Learning Objectives When you have completed this chapter you will be able to: LO1 Define a hypothesis and hypothesis testing, ‘LO2 Describe the five-step hypothess-testing procedure. L03 Distinguish between a one-tailed and a two-tailed test of hypothesis. LO4 Conducta test of hy- pothesis about a population mean, LOS Conduct a test of hy- pothesis about @ population proportion. LOG Define Type /and Type i! errors. One-Sample Tests of Hypothesis Dole Pineapple, Inc, is concemed that the 16-ounce can af sliced ineapple fs being overiled. Assume the standard deviation of he Drocesss.03 ounces. The quality control department took a random Sample of 50 cans and found that the arithmetic mean weight was 16.05 ounces. At the 5 percent evel of significance, can we conclude ‘thatthe mean weight i greater than 16 ounces? Determine the ‘P-value. (Exercise 28, LO)r ne-Sample Tests of Hypothesis 289 Eben 1: State the Null H 5 and the Alternate Hypothese an) (H) 1 The first step is to state th pe the five- 'e hypothesis beir ‘ yo 02 esing designated Hand read *H sub ae ‘ested. It is called the null hypothesis, =e and the subscript zero implies no ae cabal letter H stands for hypothesis, oo in the null hypothesis, meaning that heey Mere is usually a “not” or a “no” term hypothesis Is that the mean ruraae ttre 18 “no change.” For example, the nul ferent from 60,000. The ru ven on the steel-belted tre is not dit- it bbe written H,: = 60,000. Generally te eee Of testing. We either reject rejected unless our sample hypothesis is a statement that is not We should enone ye Convincing evidence tat i is false. the sample data, we cannot say thatthe rar ea es ba of way, fling to reject the na et ll hypothess fs tue To putt another have failed to dsprove A eee does not prove that My is tue, it means we fetal pide Helis ne without any doubt the null hypothesis is true, would have to test, nanan Cnt me fa known. To actually determine it, we feasible. The alternative i o take a sample fon tte section aly not > 1@ population. saletenulhypotesis We usually begin by stating th satedanibe ference between." or “the mean ose een or ees gant tyes cantly different from. . . .” When we select a sample from a population, the sample Statistic is usually numerically diferent from the hypothesized population parameter. ord team Peete ea ene Oe Se Os an imple of 12 glass plates is 69.5 psi, We, must make a decision about the difference of 0.5 psi. Is it a true difference, that is, a sig- nificant difference, or is the difference between the sample statistic (69.5) and the hypothesized population parameter (70.0) due to chance (sampling)? As noted, to answer this question we conduct a test of significance, commonly referred to as a test of hypothesis. To define what is meant by a null hypothesis: NULL HYPOTHESIS A statement about the value of a population parameter developed for the purpose of testing numerical evidence. ‘The alternate hypothesis describes what you will conclude if you reject the null hypothesis. It is written H, and ts read “H sub one.” tis also referred to as the research hypothesis. The altemiate hypothesis is accepted ifthe sample data provide ls with enough statistical evidence that the null hypothesis is false. ‘ALTERNATE HYPOTHESIS A statement that is accepted ifthe sample data provide sufficient evidence that the null hypothesis is fase ‘The following example will help clarify what is meant by the null hypothesis and the alternate hypothesis. A recent article indicated the mean age of U.S. commer ng Gireratt fs 15 years. To conduct a statistical test regarding this statement, the fist step isto determine the null and the atemate hypotheses. The null hypothesis feprosents the current oF reported condition. Its writen Hi: m=.18. The altemate fypotnests is that the statement isnot true, that i, Hy: 15. Ife important to remerba ht no mate ow the problem is stated, the nl hypotess wil avaye the al sign (=) wil canny Seas thera potas a re oatraey ey aes et Srzpuiteva’ fo noe i or caleuaos, Wo tne ee ae ne ee only if the data suggests the null hypothesis is untrus ae . ef290 Chapter 10 ‘ Significance is and alternate hypothesis, the next Step ig Step 2: Select a Level of ‘Ater setting up the null hypothes! state the level of significance. LEVEL is true. is designated a, the Greek letter alpha. It is also The level Soa sa ‘nay be a more appropriate term because t i level is lly true. Select level of times called the erecting the null hypothesis when it is really if take of rejecting the null hy lied to all t aera ba There is no one level of significance that is appli fests. The .05 joy fe), the .01 level, and the .10 level, are the rere iy value betweer! 0 atid 1 can Be tae Tra xd for consumer resent ee 01 for quai ing. You, the researcher, must decide on assurance, and 10 for pole Pelt. sion rule and colecting sample dag of significance before formulating a de ° fa oo ilustrate how itis possible to reject a true hypothesis, suppose a firm many. facturing personal computers uses a large number of printed circuit boards. Supp, bid on the boards, and the one with the lowest bid, awarded a sizable contract. Suppose the contrat specifies that the computer manufacturers quality. assurance department will sample all incoming ship- ments of circuit boards. If more than 6 percent of te boards sampled are defective, the shipment is rejcte The null hypothesis is that the incoming shipment of boards contains 6 percent or less defective boats. The alternate hypothesis is that more than 6 percent of the boards are defective. ‘A sample of 50 circuit boards received July 21 ‘rom Allied Electronics revealed that 4 boards, or 8 per- cent, were defective. The shipment was rejected because it exceeded the maximum of 6 percent defective printed circuit boards. If the shipment acu ally failed to meet the criteria for acceptance, then the decision to return the boards to the supplier was cottect. However, suppose the 4 defective printed circuit boards selected in the ‘sample of 50 were the only ones in the shipment of 4,000 boards. Then only too! {percent were defective (4/4,000 = .001). In that case, less than 6 percent of he entire shipment were defective, and rejecting the shipment was an error. In terms of hypothesis testing, we rejected the null hypothesis when we should have Wa ca eal eeetiont, By rejecting a true null hypothesis, we committed & 'y of committing a Type I error is a. Sa TYPE ERROR Rejecting the null hypothesis, H,, when it Is true. LO6 Define Type /and The probability of committing another t Type llertors, lgnated by the Greek letter beta (8). ail TYPE ll ERROR Accepting the null hypothesis when it is false. The firm manufacturing personal 1 enor Computers would commit a Type I St, Ae. the manufacturer, an incoming shipment of printed cicut boards fr led Electronics contained 15 percent substandard boards, yet the stl (often stated as the 5 percent ‘common levels of significance, tionally, the .05 level is selectet 'ype of error, called a Type Il eros 0°‘Tedecson rule states Decendtons when H, ie, Ono-Sample Tests of Hypothesis 2m was accepted. How could this hi (4 percent) wppen? Su fh sted poco tan he yg nh. oo tive boards, the shiprmen: eat the sample contained lesa than 6 percent eet boards selected in the sample weep |t could be that by chance the 48 good Consisting of thousands of wea "2 On acceptable ones inthe entre shipmant types of error—a Type | error, wherein the have been accepted, and a Type Il error, when it should Step 3: Select the Test Statistic ‘There are many test statistics. In this chapter we use both z and t as the test sta- tistic, In later chapters we will use such test statistics as F and x2, called chi-square. TEST STATISTIC A value, determined from sample information, used to determine whether to reject the null hypothesis. In hypothesis testing for the mean (ju) when o is known, the test statistic z is com- puted by: {1s [10-41] | ) \ a ‘The z value is based on the sampling distribution of X, which follows the normal ‘mean (1;) equal to yt, and a standard deviation os, which is equal fo o/vn. We can thus determine whether the difference between X and y. is statistically significant by finding the number of standard deviations X using formula (10-1). Step 4: Formulate the Decision Rule ‘A decision rule is a statement of the specific conditions under which the null hypoth~ esis is ejected and the conditions under which itis not rejected. The region or area of rejection defines the locaton ofall those values that are so large or so small that the probebilty of their occurrence under a true null hypothesis is rather remote. Bart 10-1 portrays the rejection region for a test of significance that will be ‘conducted later in the chapter.292 Statistics in Action Doing World Warll, | allie military plan- ners needed estimates ofthe number of German tanks. The information provided by traditional spying methods was not reli able, but statistical ‘methods proved to be valuable, For exam ple, espionage and reconnaisance led analysts to estimate that 1,550 tanks were produced during June 1941. However, ‘using the serial numbers of eaptured tanks an statistical, ‘anal, military planners estimated 244, The actual num berproduced as determined fom (German production record, was 271. The estimate using stats ‘cal analysis turned ‘out to be much more sccurate. Asmar ‘ype of analysis mas used to estimate the _number of rag tanks destroyed dirng Desert Stomn, a chapter 10 Note in the chart that: 4. The area where the null hypothesis is not rejected is to the left of 1.65. Wey explain how to get the 1.65 value shortly. ‘The area of rejection is to the right of 1.65. : ‘A one-tailed test is being applied. (This will also be explained later) The .05 level of significance was chosen The sampling distribution of the statistic z follows the normal probaiy distribution. 6. The value 1.65 separates the regions where the null hypothesis is rejected ani where it is not rejected, 7. The value 1.65 is the critical value. CRITICAL VALUE The dividing point between the region where the null hypothesis is rejected and the region where it is not rejected. Step 5: Make a Decision . ' The fifth and final step in hypothesis testing is computing the tast statisti, cor paring it to the critical value, and making a decision to reject or not to reject the nl hypothesis. Referring to Chart 10-1, if, based on ‘sample information, z is com to be 2.94, the null hypothesis is i sion to reject H, was made because 2.34 lies in the regi on, tat lo . region of rejection, beyond 1.68. We would reject the null hypothesis, reasoning that itis highly moro 4 tt a computed z value this large is due to ‘sampling error (chance). not tt | Computed Ig? Been 1.65 oF less, say 0.71, the null hypothesis bt i . It would be reasoned th ae om attributed to chance, that is, sampling eran n° Sal! computed valve accept oo'2ch only one of two decisions is possible in hypothesis testing tt accept i 'e null hypothesis. Instead of “accepting” the null hy i he decision as: “Do not reject Hy” "We Sis is rejected it hat the hal hyp Pt Be rected (2 Type | ene. Also, there & 2 es ud293 Before actually conduct one-tailed test of significan bas One-Tailed and Iwo-Tailed Tests of Significanc wish Refer to Chart 10-1. It dey 1 are cled the right (upper) tail of th 9 a test of hypothesis, we will differentiate between a 1ce and a two-tailed test. ects 9 greitalled test. The region of rejection fs only in ve. To illustrate, suppose that the packaging depart- amend test of a seal Foods Corporation is concerned that some boxes or arape Nuts pots. i ieee a, Overweight. The cereal is packaged in 453-gram boxes, so fhe null pothesis is H,: ». = 453. This is read, “the population mean (y) is equal to or less than 483." The alternate hypothesis is, therefore, H,: 4. > 453. This is read, “wu Is greater than 453.” Note that the inequality sign in the alternate hypothesis (>) points to the region of rejection in the upper tail. (See Chart 10-1.) Also note that the null hypothesis includes the equal sign. That is, H,: 4 = 453. The equality condition always appears in H, never in H,. Chart 10-2 portrays a situation where the rejection region is in the left (ower) tall of the standard normal distribution. As an illustration, consider the problem of auto- mobile manufacturers, large automobile leasing companies, and other organizations that purchase large quantities of ties. They want the tires to average, say, 60,000 miles of wear under normal usage. They will, therefore, reject a shipment of tires if tests reveal that the mean life of the tires is significantly below 60,000 miles. They gladly accept a shipment if the mean life is greater than 60,000 miles! They are not concerned with this possibilty, however. They are concerned only if they have sample evidence to con- clude that the tires will average less than 60,000 miles of useful Ife. Thus, the test is set up to satisfy the concern of the automobile manufacturers that the mean life of Region of rejection spat feject Hy ots 5 Scale of 2 Critical value CHART 10-2 Sampling Distribution forthe Statistic =, Let‘Tailed Test, {05 Level of Significance294 Tost is one-tailed tH, states > oF n <. IH, tates a directo, testis one-tald. LO4 Conduct a test of hypothesis about a population mean, Chapter 10 ‘Tis statement appears in the alt miles. Pitore the tires is no less than saeatneses In this case re : ie Pies, tenis. The null and alter: Hg w = 600% Hig w < 60,000, ion region i ; ‘ zation of the rejection region is to look at one way to determing loca ate hypothesis & pointing (ther lity Sig? he left, and the rejection region is therefore in in which the inequality Oe this problem itis pointing 'O © eg when the alternate hypothesis, 1, in summary, a test IS tion, such as: Hg: The mean in er year. Ber YO como of women stockbrokers Is greater than $55.00) Hy; The mean { i te hypothesis, we use a two: {ng direction is specified in the alterna i Je hoy Goalee the previous problem to istrate, we can say: fled tog, «The mean income of women stockbrokers is $65,000 per year. He The ear ncome of women stockbrokers Is Not equal to $85,000 per ay If the null hypothesis is rejected and Hy, accepted in the two-tailed case, ty income cadibe ‘significantly greater than $65,000 per year or it could be Sintene trom os 000 per year. To accommodate these two possibilties, the § pene) area of rejection is divided equally into the two tails of the sampling distributon shows the two areas and the critical values. Noe tere percent each). Chart 10-3 the normal distribution is 1.0000, found by .8500 + .0250 + .025 ic, ° e wi State come of women stockbrokers is less than or equa to gs ay Region of refection Do nat reject Hy ‘ejection 05 a | 95 1.96 0 1.96 Seale \ ot Crtial valve Critical value GHART 10-3 Regions of Nonrejection and Rejection for a Two-Tailed Test, : a jection f Testing for a Population Meai Known Population Standard Deviation A Two-Tailed Test ‘An example will show the details . also wis of the five-ste oe iso nih use a two-tailed test. Thats, we ap hwpotiedl eng reeds = Irterested Inrange of Smaller than the proposed population mean ate YO We bones matte iis clfeent om the proposed vale forte populate torical ona the previous chapter, vith asuaton In whi Mer ion about the population and in fact know its standard 48‘vvervoms 1eSt5 OF Hypathasls State. The Model A325 desk at the Fredonia Plant Weekly production of the follows a normal I Probability distribution With a mean of 200 and a standard devia- tion of 16. Recently, because of market expansion, new production methods have been introduced and new employees hired. ae ee President of manufacturing would ; investigate whether there has been a My Broduetion Of the Model A325 desk. Is the mean number of Tedonia Plant different from 200 at the .01 significance level? We use the statistical us hypothesis duction rate has changed from "200 cs Step 1: State the null hypothesis and the alternate hypothesis. The null 'ypothesis is “The population mean is 200." The alternate hypothesis is "The mean is different from 200" or “The mean is not 200." These two hypotheses are written: change in the weeki desks produced at 1g Procedure to investigate whether the pro- Per week, He: w= 200 Hy: w # 200 ‘This is a two-tailed test because the alternate hypothesis does not State a direction. In other words, it does not state whether the mean pro- duction is greater than 200 or less than 200. The vice president wants only to find out whether the production rate is different from 200. Step 2: Select the level of significance. As noted, the .01 level of significance is used. This is a, the probability of committing a Type | error, and itis the probability of rejecting a true null hypothesis. Step 3: Select the test statistic. The test statistic for a mean when's is known iz. It was discussed at length in Chapter 7. Transforming the production data to standard units (2 values) permits their use not only in this problem but also in other hypothesis-testing problems. Formula (10-1) for z is repeated below with the various letters identified. [10-1] i is lated by finding ‘i the decision rule. The decision tule is formul y See rant values ofz tom Appendix B.1. Since tis is a tvortaled et half of 01, or 008, Is placed in each tal. Te area where, nt rejected, located between the two tals, is therefore £8, Appendix 6.1 De a een ate el al, ite .4950 is NS ‘asp bo fie ne ne a r ic in the row and colur 4951. eats eee, ‘Appendix B.1, Areas under the Normal aa is repeated in the inside back cover.296 a of this problem are ‘shown in the diagram jp, Chen, - by Chapter 10 All the fe* 258 0 = Region of al Hy not rejected rejection Critical value rtica valu, GHART 10-4 Decision Rule for the .01 Significance Level ‘The decision rule is, therefore: Reject the null hypothesis and aoceyt the alternate hypothesis (which states that the population mean isnt 200) if the computed value of z is not between —2.58 and +2.68, Donat reject the null hypothesis ifz falls between —2.58 and +2.58. Step 5: Make a decision and interpret the result. Take a sample from the population (weekly production), compute z, apply the decision mus, and arrive at a decision to reject H, or not to reject H,. The mean nun- ber of desks produced last year (50 weeks, because the plant was tut down 2 weeks for vacation) is 203.5. The standard deviation of the population is 16 desks per week. Computing the z value from for mula (10-1): X=» _ 203.5 — 200 ofVn 16/V50 Because 1.55 does not fall in the rejection region, H, is not rejected. We conclude that the population mean isnot diferent for 200, So we vl report to the vice president of manufacturing that the sample evidence does ‘not show that the production rate at the Fredonia Plant has changed fm 200 per week. The difference of 3.5 units between the historical week P| duction rate and that last year can reasonably be attributed to error. This information is summarized in the following chart. = 1.55298 Chapter 10 Tailed Test A One-Talled ce emphasized that wo were concerned on In the previous example, Wr or there had been a change in the mean“ "eq ing to the vice president Fredonia Plant. We were not concemed with " Nnbe desks assembled at the Te Gecrease in the production. ‘“ change was an incre Shad fost, lat's change the Problem. Suppogy q i Pm Tents to know whether there has been an increase in the ny! Yer president bled. Can we conclude, because of the improved pro cn units at the mean number of desks assembled in the last 50 weeks yn than 200? Look at the difference in the way the problem is formulated, thet tase we wanted to know whether there WAS a difference in the mean assembled, but now we want to know whether there has been an Because we are investigating different questions, we will set our hypothag ferently. The biggest difference occurs In the alternate hypothesis, Before ‘stated the alternate hypothesis as “different from"; now we want to state" “greater than.” In symbols: % A two-tailed test: Hg: w = 200 Hy w # 200 Hy: w > 200 The critical values for a one-tailed test are different from a two-tailed tet a same significance level In the previous example, we split the signticance lau! half and put hal in the lower tail and half in the upper tail. In a one-tailed test ys put all the rejection region in one tail. See Chart 10-5. bs For the one-tailed test, the critical value is 2.33, found by: (1) subtract from .5000 and (2) finding the z value corresponding to .4900. i A one-tailed test: He: p< 200 / al Hee pein rejected recon A 2—hpl 258 Scale of z 0 233 Sito Cia value value CHART 10-5 Rejection Regions for Two-Tailed and One-Tailed ‘Tests, « -Value in Hypothesis Testin In testing a hy a ae " Re en we Compare the test statistic to a critical value. A é ot the null hypothesis or not to reject it. So, for example det the sion is to reject the nul ume value of the test statistic is 2.19, In rece : mation is ofan en cnutTed by the availablity of computer software, addition) eS Fo conoan chord on the strength of the rejection er acceptance: Mis @ in rejecting the null hypothesis? This approach rePaFR [Une-DUMMPTE 18ST Of Hypothesis, 2» probability (assuming that the statistic at least as ext ull hypothesis ist the probability, called the acs YAU actually ose seg & value ofthe test than the significance leva}, yo" peal 7 aH 18 level. the pi Hy ls ot rejected” lo . IF it is larger than the Sciamrente are P-VALUE The probability of observ; extreme than, the rar’, 1 Sb8erving a samplo value as exveme as, or more f Ne observed, given that the null hypothesis is true, Determining the p-valu m le i us additional insight into the jhehel Tenia in a decision regarding H,, but it gives 0001, indicates that thers nat? of the decision. very small p-élue, such as p-valve of 2039 means tas it tkelcod the Hi te. On the other hand, a is false. o |S not rejected, and there is little likelihood that it wus Eke cr ppute the p-value? To itustrate we will use the example in which Bt Fredonia aca fyPothesis thatthe mean number of desks produced per week SE ere tes 200, We dl not reject the null hypothesis, because the z value of aes 'e region between —2.58 and 2.58. We agreed-not to reject the null hypothest ne Computed value of z fell in this region. The probability of finding ick many | M 55 or more is .0606, found by .5000 — .4394. To put it another way, ie | {he probability of obtaining an X greater than 208.5 Hf. 200s 0606, To com. | edi taking the pute the p-value, we need to be concerned with the region less than —1.55 as well cies | the values greater than 1.55 (because the rejection region is in both tails). The voces | ‘WOtailed p-value is 1212, found by 2.0606). The p-value of .1212 is greater than | Fepthcaewpit| 17 significance level of 01 decided upon inital, s0 H, is not rejected, The detalls statistically sige are shown in the following graph. In general, the area is doubled in a two-sided test. wee settnot pre. | Ten the p-value can easily be compared with the significance level. The same dect- | apipuiant, | sion rule is used as inthe one-sided test Nees | ae a Rejection region ' or | i ed 155 258 Scale ofz the liketinood that His false. But how do we A ean ave? We 1 eteready said that if the p-value is less than the signif interpret a pve reject Hf fis greater than the significance level, then we Keance eve ais, f the p-value fs very large, then ti kely that Hy is true. i Bo neg nal then i ikely that His not true. The folowing box wil help ep L to interpret p-values. FITERPRETING THE WEIGHT OF EVIDENCE AGANST Ifthe p-value is less than (a) 10, we have some evidence tat Hy isnot tue 8 ‘05, we have strong evidence that Hy not tue. ee have very strong evidence that H, is not true. @ sat we nave extremely strong evidence that His not tue,ce fs changed to read: Does ths ey last sentence ethan 16 ounces? State the ma se condltions. sed ism under athestor the new conditions stated in part (a)? ‘containers re vealed the mean to Be 16.049 oy nce, o © oi yet in a single 22 te tata at What it decision regarding the nul ie 7 What You! ne ip ues Bed What is the p-value’ r rralue? Is tis the same conclusion Fo ; Exercises vestons (6h connect es 1-4 anawer the : Fe oc ret tai A el or at ing 14,2 (e) What is the p-value? Interpret it ble. Hg w= 80 Hep #50 ‘The sample mean is 49, and the sample size is 86. The population standard devon the .05 significance level. 10 ZA sample mean is 12 for a sample of 36. The population standard deviation i 2 .02 significance level. Me i sample of 36 observations is selected from a normal population. The sample mex 21, and the population standard deviation is 5. Conduct the following test of hypotess Using the .05 significance level Hy: ws 20 a Hew >20 ‘A sample of 64 observations is selected from a normal population. The sample 216, andthe population standard deviation 1s 15. Cond the following texto Np esis using the .03 significance level. : Hy: w= 220 Hy: we <220 For Exerci : Ser Sees & 6: (9 Sit te nl pothesis and the aerate hypothess (b) Sia fe Fer nae Cree val of he test statisti. () What fs your dein rege 5. The Tha arc of the X-15 steel-belted radial truck tire claims that the mean miea® dead devin, ven before the tread wears out is 60,000 miles. The populations fad fat the fens — is 5,000 miles. Crosset Truck Company bought 48 ties ad ferent from that clai oh ge for its trucks is 59,500 miles. Is Crosset's experience dt & The MscSisties eben the manufacturer at the .05 significance level? ae Wend eae Chain claims that the mean waiting time of customers is 31 (ean Population standard deviation of 1 minute. The qualty-2ssurance Cs at ing time was 2.75 minutes At the oe ce laren Road MacBurger that the mitt be 7, ating te less than 3 minutos? significance level, can we conclude thatthe m=" A recent nati 6.8 DVDs por mont, (2und that high school students watched an average (0a with @ population standard deviation of 0.5 hour. A random PFOne-Sampe Tests of Hypothesis 98 college students revealed 301 6.2. At the .05 signitcance we wer DVDS -05 significance level, can 2 MER neers a en pe NDS waked at month was wa ae 8 re a “You can avera = tion of the population distribute: 1 restaurant. the i level, "oa Grumney Family Restaurant, Beth Brigden ‘ . Oda ny ution is $3.34 07, tips.” Assume the stand il em r jard devia- aly amount of hey te yet, YS she Was employed at she is ean IPS was $84.85. At the .01 signif .85. At the 01 slgniftoanc Ting an average of mote than $80 ints? 1@ preceding exam ple, we ki cases, however, the 'e knew a, the population stand: based on prior singse ulation standard deviation is eusown. Thm. ibe ulation standard deviation in the folowing sea ee cain, the pope standard deviation is used to caliaee” Sample not kaon. 20 the, saree fo find the value of the t ¢ mula [10-1] as follows: test statistic we use the ¢ distribution and revise for- with n—1 degrees of freedom, where: X is the sample mean. vis the hypothesized population mean. 's is the sample standard deviation. nis the number of observations in the sample. We encountered this same situation when constructing confidence intervals in the previous chapter. See pages 262-268 in Chapter 9. We summarized this probier in Part 9-3 on page 265. Under these conditions the correct statistical procedure is to replace the standard normal distribution with the f distribution. To review, the major characteristics of the ¢ distribution are 4. It is a continuous distribution. 2) It is bell-shaped and symmetrical. 3. There is a famiy Of t distributions. Each time the degrees of freedom change, ibution is created : 4. fs the Gictrper of degrees of freedom increases the shape of the t distribution \dard normal distribution. proaches that of the stan 5. The ¢ distribution is flatter, or more spread out, distribution. than the standard normal The following example shows the detalls. ost to Company Claims Depariment reports the mean The MoFarland Insurance Compal ie eat at an industry comparison showed this amount Pe se process a claim is $60, anies, so the company insti y than most other insurance COMPANES. tes Supervisor of the cost-cutting measures, 3 effect of the 1996 claims processed last month. sures. To evaluate ‘the Sais Department selected a rancor sare 2 The sample information is reported ‘on the302 ‘Chapter 10 * om $5 340 a 8 58 4 0 54 a 58 51 a Scr ehere pa «it reasonable to conclude that mean cost tg = is ‘At the .01 significance level ‘a claim is now less than $60? ‘We wit use the fve-step hypothesis testing procedure, othesis and the alternate hypothesis, reece andes Population mean is at least $60. The mena pothesis atthe population mean is less than $60. We can g Popol and alternate hypotheses as follows: res, Hy» = $60 Hy» < $60 ‘The test is one-tailed because we want to determine whether there been a reduction in the cost. The inequality in the alternate hypothe points to the region of rejection in the left tail ofthe distribution, Step 2: Select the level of significance. We decided on the .01 significance Step 3: Select the test statistic. The test statistic in this situation is thet tion, Why? Firsts reasonable to conclude that the distribution of te, per claim follows the normal distribution. We can confirm this from teh: togram in the center ofthe following MINITAB output. Observe the distribution superimposed on the frequency distribution, are ATCT) Step 1: ‘Summary for Cost re 8 e 2 a et 2 3 o * sled 8 pos a = 6 220 Cee tras a ee fom We do not know th a tute the sample stance naa" deviation of the population. So we subst le standard atic Pisted by formula (10-2) deviation. The value of the test statistic is comF ‘one-Sample Tests of Hypothesis 303 Step 4: Formulate the deci B2,aportion 'on rule. The critical A inthe pent of wich s shown in “able Lae in Appendix “af” for degrees of re ofthe text Tho are coum of he ae abated umber of obescte ene number of a bs aber ations in the of degrees of freedom is the total sarmpled, Witien f= Thi hoe eee ees epee of Ree le 6 26 ord ar his case the number “of obsewatons inthe sar Or readacr st Sampled 1 popuation, so there ae 26 — 1 = 25 degrees ate degrees of fears value, first locate the row with the appropri- Satara ot teedom. This row is shaded in Table 10-4. Nex, determine felled tat 60 Odd Re ee ee rear have a one- the column with the eed of the table that is labeled “one-tailed.” Locate cance level is .01. Re ey significance level. In this example, the signifi- ‘the row with 25 /e down the column labeled “0.01” until it intersects one-aideditest Aaj iepitieninly ae is 2.485, Because this is a negative. The ry . is in the left tail, the critical value is decision rule is to reject Hy ifthe value of tis less than —2.485. TABLE 10-1. A Portion of the t Distribution Table Confidence Intervals 80% som | 95% | 98% 99% 99.9% Love of Significance for Oneal Test a | om | 00s | ozs | om | coos | 00m _ Level of Significance for Two-Tailed Test, « oa | om | 00s | oo | oo | oom f | agaa | ur | ammo | 2@ |] aan | sate m | tan | im | 20m | 2x8 | 2h 3792 | ie | ima | 2a | 2600 | 2am | ame me Tre [it | 20 | 24a | 278 | 3mm 5 1.316 4.708, | 2,060 2787 3725 sais | 170s | 2050 | 24a | 2770 3.707 Tare |= 1908 |» 2082 es BABE pa 3.600 tan | gee | 2a | 27 | ae 6 a 2 | 1318 2 | 13 tao | 205 | 2am | 2756 | 368 | tao | toa | 208 ‘ast >| 2.750__| 3646 and interpret the result. From the MINITAB outpHt to anterfne mean cost per clam for the sample of 26 Tag standard deviation ofthis sample is $10.04. formula (10-2) and compute the value off ‘Step 5: Make 2 decision the right of the histogt ‘observations is $56.42. ‘We insert thes values in _ X= _ $5642 = $60 _ 1.918 t= Sin” S1004/V25 of the critical value of cause ~1.818 les in the region to Me right it Beodas, the null hypothesis is not ejected atthe 01 signifioates eva ‘We have not demonstrated that ‘the cost-cutting measures pan yer claim to less than $60. To put itanother' ey putin Ter gase Se oamen te remands Oe uid be due to sampling ero” Te computed va aad Inchart 10-6. Itis in the region | cher the null Rypotnesis IS ,Chapter 10 2485 1.818 Critical Computed value value of ¢ CHART 10-6 Rejection Region, ¢ Distribution, .01 Significance Level \dard deviation were computed ious example the mean and the stan at si MINEED. the following example shows the details when the sample Mean a sample standard deviation are calculated from sample data, The rriean length of a small counterbalance bar is 43 millimeters. The production Merete ‘ncerned that the adjustments of the machine Producing the bars have changed. He asks the Engineering Department to investigate. Engineering selects a random sample of 12 bars and measures each. The results are reported below in millimeters. a2 39 42 4 43 40 «39 41a, \s it reasonable to conclude that there has been a change in the mean length ofthe bars? Use the .02 significance level. We begin by stating the null hypothesis and the alternate hypothesis. Hg: p= 43 Hii # 43 ‘he alternate hypothesis does not state a direction, so this is a two-tailed test. There are 11 degrees of freedom, found by n ~ 1 = 12 — 1 = 11. The t value is 2.718, found by referring to Appendix B.2 for a two-tailed test, using the .02 significance level with 1 desrees of freedom. The decision rule is: Reject the null ypothess is to the left of ~2.718 i informati herenea HB w tele oF to the right of 2.718. This information Region of rejectic a g=0 2.718 Scale of t Critical value value SHART 10-7 Regions of Rejection, Two-Tuiled ‘Test, Students ¢ Distribution, a = 2e : ~ 415-430 ae s/Vn~ “.7ea/vig ~ ~2913 null hypothesis that the Population mean is 43 mil llimeters jecte because the computed t of -2.913 lies in the area to the left of 2.718. Wie ceca the alternate hypothesis and conclude that the Population mean is not 43 millime- ters. The machine is out of control and needs adjustment, The mean life of a battery used in a digital clock is 305 days. The lves ofthe batteries follow the normal distribution. The battery was recently modified to last longer. A sample of 20 of the modified batteries had a mean life of 311 days with a standard deviation of 12 days. Did the modification increase the mean lif of the battery? (a) State the null hypothesis and the alternate hypothesis. (b) Show the decision rule graphically. Use the .05 significance level. Compute the value of t. What is your decision regarding the null hypothesis? Briefly summarize your results. xercises CONNECE 2 iver the foowinghyptess: wst0 rie Fora random sample of 10 observation, the sample mean was 1 ands stan ‘eid devon 3 Us te.05 sence ee Site ne ison les Comte sein aang te nl ypohess? jven the following hypothesis: y= 400 jn #400 n random sample of 12 observations, the sample méan wes 407 and the sam Edad devavon 6: Using tho 0 sicacelevek a. Stato the decision res te the value of the test statistic. 3 Cem ur Gein regacing the il hypaess?306 Chaptor 10 ager of Rath Publishing, Ne, a college ait alos mane raves ako aN AVTaGO of 40 aan ki, " ishing company, iin ales For tis estimate is 100 IOW. To investigate. g°Al py, Several reps $Y "2 that the mean number of calls made jce"On waves a cample Is 2.1 cals. Using the 05 signiteanes Mk dard deviation of I calls per salesperson Per week is mee A ar celta isis is consiionng Oe Imothod of asembing 6 fa. The management of A equires 42.3 minutes, On the average, to assemble gay f6 cart. The method Trandom sample of 24 carts, Using the new method, ya minut ae andar ‘Geviation of the sample was 2.7 roa Using the ig fs rite acca coche ATS rsan teh ocr oe ‘ % jimed that its plugs have - 22,1 Aisa gers cs oe vt Se A foc Finn eof Se US Sf x raed te re dean tne standard deviation was, 1500 ie, TEP SnCu etd stant vanufacturer's claim at the .05 signifi sa ha alee te a oe Foy i ‘ticket, and it costs the airline compar le than so pipe ea cee cma on fa sa, our ent ese te cocing a ea coe a Inn er et, pata errant wate Sey w fandom sample of 20 airports and collected information on the number of comply the airport had with e-tickets for ‘the month of March. The information is reporteg ring [4 “woe 2 2 w 1% 1% oe We ety te i 8 EY bed aS ee Ree {At the .05 significance level can the watchdog agency conclude the mean number of Complaints per airport is less than 15 per month? ‘2, What assumption is necessary before conducting a test of hypothesis? B. Plot the number of complaints per airport in a frequency distribution or a dot plot is reasonable to conclude that the population follows a normal distribution? ‘e. Conduct a test of hypothesis and interpret the results. A Software Solution ‘The MINITAB statistical software system, used in earlier chapters and the previous section, provides an efficient way of conducting a one-sample test of hypothesis for ‘a population mean. The steps to generate the following output are shown in the Software Commands section at the end of the chapter. —— ew200e e207 aa eteou to mitea, pests PL oe be. ‘One-Sample : Length (0.36%, B.S) 2.91 etd An additional feature of most statistical soft _Ana al f fare packay to report the p-vah® wh ahs sectional information on the null oh se the probe Using tha dene &s extreme as that computed, given that the nul hypothesis te Using the data fom the previous counterbalance bar example, tho peaue of 0148 elkelIhood of at value of ~2.91 or less plus the ikelitond of a ¢ value of 291‘une-sample Tests of Hy 307 larger, given a popui Of 43, eal er the null hypothestt®: ‘Comparing the p-value to the significance Fe etian the: war inn 28 close to being rejected, barely rejected, or shaded are; ‘ ted Deval of ‘O14 ie eee licens lover ent, Te talus of O14 eth darker Tejected. Had the p-vaiug at lanicance lacy seated, 68, Because the 57 the null hypothesis woury «3 than the signiicancs level”-say "00, te oy been selected as .01, the mal typethoss ee the signicance level hd intaly Ould not be rejected. 2918 Scale oft ang In the i rejection areas in ‘Doth tee the altemate hypothesis was two-sided, so there were p-value, it was necesean: eee tail and the upper (right tall, To determine the bution with 11 degr sSary to determine the area to the left of —2.913 for a t distri- eect legrees of freedom and add to it the value of the area to the right What if we were conducting a one-sided test, so that ‘the entir tc would be in either the upper or the lower tal? in that case, we would por the sea from only the one tail. In the counterbalance example, if H, were stated as y. < 43, the inequality would point to the left, Thus, we would have reported the p-value as the area to the left of -2.913. This value is .007, found by .014/2. Thus, the p-value for a one-tailed test would be .007. How can we estimate a p-value without computer software? To illustrate, recall that, in the example regarding the length of a counterbalance, we rejected the nul hypothesis that ». = 43 and accepted the alternate hypothesis that . # 43, The sig- nificance level was .02, so logically the p-value is less than .02. To estimate the p-value more accurately, go to Appendix B.2, and find the row with 11 degrees of freedom. The computed t value of 2.913 is between 2.718 and 3.106. (A portion of Appendix B.2 is reproduced as Table 10-3.) The two-tailed significance level TABLE 10-3 A Portion of Student's ¢ Distribution ‘Confidence Intervals som | ore | oom | oom | oo | mon | Level of Significance for One-Tailed Test, a | oan | 005 | 002 | amr | 0005 | oom i Level of Significance for Two-Taled Test, [ez [0x0 aos | 002 | oot | ooo Li : : a250 | 4781 yge3 | 1.893 | 2262 | 2821 ao | ug | ter | 22a | 274 | sta | 4st 1.363 1.796 2201 278 | 3.106] | 4437 ADf es taeo |. tae, | -2t70 [zeit | Sisal. eo] aa | azn | zim | 2680 soz | Az as | 17et | 2 wigs | ae | Be | am | ame | oe308 Chapter 10 it is .01. Therefore, the p-value ig , and for 3.106 value i beta, comeaponding 0 2.71818 9 O° 6 repor that the P Sraliets Ese ae eRe 01 and 02, The USt oSo we would report “the P een the two significance levels. jedicine. A sample of ei tie with 2.0 grams of Ti att bo. Selb Review 10-4 A machine ie soto fil a small BAN Tin Gach bottle tles revealed the following 2" SE A BSS BBB ces ide that the mean weight is 165s than 9.0 grap, i .ce level, can we conclu =e 3 ae hypotes an the aerate Hypothes (b) How many degrees ‘of freedom are there’ a eae vae tC What is your decision regarding the null hypothesis (©). Estimate the p-value. a0 Exercises 18. Given the following hypothesis: ‘A random sample of five resulted in the following values: 18, 15, 12, 19, and 21. Using the .01 significance level, can we conclude the population mean is less than 20? a. State the decision rule. 'b. Compute the value of the test statisti. ©. What is your decision regarding the null hypothesis? 4. Estimate the p-value, 16. Given the following hypothesis: Hy w= 100 Hew # 100 A random sample of six resulted in the following values: 118, 105, 112, 119, 105, at! 111, Using the .0S significance level, can we conclude the mean is different from 100? @. State the decision rule. Compute the vale ofthe test statistic, > What is your decisi eatmate he eral” Tegaraing the null hypothesis? 17. Experience raising New Jers at five months is 4.35 poun¢ increase their weight, a sp Weights of a sample of fiv ey Red chickens revealed the mean weight of the chickers is. The weights follow the normal distribution. In an ecial additive is added to the chicken feed. The subsecue™ ‘e-month-old chickens were (in pounds): [AN 437433435430 499 agp ams 8 : ; exes Estimate the p-value, "> SP&lal additive increased the mean weight of the otic the (80 days), As ‘was added © chlorine to find whether ). AS an experiment, Holdlonger tet had these shelf ives (fy ia) 'ncrease the shelf life. A sample of nine jug8 2159 2170 2189 21702160 ayer aan anon 28 te At the .025 level, has nate p-value, ololonger increased the sheit life of the chlorine? Etos Conducta testoF ss about @ we proportion, ‘ne and a(t — 1) must beatleats, One-Sample Tests of Hypothoss 309 19. Wyoming fisheries day of fly-fishing onthe Se that the moan number of cutthroat trout caught during a full area Is 4.0, To make hoe eee ‘and other rivers and streams in the Jackson Hole fshermen to keep a count Spat UPdate the fishery personnel asked a sample of ty 4,3, 2, 6, 8, 7, 1, 9, 3, 4, and 6 meet Caught during the day. The numbers were: 4, ber caught Is greater than wna 6-At the .05 level, can we conclude that the mean num 20. Hugger Pols Gontends theta tats the pal very Wook, A seenca et 89 agent conducts a mean of 69 in-depth home surveys uate its effectiveness, The mene om has been intreduced, and Hugger wants to eval random sample of agents en” ©! -depth surveys conducted during a week by @ 53 57 50 55 88 54 oy 52 se 2 60 60 SI 59 56 At the 05 level of signi anes, can we conc i conduct conclude that the mean number of Interviews lucted by the agents is more than 83 per week? Estimate the p-value. Tests Concerning Proportions init revo chapter we discussed confidence intervals for proportions. We can Fie luct a test of hypothesis for a proportion, Recall that a proportion is the ratio ithe: number of successes fo the number of observations We let X refer tothe num ber of successes and the rumber of observations, so the proportion of successes *d number of trials is X/n. Thus, the formula for computing a sample propor- tion, p, is p = X/n. Consider the following potential hypothesis-testing situations. * Historically, Wells Fargo Bank reports that 70 percent of leased vehicles are returned with less than 36,000 miles. A recent sample of 200 vehicles returned at the end of their lease showed 158 had less than 36,000 miles. Has the pro- portion increased? ‘+ The American Association of Retired Persons (AARP) reports that 60 percent of retired people under the age of 65 would return to work on a full-time basis if a suitable job were available. A sample of 500 retitees under 65 revealed 315 would return to work. Can we conclude that more than 60 percent would return to work? + Able Moving and Storage, Inc., advises its clients for long-distance residential moves that their household goods will be delivered in 3 to 5 days from the time they are picked up. Able's records show it is successful 90 percent of the time with this claim. A recent audit revealed it was successful 190 times out of 200. Can the company conclude its success rate has increased? ‘Some assumptions must be made and conditions met before testing a population proportion. To test a hypothesis about a population proportion, a random sample is ehesen from the population. It is assumed that the binomial assumptions discussed in Chapter 6 are met. To review, these conditions are: (1) the sample data collected lave the result of counts; (2) the outcome of an experiment is classified into one of wo mutually exclusive categories—a “success” or a “failure”; (8) the probability of a guacess is the same for each trial; and (4) the trials are independent, meaning the Gutcome of one trial does not affect the outcome of any other tial. The test we will conduct shortly is appropriate when both nm and n(i ~ 7) are at least 5. nis the Semple size, and 7 is the population proportion. It takes advantage ofthe fact that S tiromial distribution can be approximated by the normal distribution tions ir it didate ‘suppose prior elections in a certain state indicated it is necessary for a can forgoveror “o receive at least 80 percent of the vote in the northern section of the 10" Bt be elected. The incumbent governor is interested in assessing his chances stale tering fo office and plans to conduct a survey of 2,000 registered voters in the northern section of the state, : Using the hypothesis-testing procedure, assess the governor's chances of reelection.310 Chapter 10 % 1's reelection meets the binomia) jing the governor 1 ara eto 1 ese “outcomes. That IS, @ sampled Voter vant, + There are ith 5 the govermor. 4 same for each trial. In this oy + ip Weal a ses ‘support reelection is .89,. “8° the ee hood 3 Pa ependent. This means, for example, the lklnoed the ae si ‘The tha suppor reelection is not affected by what © 24th oF 2nd ot sampled wil ‘data is the result of counts. We are going to count the we + Tne eaiho support reelection in the sample of 2,000. fs cen use the normal approximation 10 the bine tbutn, gn. Shasta 7 because Both rv and nt ~ 2) exceed 6. Inti ease, n = pat cher (isthe proportion of the vote in the northern part of the state," se ‘cent, needed to be elected). ‘ Thus: nx = 2,000(80) = 1,600 and n(1 ~ =) = 2,000 (1 ~ is Both 1;600 and 400 are greater than 5. Stop 1: State the null hypothesis and the alternate hypothesis, BYE) hypothesis, H,, is that the population proportion = is .80 or The ny alternate hypothesis, H,, is that the proportion is less than ape $ . Fre ractical standpoint, the incumbent governor is concerned o the proportion is less than .80. If itis equal to or greater tee”, A will have no probiem; that i, the sample data would incoat pl probably be reelected. These hypotheses are written symbolicaly H, states adrection. Thus, as noted previously the testis one the inequality sign pointing to the tall ofthe distribution containing region of rejection. “I Step 2: Select the level of significance. The level of signticance is.05. Theis the likelihood that a true hypothesis will be rejected. Step 3: Select the test statistic. zis the appropriate statistic, found where: tz is the population Proportion, is the sample proportion, ‘nis the sample size, Step 4: Formulate the decision rule, The critical value or valu rm the de viing Point or points between the regions where is ania 7 Ae rejected. Because the alteriate hypothesis states a decton, tsi felt sdene a f65t. The sign of the inequality points to the left, so on) He eft side of the curve is used. (See Chart 10-8.) The significance level wes Given as .05 in step 2. This Probability is in the left tail and determines te etween zero and the critical value is 500, es ae ston and interpret the result. Select a sample and mae? Dat of the stato sero SUEY oF 2,000 potential voters inthe ne state revealed that 1,560 Planned to vote for the incur!7 0 8 MYpathesis 3u CHART 10-8 Rejection Region for the .05 Level of Signifi esp cance, One:Tailed Test Rare. ane Proportion of .775 (found by 1,850/2,000) close ieee e that the difference is due to Sampling error? In Pp 778, the tion ir governor ® Proportion in the sample who plan to vote for the 17 is 2,000, the number of voters si 5 surveyed, 1 is .80, the hypothesized population proportion, 2 is a normally distributed test statisti ie when t true and the other assumptions are true. eee Using formula (10-3) and computing z gives. ‘The computed value of z (~2.80)is in the rejection region, so the null hypothesis is rejected at the .05 level. The difference of 2.5 percentage points between the sample percent (77.5 percent) and the hypothesized population percent in the northern part of the state necessary to carry the state (80 percent) is statistically significant. itis probably not due to sam- pling variation. To put it another way, the evidence at this point does not Support the claim that the incumbent governor will return to the gover- rnor's mansion for another four years. “The p-value is the probability of finding az value less than ~2.80. From Appendix B.1, the probabilty of az value between zero and ~2.80 ig 4974, So the p-value is .0026, found by .6000 — .4974. The governor ‘cannot be confident of reelection because the p-value is less than the significance level. ‘eho 0 Aco wus erin pe rr ae Te oer ay ‘too large. A sample ‘of 200 traffic accidents this year showed 74 teens were re too large nother accident within the last five years, Use the Ot signlfeance (a) an wo use z 2s the test static? Tel why or why not (b) State the ‘null hypothesis: and a ‘hypothesis. sete te 6 poet eye of zand ‘state your decision regarding the null hypothesis. (e) Determine and interpret the p-value. *