You are on page 1of 8
CHAPTER 10 ESTIMATION: SINGLE POPULATION 387 a q ‘The sample mean is X = 4, and the median is m= q ‘The 95% confidence interval estimates using the, 4 B ‘As you can see, the interval based on the sample mean is narrower; as we pointed out previously, natrower intervals provide more precise information. To understand why the sample mean produces better estimators than the sample median, recall how the median is calculated. We simply put the data in order and select the observation that falls in the middle. ‘Thus, as far as the median is concerned the data appear as 1 2 3 4 6 6 7 8 8 By ignoring the actual observations and using their ranks instead, we lose information. ‘With less information, we have less precision in the interval estimators and so ultimately ‘make poorer decisions. Legfhing the techniques Ina random sample of 400 observations from ‘ population whose variance is a? = 100, we calculated X = 75, Find the 95% confidence interval ‘estimate of the population mean us ‘Suppose that a rencom sample of five observations was taken from a normal population whose variance is 25. The results are 8, 16, 12, 6, 7 Find the 99% confidence intarval estimate of the population mean. Arrandom sample of 400 observations from a population whose standard deviation is 90 produced X = 1600. Find the 90% confidence. interval estimate of p. ‘The following observations were drawn from a ‘normal population whose variance is 100: 12.8 22 15 30 6 39 48 Determine the 90% confidence interval of the population mean. Describe what happens to the width of a confidence intarval estimate of u when each of the following ocours: ‘a The confidence level inareases fram 95% to 99%. b The sample size decreases, The value of a increases. Exercises 10.14-10.21 are ‘whatif analyses’ designed to determine what happens to the interval estimate when the confidence level, the sample size and the standard deviation change. These problems can be solved ‘manually or by using the Exce!2-Estimate_Mean worksheet in the Estimators workbook. 10.14 a A statistics practitioner took a random sample (of 50 observations from a population whose standard deviation is 25 and computed the sample mean to be 100. Estimate the population mean with 90% confidence, Repeat part (aj using a 95% confidence level Ropeat part (al using a 99% confidonce level Describe the effect on the confidence interval ‘ostimate of increasing the confidence level ‘A andom sample of 25 observations was drawn from a normal population whose standard deviation is 50. The sample mean was 200. Estimate the population mean with 959% confidence. Repeat part (a changing the population standard deviation to 26. Repeat part (a) changing the population standard deviation to 10, Describe what happens to the contidence interval estimate when the standard deviation is. decreased, ‘Arandom sample of 25 was drawn from a normal distribution whose standard deviation is 5. The sample mean was 80. Determine the 95% confidence interval estimate of the population mean. 388 PART 2 STATISTICAL INFERENCE tb Repeat part (a) with a sample size of 100. Roopeat part (a) with a sample size of 400. Describe what happens to the confidence interval estimate when the sample size increases. ae 10.17 Given the following information, determine the 98% confidence interval estimate of the population mean: ¥=500 o=12 n=50 b Repeat part (a) using @ 95% confidence level © Repeat part (a) using a 90% confidence level d Review parts (a) to (c] and discuss the effect on the confidence interval estimator of decreasing the confidence level 10.18 a The mean of a sample of 26 was calculated as X = 500. The sample was randomly drawn from ‘ normal population whose standard deviation is 18. Estimate the population mean with 999% confidence 'b Repeat part (a) changing the population standard deviation to 30, © Repeat part (a} changing the population standard deviation to 60. d_ Desoribe what happens ta the confidence interval estimato when the standard deviation is increased H0.19 a Astatistics practitioner randomly sampled 100 ‘observations from @ normal population whose standard deviation is 5 and found that X = 10, Estimate the population mean with 90% confidence, Repeat part (a) with a sample size of 26. © Repeat part {a} with a sample size of 10. d_ Deseribe what happens to the confidence interval estimate when the sample size decreases, 10.20 a From the information given here for @ sample taken from a normal population, determine the 95% confidence interval estimate of the population mean: 100 9 =20 n=25 b Repeat part (a) with X = 200. © Repeat part (a) with X = 600. d_ Describe what happens to the width of the confidence interval estimate when the sample mean increases. 10.21 a Arandom sample of 100 observations was randomly drawn from a population whose standard deviation is 8. The sample mean was calculated as X = 400, Estimate the population mean with 99% confidence, by figpeat part (a) with X = 200. Repeat part (a) with X = 100, Describe what happens to the width of the ‘confidence interval estimate when the sample mean decreases Exercises 10.22 ~10.26 are based on the optional subsection ‘Estimating the population mean using the ‘sample median’ on pages 386-7. All exercises assume that the population is norma 10.22 Is the sample median an unbiased estimator of the population mean? Explain 10.23 Is the sample median a consistont estimator of the population mean? Explain. 10.24 Show that the sample mean is relatively more: efficient than the sample median when estimating the population mean, 10.25 a Given the following information, determine the 90% confidence interval estimate of the Population mean using the sample median, Sample median =500 o=12 n= 60 b Compare your answer in pat (a) with that produced in part () of Exercise 10.17, Why is the confidence interval estimate based on the ‘sample median wider than that based on the sample mean? Applying the techniques 10.26 Self-correcting exercise, In a survey conducted 10 determine, arnong other things, the cost of holidays, 164 individuals were randomly sampled. Each was asked to assess the total cost of his or hor most recent holiday. The average cost was 82772. Assuming that the population standard deviation was $800, estimate the population mean ‘cost of holidays with 999% confidence, 27 Asurvey of 20 Australian companies indicated that the average annual incorne of company secretaries was $120000. Assuming that the population standard deviation is $7500 and that the annual | comes are normally distributed, calculate the 909% Confidence interval estimate of the average annual income of all company secretaries 10.28 In a random sample of 70 students ina large University, a dean found that the mean weekly i time devoted to homework was 14.3 hours. If wo assume that homework time is normally distributed, with a population standard deviation of 4.0 hours, find the 99% confidence interval ostimate of the weekly time spent doing homework forall the university's students. 410.29 To dotermine the mean waiting time for his ‘customers, a bank manager took e random sample Cf 50 customers and found that the mean waiting time was 7.2 minutes. Assuming that the population standard deviation is known to be 5 minutes, find the 90% confidence interval estimate of the mean waiting time forall of the bank's customers, Computer/manual applications The following exercises may be answered manually or with the assistance of a computer and software. Excel's Estimators workbook can also be used for this purpose. 10.30. XR10-30 The following data represent a random ‘sample of 9 marks (out of 10} on a statistics quiz. The marks are normally distributed with @ population standar vation of 2, Estimate the population mean with 90% confidence. 7] [7 [5 {4 ja [3 fo ja | 10.31. XRI0-21 The following observations are the ages of a random sample of eight men in a bar. It is known that the ages are normally distributed with a population standard deviation of 10. Determine the {95% confidence interval estimate of the population ‘mean, Interpret the interval estimate. 52 | 68 | 2 [ae | ae 30 | 56 10.32. XRI0-32 How many rounds of golf do physicians {who play golfl play per year? A survey of 12 physicians revealed the following numbers: a fa] i] [as [ar]ve| 8] 17] [20/61] Estimate with 96% confidence the mean number cof rounds per year played by these physicians, ‘assuming that the number of rounds is normally distributed with s population standard deviation of 12. 10.33 XRI0-13 Among the most exciting aspects of a University lecturer's lif are the departmental ‘meetings where such critical issues as the colour the walls will be painted and who gots a new dosk are decided. A sample of 20 lecturers was ‘asked how many hours per year are devoted to CHAPTER 10 ESTIMATION: SINGLE POPULATION 389 34 10.35 10.36 10.37 these meetings. The responses are listed here. ‘Assuring that hours spent on meetings is normally distributed with a standard deviation of 8 hours, estimate the mean number of hours spent at doparimental meatings by all lecturers. Use a confidence level of 90%. wivtale|[w 7{e{ol|s[uj|wiwl als] a] 3/8 XR10-34 Tho numbor of used cars sold annually by salespeople is normaly distributed with a standard deviation of 18. Arandom sample of 15 salespeople ‘was taken and the number of cars each sold is listed here. Find the 95% confidenes interval estimate of the population mean. Interpret the interval estimate 7 | 42 | be | oo) wil | 7m | se | 71 | 60 [101 | m | 55 | a8 | XRID-35 It is known that the amount of time needed to change the ol ina car is normally distributed vith a standard deviation of 6 minutes. The amount of time {in minutes) to complete @ random sample of 10 oll changes was recorded and listed here. Compute the 99% confidence intarval the mean of the population, ate of [a [10 | 6 [15 | 18 | 12 | 25 [ 20 | 18 | 24] ARI0-36 Suppose that the emount of time teenagers spend weekly working at part-time jobs is normally cistributed with e standard deviation of 40 minutes. random sample of 15 teenagers was drawn and ‘each reported the amount of time (in minutes) ‘spent at parttime jobs, These are listed here. Determine the 95% confidence interval estimate of the population mean. 1a0 | 190 | 160) 165 | wo | 190) 720 | 60 | 00 | 180 | 80 | 240 | 210 | 150 6 |_| XR10-37 Ono of the few negative side effects of quitting smoking is weight gain, Suppose that the ‘weight gain in the 12 months following a cessation in smoking is normally distributed with a standard doviation of Skg. To estimate the mean weight gain, a random sample of 13 quitters was drawn and their weight gains recorded and listed here. Determine the 90% confidence interval estimate of ‘the mean 12-month weight gain forall quiters. a[ala[i[7 ils eye lelal* 17 400 PART 2 STATISTICAL INFERENCE The following exercises can be solved manually or by 10.48 a A random sample of 8 obsorvations wes drawn report their income, people often refuse to divulge this information. This is a troublesome issue for statistics practitioners. We can't force people to answer our questions. However, if the number of non-responses is high, the results of our analysis may be invalid because the sample is no longer truly random, To understand why, suppose that people who are in the top quarter of household incomes regularly refuse to answer questions about thelt incomes, ‘The estimate of the population household income mean will be lower than the actual value, The issue can be complicated, There are several ways to compensate for non-responses, ‘The simplest method is simply to eliminate them. To illustrate, suppose that in a political survey, respondents are asked for whom they intend to vote ina two-candidate race, Surveyors record the results as 1 = candidate A, 2 = candidate B, 3 = ‘Don't know, and 4 ~ ‘Refuse to say’ Ifwe wish to infer something about the proportion of decided voters who will vote for candidate A, we can simply omit codes 3 and 4. If we are doing the work manually, we will count the number of voters who prefer candidate A and the number who prefer candidate B, ‘The sum of these two numbers is the total sample size, In the language of statistical software, nonresponses that we wish to eliminate are collectively called missing data. Software packages deal with missing data in different ways The Appendix to this chapter presents Excel instructions for Missing Data, and Recoding Data describes how to address the problem of missing data in Excel as well as how to recode data. In Excel, the nonresponses appear as blariks, We complete this section with a review of how to identify this technique. To recognise when touse the t-confidence interval estimator ofa population mean, remember the following factors. TET Factors that identify the interval estimator of a population mean 1 Problem objective: to describe a single population I 2 Data type: numerical (quantitative) | 3° Type of descriptive measurement: centval location 4 Population variance: unknown using the Excel Estimators workbook that is available ‘rom a normal population. The sample mean trom the companion website. ‘and sample standard deviation are X = 40 and 0. Estimate the population mean with 95% Learning the techniques confidence, 10.46 XRID-38 The following data were drawn from a 1 Ropest part) assuming that you know thatthe population standard deviation is o = 10. normal population, 4)ela Estimate the populat confidence, 47 You ate given the following statistics: X= 166 Estimate the population mean with 95% confidence. niw[e[wlele Explain why the intorval estimate produced in part (b is narrower than that in part (a 10.49 @ Estimate the population mean with 90% Confidence given the following statisti: 30; S=149 n=12 b Repeat part (a) assuming that you know that the Population standard deviation is o = 30. © Explain why the interval estimate produced in part (ois narrower than that in part (a. n=5, 10.50 a Ina random sample of 500 observations drawn from @ normal population, the sample mean and sample standard deviation were caleulated as X = 360 and s = 100, Estimate the population ‘mean with 99% confidence. b Repeat part (a) assuming that you know that the population standard deviation is o = 100, © Explain why the interval estimates were virtually Identical. 40.51. XR10-81 A parking officer is conducting an analysis Of the mount of time left on parking meters, A, ‘uick survey of 16 cars that have just left their ‘metered parking spaces produced the following times {in minutes). Estimate with 95% confidence the mean amount of time left forall the vacent maters. 2 ea nia [ 20 | 2 10.52. NRIO-2 Porto a uivesity lecturer's job sto publish bis or her research. Ths task oft entails roading a variety of journal articles to keep upto date, Ta help determine faulty standards, tho dan ofa business school surveyed random sample of 12 lecturers across the country and asked them to count the numberof joural articles they road h a ‘ypicel month, These data af sted here. Estimate with 80% confidence the mean numberof journal articles read monthly by lecturers. es [ee 20] 58 1.5 Most onnes of data emer str the plore on the camera Some wil ven downoad these os camp or it thm sng the ovm prnr or use a commer re A fm roceodngcomony wate Kw haw many ples were sted on carer, Aondom ‘empl of 10 gt camera owne's produced he data gen hr. Etna th 8% confidence the ten umber of pts sted on igh comeres eee sla ee Tee] 10.54 a A random sample of 25 obsorvations was

You might also like