Anda di halaman 1dari 8
154 Chapter 4 Probabitly survey, respondents were asked to 15. CD) = 0.450, hire PCD) is 16, Flirting Survey In a Microsoft Instant Mesaging survey, resp ‘rth dra eg choo theo fan way fi an i found that F() = 0.55, where D is day in soonest 80 person If someone is andormly slected, what dos P(D) represent and what ists vlue inp extr m web 38, Sovetychecosint When tear bene» shri chest cond 14 PC) dete peyote by the Dutchess County Sherif Deparment, he sw tht 676 drives were cron and fnotinr ool gtht Wate 6 were ated for driving while intoricated. Based on those results, we can estimate chat iat nad #1) PUL) = 0.00888, where / denotes the even of reening driver and getting someone who is {99112 = 099 (dd) Intoxicated. What does P(2) denote, and whac ists value? @ in Exercises 17-20, use the drug sereening data given in Table 41, which is inclided swith the Chapter Problem. mt 17. Drug Screening If one ofthe est subjects is randomly selected, ind the probable chat the subjecchad positive tes tesul ora negative est res 1.0994 18, Drug Screening If one ofthe et subjects is randomly selected find the probaliicy chat the subjet had positive restr or does nt use drugs 19,0986 19, Drug Screening If one ofthe subjects i randomly selected ind the probabilcy shat the subject ad a negative tx esulo docs no we drugs, nono 20, Drug Screening If one of te sabjets is tandomly selected find the probabil cha the subjet hada negative et rel owe rugs. Dosage Calculations, Jn Exeries 21-26, ute the data in the accompanying tae, sehic lise the numbers of correct and wrong dasage amounts calculated by physicians. Ina research experiment, one group of physicians was given botles of epinephrine beled swith a concentration of "I milligram in I milter solution,” and another grow) of physicians was given botles labled witha ratio of“ miller of 1:1000 solu The two labels describe the exact same amount, and the physicians were instruc! 40 advinister 0.12 milligraras of epinephrine, The recults were reported in ‘The New York Times. Comect Dosage Calculation | Wrong Dosage Calculation Concentration Label 1 js (megan int iter seliton) Ratio Labet , (C1 milter of a 1:1000 2 scion) i es eh a 2113/2. 0464 Ta yolbly’s 24, Correct Dosage If one ofthe physicians is andomly sled, what sche probably of al sigh sit shal be, ftting one who cleulated the dose correc? Is that probability as high ait should be? 22.15/25 e 05%. tht potaitys 2. Wrong Dosage If one ofthe phy me 3 a scans ie randomly selected, what ithe probabiliy of tb al seeing one who calulated the done incor? Is that proabiiy slew irkechl be” B.6/hesraosn 23, Correct or Concentration If one ofthe phy eee cee wes lation or was given the bottle wt ee 24, Wrong Dosage or Ratio If Ta hss ga obi ih Sf ering ane who made ewer nee Physicians is randomly selected, find the probability canton peek dine OF B*tnB one who made a wrong dosage clusion or was given the eee ahs eo abel tmchite ve aggst 5, Which Group Did Bett ‘thet labels described os concentra ae if Sambi ees, For che physicians gven the bores labeled —-_ 4-3 Addition Rule For the physicians given the bottles labeled with a ratio, find the percentage of correct dos- agecalulations; chen express it as a probabiliey ¢.Doesit appear ha citer group did beter? What does the result sugest about drug labels? 28. Which Group Did Worse? a. For the physicians given the bots labeled with a concentration, find che percentage of wrong dosage calculations then express i asa probability. ’. For the physicians given the bores labeled with a ratio, find the percentage of wrong dos- age calculations: dhen express ic as a probability. 20 cither group did worse? What does the result suggest about drug label? «Doss it appear Survey Refusels. In Exercises 27-32, refer to the following table summarizing results from a study of »-ople who refused to answer survey questions (based on data from “I Hear You Ksocking but You Can't Come In,” by Ftegerald and Fuller, Sociological Methods and fssarchs Vol 11, No. Dn each ese, assume that one of the subjects is randomly slo ee 121 | 2220 | 2090 | 4049 | 8059 | GOandowr feapondes = 7) 5S MSHS ae Fetused 4 2 * 100 sof 4 21, Survey Refusals What is the probability that the selected person refused 10 an- swer? Does that probability value suggest tha refusals are a problem for pollsters? Why or why no? 2, Survey Rofusals A pharmaceutical company is interested in opinions of the elderly, because they are cither receiving Medicate or wil receive it soon. What isthe probability that the selected subject is someone 60 and over who responded? 28, Survey Rofusals What is the probability thatthe selected person responded or isin the 18.21 age brackev 30. Survey Refusals What isthe probability thatthe selected person refused 10 respond or isover 59 yeas of age? ‘91, Survey Refusals A markec researcher is interested in responses, especially from those be- ‘ween the ages of 22 and 39, because they are the people mor likely to make purchases. Find ‘he probability chat a selected subjet responds ors between the ages of 22 and 39. 32. Survey Refusals A market researcher is not interested in refusals or subjects below 22 ean of age or over 59, Find the probably that che selected person refused to answer ori 22oris older than 59. In Exercises 33-38, use these results from the “I-Panel-THC” tes for marijuana use, ‘which is provided by the company Drug Test Success: Among 143 subjects with positive ‘est results, there are 24 false positive results: among 157 negative results, there are fale negative results. (Hint: Construct a table similar to Table 4-1, which is included ‘ith the Chapter Problem.) 3% Screening for Marijuana Use How many subjects ae inl in the sud)? 4 How many mbjets did no se mariana? © What isthe probability chat a randomly selected subject did not use marijuana? 155 140.0214 057 «The hss nh be wth ts per ve don much ore Theres ges nt es Aes rasa ech wos Hon ch eed sees. 11, 196/1205 = 0108. Ys igh «ef ee essing sae thts tot oases fhe ‘nan, bcos hese who ree ny velco pra grup vith pins diferent ram tes 28, 102/105 = 0168 29, 1060/1205 = 0880 3, 398/1005 = 0297 1 Nan/ims = 0915, 32.401/1005 = 038 380.300 aim 178/300 = 0588 46 Chapter 4 Probabilty 15. 89/945, <0830. 1, 2, XSORT Gender Selection Miao’ XSORT gedessonecnigu i sign sigue bev to increase the ikeihood thara baby wll ea gel. In updated result (as of cis writin) of the XSORT gender-election technique, 945 births consisted of 879 baby girs and 66 baby boys (based on data fiom the Genetics &¢ IVE Institue). Based on these results, what i she prob ability of a gel boen roa couple using MicroSor’s XSORT method? Does ic appear chat the technique i effective in increasing the Ukeibood that a baby willbe a girl? 16, 138/291 0 080.1, be 26, YSORT Gender Selection MicroSon"’s YSORT gender-slection technique is dcsigned ut pps eee to increase the likelihood that a baby willbe a boy. In updated results (as of this writing) from 4 test of MicroSor’s YSORT gender section technique, 291 births consisted of 239 baby boys and 52 baby gists (based on data from the Genetics & IVF Institute). Based on these resus, what isthe probability of a boy born to couple using MicroSor's YSORT method? Does i appeat tha the technique is elective in incretsng the liklthood that a baby wil be boy? 20100001 Note pebobiy of 27, Struck by Lightning Ina recent yar, 304 ofthe approximately 300,000,000 pople in tein sucks mh retron on the United States were struck by lightning. Estimate the probability that a randomly selected opengl edi ed person inthe Unite States will estrck by lightning this year. Isa golfer reasoning correctly storm. The gle sho ek hb. ihe or she is caught out in a thunderstorm and does not seek shelter from lightning because the probability of being struck i 0 small 28 404/80 = 0738; 1, 28, Mendelian Genetics When Mendel conducted his famous genetics experiments with eas, one sample of offspring consised of 428 green peas and 152 yellow peas. Based on those results, estimate the probabil of geting an offapring pea that ably close to the expected value of 3/4 claimed by Mendel? green, Is the result ccason- Using Probability to Identify Unlikely Events. In Exercises 29-36, consider «1 event to be “unlikely” ifits probability is less than or equal to 0.05. (This is equivalent 0 the same criterion commonly used in inferential statistics, but the value of 0.05 ‘not absolutely rigid, and other values such as 0.01 are sometimes wed instead.) 01/365 29. Guessing Birthdays On their fist dae, Kelly asks Mike to guess the date of her birth, hte ot inching the yea. ‘tae 8 What the probity hat Mike wl gus correct? (Ignore leap years) , Would ic be unity for him wo gues corey om hiss ey? «. Ifyou were Kelly, and Mike did guess corey on his fi shat nhs is, would you believe his aim he made nck gus or woul yo be convinced that heady knew when you were bor? 4 I Kelly aks Mike wo guess her age, and Mike's gus is oobi ears, what is he probably that Mik and Key wie cond Sas? NEN OY 192" 20.0670. tis ey feowe 90, Credit Card Purchasas In a survey, 169 z suney, 169 respondents sy that chey never use a cit Senamcsfe dey e127 uy ht ty som an 2834p hace hese Wn roe: the probably that a randomly elected person wes a cet cand frageny cule ot 31.01.0767. No, « rash isnot unliely. ' inileeninbass Sts woe 0427aohenbe Pad ge aceon ene ne Ud call phone texting ‘What does this suggest about driving? eR ‘cre fatty ism igh 32. Air Travel Fatalities One measure of air travel safety is this: e 7 facalities: Sortson tengo _prblln asng ih Een ter et The 17 in shorts distonces than his byt. billion trips? Is this comparison fair? POTS o = 4-2 Basic Concepts of Probability 38 Testing While Driving In a New York Time/CBS News pol respondents were asked ic should be leva or illegal to send a teat message while driving. Eight said that i should belegal and 804 said that i should be ileal. Whi the probabily of randomly selecting someone who believes it should be legal to text while driving Isic unlikely co randomly select someone with 34, Gell Phones While Driving In a New York Time:/CBS News poll, respondents were asked if it shoul be legal or illegal to use hand-held cell phones while driving, One hundred fony-one sid ta it should be legal, and 663 said that it should be illegal, What is the prob- abiliy of randomly selecting someone who believes i should be legal ro use a hand-held cell phone while driving’ Is i unlikely to randomly select somone with that belief? 35. Favorite Soat om a Plane Among respondents asked which i cheir favorite seat on a plane, 492 chose she window sea, 8 chose the middle seat, and 306 chose the aisle seat (based on data from U4 Tada). What isthe probability that a passenger prefers the mide seat? Is, isunlikly for passenger vo prefer the middle seat? [Fs0, why do you think the middle seat is sounpopular? 36, At the Endl of the Day In a Marist poll, respondents chose the most annoying phrases wed in conversation, Nineteen chose “at the end of the day,” 441 chose “whatever,” 235 chose “you knoy,” 103 chose “itis wha i is,” 66 chose “anyway,” and 75 were unsure. Based on these resuls, what is the probability of selecting someone who considers “a the end of the diy" co be the os: annoying phase? AC the end of the day sit unlikely to select someone with that cho'e Probability ‘oon a Sample Space. In Exercises 37-42, ue the given sample space or ‘conatruct the exwired sample space ro find the indicated probability. 37, Three Children Use this sample space listing the eight simple events that are possible hen a couple has chree children (a in Example 1): (bbb, bbg,beb, bag, gbb, gbe, gb ges Assume that boys and girl are equally likely, so chat the eight simple events are equaly likely. Find the probability chat when a couple has three children, ther is exactly one gil ‘$8. Three Children Using the same sample space and assumption from Exercise 37, find the probability char when a couple has thre children, there are exactly two gies, 39. Two Children Exercise 37 list the sample space fora couple having three children, First identify the sample space for a couple having «wo children, then find the probability of get- lg one child ofeach gender. Again assume that boys and gts are equally key. 40.Four Children Exercise 37 lists the sample space fora couple having thice children First identify che sample space for a couple having four children, then find the probability of {tting three girls and one boy (in any ord) 41, Genetics: Bye Color Each of rwo parents has the genorype brown/bluc, which consss fhe pir oF alleles ehat determine eye color, and each parent contributes one of chose alleles {©2 child, Assume chat if the child has a least one brown allele, that color will dominate and the eyes will be brown. (The actual determination of eye color is somewhat more complicated) 8 Lis the diferent posible outcomes. Assume that these outcomes ae qual ky ‘What isthe probability tha a child of these parents will have the blue/blue genotype? ‘© Whats the probability ha he child wll have brown eyes? ‘2-XLinked Genetic Disease Men have XY (or YX) chromosomes and women have XX Shromosomes. X-linked recessive genetic diseases (such as juvenile retinoschisis) occur when theteis a defective X chromosome that occurs withour a paired X chromosome that is good. nthe following, representa defective X chromosome with lowercase x, 0 child with the xY 147 ‘33. 0.00985. Nis unk ‘4. 0.175. his nt unl. 35, 10099. Yes is nay, The mile sat aks on oui vw xy cess tothe cl, do pss the ide set has pasengers on bah ses isted of on oe sie xy. ‘36, 0.0002 Yes ull, 7.3/8, 00375 8.3/8, 010375 39 tb ba ho 1/2 006. Independent events DEFINITION Exercises 59 5a 55 56 138 Gop 5 Poy nd Poy ets Then Pian) ‘Suppose that the probability of event A isthe same regarles of whether event B thas occurred; that is, suppose that P(ALB) = P(A), ‘Then we say thatthe occurrence of event A isnot dependent on the occurrence of, ‘event B or, more simply, that A and B are independent events Independent Events Two events A and B are independent events if PCAIB) = P(A) orif PBA) = PCB). P(A) then P(B|A) = P(B); and ‘The concept of independence is of particular importance in sampling. Subse- ‘quently, we will draw samples from two (or more) populations in order to compare Population means, population variances, or some other population parameters, For ‘most ofthese applications, we will select samples in such a way that the observed values in one sample are independent of the valves that appear in another sample, We call these independent samples. esi Tecigues A coins be ppd the tins. List the posible cutzomes in he following form: (esl ‘0 05 result on toss 2, eat on os 3) For the data in Exercise 5.3, assume that each outcome has lity } of occurring. Find ‘the following probabilities. reas ‘: A: Observe exactly 1 heads 1b B: Observe 1 or more heads © C Observe no heads Refer io Exercise 54, © Compute he probability ofthe complement of event A, event B, and event C 1 Determine whether events A and # are mutually exclusive. Determine the following conditional probabilities forthe events of Exercise 54 « PiAiB) o 59 5:10 sa 512 513 54 55° Cndtral Plait od dip 159 b Pac) © PBIO, Refer w Exercise 56. Are events A and B independent? Why or why not? What about A and (C7 What bout B and C? {A dics tobe rolled and we are to observe the number that falls face up, Find the probabilities Tor these events 8 A: Observe a6 1b B: Observe an even number. © C: Observe a number greater than 2. 44D: Observe an even number and a number greater than 2, Refer to Exercise 5.8, Which of the events (A, B, and Cae independent? Which ae mutually ‘exclusive? ‘Consier the following outcomes for an experiment Ome 1 2 3 4 5 Probebty 20-25 1S 1030 and 5, and lt event B consis of outcomes 4 and 5. Let event A consist of outcomes 1 Find P(A) and PCB). Find P(both A and B occur). Find P(cither A or B occur). Refer to Exercise 5.10. Does P(cither A or B occurs) = P(A) + P(B)? Why or why not? ‘Applications | stodent has to take an accounting course and an economics course next fem. Assuming that there are no schedule conflicts, describe the possible outcomes for selecting one section of the accounting course and ane of the economies course if there are four possible accounting sections and three posible economics sections “The emergency room of hospital has two backup generators, either of which can supply enough electricity for basic hospital operations. We define events A and B as follows: Event A: Generator 1 works propels Event B: Generator 2 works properly. Deserbe the following events in words © Complement of A b BA & Bither Aor B |A survey of a number of lage corporations gave the fllowing probability table fr events related tthe offering of promotion involving a transfer Manied Promton Twewaeer Onrcrer Treat Manioge Mtoe Uanoried—_Tetl Rejected 184 05551702565 Accepied 276 31451830 7435 Toul 46 37 17 sas 56 sar ss 59 160 Gg Pity Pi sis Use the probabilities to answer the following questions Whatisthe probability data professional selected at random) will accept the promatig? Reject? 'b What isthe probability tht a professional (selected at random) is part ofa two-cares marrige? A one carer marriage? An institutional investors considering large investment in two of five compenies. Suppose tha unknown tothe investor, two ofthe ve firms are on shaky ground with regard the ‘development of new prodocts. {© List the posible outcomes fortis situation, 'b Determine the probability of choosing to ofthe three firms tha are on beter ground, {€ Whats the probability of choosing one of the two firms on shaky ground! 4 Whatisthe probability of choosing both ofthe two shaky firms’? ‘A survey of workers t two manufacturing sts of firm included the fllowing question: How Cccive is manangement in responding to egiiate grievances of workers? The ress ae Shown hee, Nae Srveyd mb Responding “Poe” Stel 192 48 Ste2 248 % Let Abe the event hat a worker comes fom Site 1, and let B be the event that the responses “poor” Compute P(A), P(B),and P(A NB) Refer to Exercise 5.16 © Areevens A and B independent? 1b Find PUB|A) and PCA). Ate they equal? A large corporation has spent considerable time developing employce performance ring ‘cals to evaluate employes’ jb performance on regular bass so thal major adjustments «an be made when needed and so that employees who shouldbe considered for “iast-tack !avancement can be isolated. Keys othe atier determination ate ratings on the employee's biliy to perform to his rhe capabilities and on his or her formal training forthe od. Formal ig Mla Cpdty Wane Ute Some Exesve Low Ole Medium 05-0607 “10 High 10 1S 16 22 ‘The pobsilts fr eng lcd om a ack ar as nda for he 12 cago ‘ova ccs onli Te wig ree Ba Caco A: An mplye worst he high apy ve 8 Anemplye fs ino the igh xen oma ting ae An emplye site on formal aig md works bw igh opacy 4nd F(A), PCB. an P10) b Find P(A|B), P(A\B), and PEBIC). © Find BAUD) PCAN Chand PCB, The uty company in are metropolitan area im ‘es uit company i lage metropolitan area nds hat 705 of ts customers pay 88 © Suppose ha two customers the probability that both es are chosen at random from thelist of al customers. Wt * tomers wil pay thei ates raonhly bl in fll? 5.20 5.6 56 oc Vihes 161 ‘b What is the probability that atleast one of them wil payin ful? Refer to Exercise 5.19. A more detailed examination of the company records indicates that ‘954 ofthe cusiomers who pay one monthly bil in full will pay the next monthly il in fll, tos: aly 10 of ae who py stan th anou on omy il ay in le ex 1 Bind he rot that ese st at anom wl ay vo coneete mots b Fide probity hata cuore elect anom wl pay mite f wo cmeete ‘months in full ‘© Find dhe probability that a customer chosen at random wl pay xactty one month in fl Random Variables “The basic language of probability developed in this chapter deals with many different kinds of events. We are interested in calculating the probabilities assocised with both ‘Guanttative and qualitative events, For example, we developed techniques that could be used to calculate the probability that a person selected at random for a Nilsen survey of television viewing haits would favor the ABC nightly news program (as ‘opposed to that of CBS or NBC), These same techniques are also applicable to finding the probability that a person selected forthe Nelsen survey watches television more than 30 hours per week. “These qulitaive and quantitative events canbe clasifed as events (or outcomes) associated with qualitative and quantitative variables. For example, in the Nielsen survey, responses tothe question “Which evening television news program do you prefer: ABC, CBS, or NBC?” are observations on u qualitative arable, since the possible responses vay in kind but not in any numerical degree. Because we cannot predict with certainty what a particular person's response will be, the variable is Classified as «qualitative random variable. Other qualitative random variables that are commonly measured include political party affiaton, socioeconomic status, geographic location, and sex/ace classification. “A finite (and typically quite small) number of possible outcomes are asociated ‘with any qualitative variable. Using the methods of this ehaper, itis possible to caleulate the probabilities associated with these events “Many times the events of interest in an experiment are quantitative outcomes associated with a quantitative random variable, since the possible esponses vary in numerical magnitude. For example, in a Nielsen survey responses the question “How many hours a week do you watch television?” are observations on a quantitative random variable. Events of interest, suchas viewing television more than 30 hours per week, ate measured by this quantitative random variable. Other quantitative random Yarables include the change in earings per share of a stock over the next year, the increase in total sales over the next year, and the number of persons voting for the incumbent in an upcoming election, Again, the methods of ths chapter can be applied to calculate the probability associated with any particular event. "There are major advantages to dealing with quantitative random variables. The ‘numerical yardstick underlying a quantitative variable makes the mean and the tan- dard deviation (for instance) sensible. With qualitative random variables, there isnt