆ = 0.73 ( 0.27 ) ≈ 0.072 38 continuity correction = 0.5 0.5 = = 0.013 n 38 P ( pˆ ≥ 0.667 ) ≈ P ( x ≥ 0.667 − 0.013) = P ( x ≥ 0.654 ) 0.654 − 0.73 ⎞ ⎛ = P⎜z ≥ ⎟ 0.072 ⎠ ⎝ = P ( z ≥ −1.06 ) = 0.8554 (b) n = 45, p = 0.86 np = 45 ( 0.86 ) = 38.7, nq = 45 ( 0.14 ) = 6.3 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.86, σ pˆ = 0.86 ( 0.14 ) ≈ 0.052 45 continuity correction = 0.5 0.5 = = 0.011 n 45 P ( pˆ ≥ 0.667 ) ≈ P ( x ≥ 0.667 − 0.011) = P ( x ≥ 0.656 ) 0.656 − 0.86 ⎞ ⎛ = P⎜z ≥ ⎟ 0.052 ⎠ ⎝ = P ( z ≥ −3.92 ) ≈1 (c) Yes, both np and nq exceed 5 for men and for women. 5. n = 55, p = 0.11 np = 55 ( 0.11) = 6.05, nq = 55 ( 0.89 ) = 48.95 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.11, σ pˆ = 0.11( 0.89 ) ≈ 0.042 55 continuity correction = 0.5 0.5 = = 0.009 n 55 Copyright © Houghton Mifflin Company. All rights reserved. 438 Instructor’s Resource Guide Understandable Statistics, 8th Edition (a) P ( pˆ ≤ 0.15 ) ≈ P ( x ≤ 0.15 + 0.009 ) = P ( x ≤ 0.159 ) 0.159 − 0.11 ⎞ ⎛ = P⎜z ≤ ⎟ 0.042 ⎠ ⎝ = P ( z ≤ 1.17 ) = 0.8790 (b) P ( 0.10 ≤ pˆ ≤ 0.15 ) ≈ P ( 0.10 − 0.009 ≤ x ≤ 0.15 + 0.009 ) = P ( 0.091 ≤ x ≤ 0.159 ) 0.159 − 0.11 ⎞ ⎛ 0.091 − 0.11 = P⎜ ≤z≤ ⎟ 0.042 0.042 ⎠ ⎝ = P ( −0.45 ≤ z ≤ 1.17 ) = P ( z ≤ 1.17 ) − P ( z ≤ −0.45 ) = 0.8790 − 0.3264 = 0.5526 (c) Yes, both np and nq exceed 5. 6. n = 28, p = 0.31 np = 28 ( 0.31) = 8.68, nq = 28 ( 0.69 ) = 19.32 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.31, σ pˆ = 0.31( 0.69 ) ≈ 0.087 28 continuity correction = 0.5 0.5 = = 0.018 n 28 (a) P ( pˆ ≥ 0.25 ) ≈ P ( x ≥ 0.25 − 0.018 ) = P ( x ≥ 0.232 ) 0.232 − 0.31 ⎞ ⎛ = P⎜z ≥ ⎟ 0.087 ⎠ ⎝ = P ( z ≥ −0.90 ) = 0.8159 (b) P ( 0.25 ≤ pˆ ≤ 0.50 ) ≈ P ( 0.25 − 0.018 ≤ x ≤ 0.50 + 0.018 ) = P ( 0.232 ≤ x ≤ 0.518 ) 0.518 − 0.31 ⎞ ⎛ 0.232 − 0.31 = P⎜ ≤z≤ ⎟ 0.087 ⎠ ⎝ 0.087 = P ( −0.90 ≤ z ≤ 2.39 ) = P ( z ≤ 2.39 ) − P ( z ≤ −0.90 ) = 0.9916 − 0.1841 = 0.8075 (c) Yes, both np and nq exceed 5. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 439 7. (a) n = 100, p = 0.06 np = 100 ( 0.06 ) = 6, nq = 100 ( 0.94 ) = 94 p̂ can be approximated by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.06, σ pˆ = 0.06 ( 0.94 ) ≈ 0.024 100 continuity correction = 0.5 = 0.005 100 (b) P ( pˆ ≥ 0.07 ) ≈ P ( x ≥ 0.07 − 0.005 ) = P ( x ≥ 0.065 ) 0.065 − 0.06 ⎞ ⎛ = P⎜z ≥ ⎟ 0.024 ⎠ ⎝ = P ( z ≥ 0.21) = 0.4168 (c) P ( pˆ ≥ 0.11) ≈ P ( x ≥ 0.11 − 0.005 ) = P ( x ≥ 0.105 ) 0.105 − 0.06 ⎞ ⎛ = P⎜z ≥ ⎟ 0.024 ⎠ ⎝ = P ( z ≥ 1.88 ) = 0.0301 Yes, since this probability is so small, it should rarely occur. The machine might need an adjustment. 8. (a) n = 50, p = 0.565 np = 50 ( 0.565 ) = 28.25, nq = 50 ( 0.435 ) = 21.75 p̂ can be approximated by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.565, σ pˆ = 0.565 ( 0.435 ) ≈ 0.070 50 continuity correction = 0.5 0.5 = = 0.01 50 n (b) P ( pˆ ≤ 0.53) ≈ P ( x ≤ 0.53 + 0.01) = P ( x ≤ 0.54 ) 0.54 − 0.565 ⎞ ⎛ = P⎜z ≤ ⎟ 0.070 ⎠ ⎝ = P ( z ≤ −0.36 ) = 0.3594 Copyright © Houghton Mifflin Company. All rights reserved. 440 Instructor’s Resource Guide Understandable Statistics, 8th Edition (c) P ( pˆ ≤ 0.41) ≈ P ( x ≤ 0.41 + 0.01) = P ( x ≤ 0.42 ) 0.42 − 0.565 ⎞ ⎛ = P⎜z ≤ ⎟ 0.070 ⎠ ⎝ = P ( z ≤ −2.07 ) = 0.0192 (d) Meredith has the more serious case because the probability of having such a low reading in a healthy person is less than 2%. 9. total number of successes from all 12 quarters total number of families from all 12 quarters 11 + 14 + … + 19 = 12 ( 92 ) p= 206 1104 = 0.1866 = q = 1 − p = 1 − 0.1866 = 0.8134 µ pˆ = p ≈ p = 0.1866 σ pˆ = pq ≈ n 0.1866 ( 0.8134 ) pq = ≈ 0.0406 n 92 Check: np = 92 ( 0.1866 ) = 17.2, nq = 92 ( 0.8134 ) = 74.8 Since both np and nq exceed 5, the normal approximation should be reasonably good. Center line = p = 0.1866 Control limits at p ± 2 pq n = 0.1866 ± 2 ( 0.0406 ) = 0.1866 ± 0.0812 or 0.1054 and 0.2678 Control limits at p ± 3 pq n = 0.1866 ± 3 ( 0.0406 ) = 0.1866 ± 0.1218 or 0.0648 and 0.3084 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 441 P Chart for r 3.0SL=0.3084 0.3 Proportion 2.0SL=0.2678 0.2 – P=0.1866 0.1 –3.0SL=0.1054 –3.0SL=0.06474 5 Sample Number 0 10 There are no out-of-control signals. 10. total number of defective cans total number of cans 8 + 11 + … + 10 = 110 (15 ) p= 133 1650 = 0.08061 = q = 1 − p = 1 − 0.08061 = 0.91939 µ pˆ = p ≈ p = 0.08061 σ pˆ = pq ≈ n pq = n ( 0.08061)( 0.91939 ) ≈ 0.02596 110 Check: np = 110 ( 0.08061) = 8.9, nq = 110 ( 0.91939 ) = 101.1 Since both np and nq exceed 5, the normal approximation should be reasonably good. Center line = p = 0.08061 Control limits at p ± 2 pq n = 0.08061 ± 2 ( 0.02596 ) = 0.08061 ± 0.05192 or 0.02869 and 0.1325 Control limits at p ± 3 pq n = 0.08061 ± 3 ( 0.02596 ) = 0.08061 ± 0.07788 or 0.00273 and 0.1585 Copyright © Houghton Mifflin Company. All rights reserved. 442 Instructor’s Resource Guide Understandable Statistics, 8th Edition P Chart for r 3.0SL=0.1585 0.15 Proportion 2.0SL=0.1325 0.10 – P=0.08061 0.05 –3.0SL=0.02869 0.00 –3.0SL=0.002738 0 5 10 Sample Number 15 There are no out-of-control signals. It appears that the production process is in reasonable control. 11. total number who got jobs total number of people 60 + 53 + … + 58 = 75 (15 ) p= 872 1125 = 0.7751 = q = 1 − p = 1 − 0.7751 = 0.2249 µ pˆ = p ≈ p = 0.7751 σ pˆ = pq ≈ n pq = n ( 0.7751)( 0.2249 ) ≈ 0.0482 75 Check: np = 75 ( 0.7751) = 58.1, nq = 75 ( 0.2249 ) = 16.9 Since both np and nq exceed 5, the normal approximation should be reasonably good. Center line = p = 0.7751 Control limits at p ± 2 pq n = 0.7751 ± 2 ( 0.0482 ) = 0.7751 ± 0.0964 or 0.6787 to 0.8715 Control limits at p ± 3 pq n = 0.7751 ± 3 ( 0.0482 ) = 0.7751 ± 0.1446 or 0.6305 to 0.9197 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 443 P Chart for r 1 3.0SL=0.9197 Proportion 0.9 2.0SL=0.8715 0.8 – P=0.7751 0.7 –3.0SL=0.6787 0.6 –3.0SL=0.6305 1 0 5 10 Sample Number 15 Out-of-control signal III occurs on days 4 and 5, Out-of-control signal I occurs on day 11 on the low side and day 14 on the high side. Out-of-control signals on the low side are of most concern for the homeless seeking work. The foundation should look to see what happened on that day. The foundation might take a look at the out of control periods on the high side to see if there is a possibility of cultivating more jobs. Chapter 7 Review 1. (a) The x distribution approaches a normal distribution. (b) The mean µ x of the x distribution equals the mean µ of the x distribution, regardless of the sample size. (c) The standard deviation σ x of the sampling distribution equals σ n , where σ is the standard deviation of the x distribution and n is the sample size. (d) They will both be approximately normal with the same mean, but the standard deviations will be σ 50 and σ 100 respectively. 2. All the x distributions will be normal with mean µ x = µ = 15. The standard deviations will be: n = 4: σ x = n = 16: σ x = n = 100: σ x = σ n σ n σ n = = = 3 4 3 16 3 = 3 2 = 3 4 100 = 3 10 3. (a) µ = 35, σ = 7 40 − 35 ⎞ ⎛ P ( x ≥ 40 ) = P ⎜ z ≥ ⎟ 7 ⎠ ⎝ = P ( z ≥ 0.71) = 0.2389 Copyright © Houghton Mifflin Company. All rights reserved. 444 Instructor’s Resource Guide Understandable Statistics, 8th Edition (b) µ x = µ = 35, σ x = σ n = 7 7 = 9 3 40 − 35 ⎞ ⎛ P ( x ≥ 40 ) = P ⎜ z ≥ ⎟ 7 ⎜ ⎟ 3 ⎝ ⎠ = P ( z ≥ 2.14 ) = 0.0162 4. (a) µ = 38, σ = 5 35 − 38 ⎞ ⎛ P ( x ≤ 35 ) = P ⎜ z ≤ ⎟ 5 ⎠ ⎝ = P ( z ≤ −0.6 ) = 0.2743 (b) µ x = µ = 38, σ x = σ n = 5 = 1.58 10 35 − 38 ⎞ ⎛ P ( x ≤ 35 ) = P ⎜ z ≤ ⎟ 1.58 ⎠ ⎝ = P ( z ≤ −1.90 ) = 0.0287 (c) The probability in part (b) is much smaller because the standard deviation is smaller for the x distribution. 5. µ x = µ = 100, σ x = σ = 15 = 1.5 n 100 P (100 − 2 ≤ x ≤ 100 + 2 ) = P ( 98 ≤ x ≤ 102 ) 102 − 100 ⎞ ⎛ 98 − 100 = P⎜ ≤z≤ ⎟ 1.5 1.5 ⎠ ⎝ = P ( −1.33 ≤ z ≤ 1.33) = P ( z ≤ 1.33) − P ( z ≤ −1.33) = 0.9082 − 0.0918 = 0.8164 6. µ x = µ = 15, σ x = σ = 2 = 0.333 n 36 P (15 − 0.5 ≤ x ≤ 15 + 0.5 ) = P (14.5 ≤ x ≤ 15.5 ) 15.5 − 15 ⎞ ⎛ 14.5 − 15 = P⎜ ≤z≤ ⎟ 0.333 ⎠ ⎝ 0.333 = P ( −1.5 ≤ z ≤ 1.5 ) = P ( z ≤ 1.5 ) − P ( z ≤ −1.5 ) = 0.9332 − 0.0668 = 0.8664 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 7. µ x = µ = 750, σ x = 445 σ = 20 = 2.5 n 64 750 − 750 ⎞ ⎛ (a) P ( x ≥ 750 ) = P ⎜ z ≥ ⎟ 2.5 ⎠ ⎝ = P ( z ≥ 0) = 0.5000 755 − 750 ⎞ ⎛ 745 − 750 ≤z≤ (b) P ( 745 ≤ x ≤ 755 ) = P ⎜ ⎟ 2.5 2.5 ⎠ ⎝ = P ( −2 ≤ z ≤ 2 ) = P ( z ≤ 2 ) − P ( z ≤ −2 ) = 0.9772 − 0.0228 = 0.9544 8. (a) Miami: µ = 76, σ = 1.9 77 − 76 ⎞ ⎛ P ( x < 77 ) = P ⎜ z < ⎟ 1.9 ⎠ ⎝ = P ( z < 0.53) = 0.7019 Fairbanks: µ = 0, σ = 5.3 3−0⎞ ⎛ P ( x < 3) = P ⎜ z < ⎟ 5.3 ⎠ ⎝ = P ( z < 0.57 ) = 0.7157 (b) Since x has a normal distribution, the sampling distribution of x is also normal regardless of the sample size. σ = 1.9 = 0.718 Miami: µ x = µ = 76, σ x = n 7 77 − 76 ⎞ ⎛ P ( x < 77 ) = P ⎜ z < ⎟ 0.718 ⎠ ⎝ = P ( z < 1.39 ) = 0.9177 Fairbanks: µ x = µ = 0, σ x = σ = 5.3 = 2.003 n 7 3−0 ⎞ ⎛ P ( x < 3) = P ⎜ z < ⎟ 2.003 ⎠ ⎝ = P ( z < 1.50 ) = 0.9332 Copyright © Houghton Mifflin Company. All rights reserved. 446 Instructor’s Resource Guide Understandable Statistics, 8th Edition (c) We cannot say anything about the probability distribution of x , because the sample size is not 30 or greater. Consider using all 31 days. σ = 1.9 = 0.341 Miami: µ x = µ = 76, σ x = n 31 77 − 76 ⎞ ⎛ P ( x < 77 ) = P ⎜ z < ⎟ 0.341 ⎠ ⎝ = P ( z < 2.93) = 0.9983 Fairbanks: µ x = µ = 0, σ x = σ = 5.3 = 0.952 n 31 3−0 ⎞ ⎛ P ( x < 3) = P ⎜ z < ⎟ 0.952 ⎠ ⎝ = P ( z < 3.15 ) = 0.9992 9. (a) n = 50, p = 0.22 np = 50 ( 0.22 ) = 11, nq = 50 ( 0.78 ) = 39 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.22, σ pˆ = continuity correction = 0.22 ( 0.78 ) 50 ≈ 0.0586 0.5 0.5 = = 0.01 n 50 P ( 0.20 ≤ pˆ ≤ 0.25 ) ≈ P ( 0.20 − 0.01 ≤ x ≤ 0.25 + 0.01) = P ( 0.19 ≤ x ≤ 0.26 ) 0.26 − 0.22 ⎞ ⎛ 0.19 − 0.22 = P⎜ ≤z≤ ⎟ 0.0586 ⎠ ⎝ 0.0586 = P ( −0.51 ≤ z ≤ 0.68 ) = P ( z ≤ 0.68 ) − P ( z ≤ −0.51) = 0.7517 − 0.3050 = 0.4467 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 7 447 (b) n = 38, p = 0.27 np = 38 ( 0.27 ) = 10.26, nq = 38 ( 0.73) = 27.74 Approximate p̂ by a normal random variable since both np and nq exceed 5. µ pˆ = p = 0.27, σ pˆ = continuity correction = ( 0.27 )( 0.73) 38 ≈ 0.0720 0.5 0.5 = = 0.013 n 38 P ( pˆ ≥ 0.35 ) ≈ P ( x ≥ 0.35 − 0.013) = P ( x ≥ 0.337 ) 0.337 − 0.27 ⎞ ⎛ = P⎜z ≥ ⎟ 0.0720 ⎠ ⎝ = P ( z ≥ 0.93) = 0.1762 (c) n = 51, p = 0.05 np = 51( 0.05 ) = 2.55 No, we cannot approximate p̂ by a normal random variable since np < 5. Copyright © Houghton Mifflin Company. All rights reserved.