Samples, Good and Bad

Samples,Good and Bad

As discussedin Chapter'1, since 1960 s. HelenGrace Chocolates has conducted a jellybean poll to predictthe winnerof the presiden- B tial election.President and CEOJim Gracestated in an onlinearticle in October2OOO that the jellybean poll hadcorrectly picked the electionwinner for over 3 30 years.When first initiatedin 1960,the poll fore- castedJohn F. Kennedy's narrow victory over Richard Nixon.lt accuratelyforetold Ronald Reagan's victory overJimmy Carter in 1980 and George Bush's win over MichaelDukakis in 1988. InOctober 2OOO the pollshowed George W. Bush havinga 2.2o/olead over Vice PresidentAl Gore.The poll wasagain correct in pickingthe electionwinner. Thepoll has an extraodinary record, so does this imply that the poll producesgood data?By the end of this chapteryou will learnhow to assesswhether the data from the jelly beanpoll are good or bad.r

How to samplebadly For many years in Rapides Parish, Louisiana, only one company had been allowed to provide ambulance service. In 1999, the local paper, the Town Talk, asked readers to call in to offer their opinion on whether the company should keep its monopoly. Call-in polls are generally automated: call one telephone number to vote'Yes" and caII another number to vote "No." Telephone companies often charge callers to these numbers. The Town Talk got 3763 calls, which suggestsunusual interest in ambulance service. Investigation showed that 638 calls came from the ambulance company office or from the homes of its executives. Many more no doubt came from lower-Ievel employees."We've got employeeswho are concerned about this situation, their job stability and their families and maybe called more than they should have," said a company vice president. Other sources said employeeswere told to, as they say in Chicago, "vote early and often."

21 22 CHAPTER2 Samples, Good and Bad

As the Town Talk learned, it is easier to sample badly than to sample well. The paper relied on uoluntary response,allowing people to call in rather than actively selecting its own sample. The result was biased,-the sample was overweighted with people favoring the ambulance monopoly. Voluntary response samples attract people who feel strongly about the issue in question. These people, like the employees of the ambulance company, may not fairly represent the opinions of the entire population. There are other ways to sample badly. Suppose that we sell your company several crates of oranges each week. You examine a sample of oranges from each crate to determine the quality of our oranges. It is easy to inspect a few oranges from the top of each crate, but these oranges may not be representative of the entire crate. Those on the bottom are more often damaged in shipment. If we were less than honest, we might make sure that the rotten oranges are packed on the bottom with some good ones on top for you to inspect. If you sample from the top, your sample results are again biased-the sample oranges are systematically better than the population they are supposedto represent.

:EFrUl:f:.i!li:i;'i:.l.:1.:Fri.:'!liij!:irii:1':.r: ;:r: ,ii;:iii:.:i.'i;ii::':.i:.f,i: i::::ii::;:;i::i-::::i:!ril: Biasedsampling methods The design of a statistical study is biased if it systematically favors certain outcomes. Selection of whichever individuals are easiest to reach is called convenience sampling. A voluntary response sample choosesitself by responding to a gen- eral appeal. Write-in or call-in opinion polls are examples of voluntary responsesamples. Conveniencesamples and voluntary responsesamples are often biased.

EXAMPLEI Interviewingat the mall o q Squeezing the oranges on the top of the crate is one example of convenience sampling. MalI interviews are another. q Manufacturers and advertising agencies often use interviews at shopping malls to gather information about the habits of consumers and the effectivenessof ads. A sample of mall shoppers is fast and cheap. But people contacted at shopping malls are not representative of the entire U.S. population. They are richer, for example, and more likely to be teenagers or retired. Moreover, the interviewers tend to select neat, safe-looking individuals from the stream of Simplerandom samples 23

customers. Mall samples are biased: they systematically overrepresent some parts of the population (prosperous people, teenagers, and retired people) and underrepresent others. The opinions of such a convenience sample may be very different from those of the population as a whole.

EXAMPLE2 Write-inopinion polls Ann Landers once asked the readers of her advice column, "If you had it to do over again, would you have children?" She received nearly 10,000 responses, almost 70% saying "NO!" Can it be true that 70o/oof parents regret having children? Not at all. This is a voluntary response sample. People who feel strongly about an issue, particularly people with strong negative feelings, are more likely to take the trouble to respond. Ann Landers's results are strongly biased-the percentage of parents who would not have children again is much higher in her sample than in the population of all parents.

Write-in and call- in opinion polls are almost sure to lead to strong bias. In fact, only about I5o/oof the public have ever responded to a call-in poll, and these tend to be the same people who call radio talk shows. That's not a representative sample of the population as a whole. "Hey,Pops, whot wosthot letter you sent off to Ann Londersyesterday?"

Simplerandom samples In a voluntary response sample, people choose whether to respond. In a convenience sample, the interviewer makes the choice. In both cases,personal choice produces bias. The statistician's remedy is to allow impersonal chance to choose the sample. A sample chosen by chance allows neither favoritism by the sampler nor self-selection by respondents. Choosing a sample by chance attacks bias by giving aII individuals an equal chance 24 CHAPTER2 Samples, Good and Bad

to be chosen. Rich and poor, young and old, black and white, all have the same chance to be in the sample. The simplest way to use chance to select a sample is to place names in a hat (the population) and draw out a handful (the sample). This is the idea of simple random sampling.

Simplerandom sample A simple random sample (SRS) of size nconsists of nindividuals from the population chosen in such a way that every set of n individuals has an equal chance to be the sample actually selected.

An SRS not only gives each individual an equal chance to be chosen (thus avoiding bias in the choice) but also gives every possible sample an equal chance to be chosen. Drawing names from a hat does this. Write 100 names on identical slips of paper and mix them in a hat. This is a population. Now draw 10 slips, one after the other. This is an SRS, because any 10 slips have the same chance as any other 10.

2.1 Sampling my class. There arc 20 studentsin my class. They sit in assignedseats, consisting of four rows of five students each. I want to take a simple random sample consisting of 4 of the students in my class. To do this I select a single student from each row as follows. I write the numbers 1 to 5 on identical slips of paper. I mix the slips in a hat and draw one at random. I count this number of seats in from the left in the first row and select the student in this seat. For example, if the number selected is 3, I select the third student in from the left in the first row. I replace the slip in the hat, again mix the slips, and draw a new number. The student seated this many seats in from the left in the second row is selected.I repeat this processfor the remaining two rows. Every student in the class has a 1-in-5 chance of being selected when I come to their row. Thus, every student has the same chance of being selected.Is the sample a simple random sample? Explain.

Drawing names from a hat makes clear what it means to give each individual and each possible set of z individuals the same chance to be chosen. That's the idea of an SRS. Of course, drawing slips from a hat would be a bit awkward for a sample of the country's 111 million households. In practice, we use computer-generated random digits to choose samples. Many Simple random samples 27

:r+l i: 4:'i:+: l: F ii.:lii itii.: i .;',' ;:i li ii Choosean SRSin two steps Step 1: Label. Assign a numerical Iabel to every individual in the population. Be sure that all Iabels have the same number of digits if you plan to use a table ofrandom digits. Step 2: Software or table. Use random digits to select labels at random.

You can assign labels in any convenient manner, such as alphabetical order for names of people. When using a table of random digits, as long as all Iabels have the same number of digits, all individuals will have the same chance to be chosen.Use the shortest possible labels: one digit for a population of up to 10 members, two digits for 11 to 100 members, three digits for 101 to 1000 members, and so on. As standard practice, we recommend that you begin with label 1 (or 01 or 001, as needed).You can read digits from Ta- ble A in any order-across a row, down a column, and so on-because the table has no order. As standard practice, we recommend reading across rows. Sample surveys use computer software to choosean SRS, but the software just automates the steps in Example 3. The computer doesn't look in a table ofrandom digits becauseit can generate them on the spot.

EXAMPLE4 Howto choosean SRSusing software To see how software can be used to generate an SRS, go to the Research Randomizer at www.randomizer.org. Click on "Randomize Now" and fill in the boxes.You can even ask the Randomizer to arrange your sample in order. To repeat Example 3, we asked the Randomizer to generate one set of numbers with five numbers per set. We specified the number range as 1 to 30. We requested that each number remain unique and that the numbers be sorted from least to greatest. We asked to view the outputted numbers with the place markers off. See Figure 2.1. After clicking on the "RandomizeNow!" button, we obtained the digits 4,7, 14,16, and 23. This time, the sampleis the clients labeled04,07 ,14, \6,23. Theseare Anderson Construction, Bennett Hardware, Fleisch Realty, JL Records, and River City Books.

2.2 Evaluating teaching assistants. Toassess how its teaching assistantsare performing,the statistics department at a large university randomlyselects 3 of its teachingassistants each week and sendsa faculty memberto visit their classes.The current list of 20 teaching assistants is given below. Use either software (for example, the Research -

28 CHAPTER2 Samples. Good and Bad

mi -RnsETRCH '*d R;i;i;Nrrffn Instructinns T{) gtncr;llc li s!'t {r1 t;tfirk}nt nultrl^-ts. rirn;th ttttr.r r rirr Stl!'ctxltlsiitttt'!Ct riiltrci tlrlr l: T, grnau{* a *l.ri ft qdour nurnbrrs. simFll- llrttr rtt;rrtt\fl\ r,t rlUnlhft\ rlrr \(,u itll {!ul the !$nn. radicatirg: $ .illt hl g!'n!-fitc l it,:i

y' Horv may s15 s1i llrrr 1pn1 nuntltrs pr *-tl t taudom numbrrs rou r*nrlei liLc :1,,i

y' lftru marr numbi'rt \umhcrr.rngc (c.g.. l-5{)t Iiurn ;-'ou wrct in each *!

To: lr-, y' Thr nngr vithin *Ai{n !.L:;: t0u $ftrl !{tua nultrr'tr lir fill

D{r \ r,U\\ r'h c.{h nurnlr.'rlD .l xt lil \ F5 f _i- y' ltrlrthcrornu vou rr)lhilI Unlqur'l sdnl r'aLh nashtr iB -;-. a idt t0 bf uriqlr I.)o r ou n ish ttl tr1tl \'rl$r Y€! LtllrtcCreslr+l -:. y' ilutlnrttid nuntruLrs.' I j,jL l!'itcthcr rou *uukl tikc llrt rruubrrr in -:- cflch et s'll€d. citlr.-f llori.do |ou $ irh tit r ic$ rnur pld(rurrker$ ott fisrrl l-!as{ tn {;r*itr\t rl{ltpUllr'rjnttlrirt'r r' -: :J.i !r l+n1 (iRat{\l tr1 l-tasl

y' l{.t*}ourrirhtorier t,tur outFullfd fitnhr'rs

t ,,1 , l

FIGURE2.1 Usingthe ResearchRandomizer at www.randomizer.orq.

Randomizer) or Table A at line 116 to choose3 to be visited this week. Re- member to begin by labeling the teaching assistants from 01 to 20. Bautista Frazier Miller Walters Bolen Kumar Podboy Wang Clottey Lam Roy Weimer Counts Lovesky Schumacher Yu Draper Marin Tower Zhang

Canyou trust a sample?

The Town Talh, Ann Landers, and mall interviews produce samples. We can't trust results from these samples, because they are chosen in ways Simplerandom samples 25 statistical software packages have random number generators that generate random digits. Some also allow one to choosean SRS. If you don't use software, you can use a table of random digits to choose small samplesby hand.

Randomdigits A table of random digits is a long string of the digits 0, 1,2,3, 4, 5, 6,7, 8,9 with thesetwo properties: 1. Each entry in the table is equally likely to be any of the 10 digits 0 through 9. 2. The entries are independent ofeach other. That is, knowledge of one part of the table gives no information about any other part.

Table A at the back of the book is a table of random digits. You can think of Table A as the result of asking an assistant (or a computer) to mix the digits 0 to 9 in a hat, draw one, then replace the digit drawn, mix again, draw a second digit, and so on. The assistant's mixing and drawing save us the work of mixing and drawing when we need to randomize. Table A begins with the digits 19223950340575628713.To make the table easier to read, the digits appear in groups of five and in num- bered rows. The groups and rows have no meaning-the table is just a long list of randomly chosen digits. Here's Are theserandom digits reallyrandom? how to use the table to choosean SRS. Not a chance.The randomdigits in TableA were producedby a computerprogram. Computer programs do exactly what you tell EXAMPLE5 Howto choosean SRS them to do. Givethe programthe same input and it will produce Joan's small accounting firm serves 30 business exactlythe same"random" digits. clients. Joan wants to interview a sample of 5 clients Of course,clever people have de- to find ways to improve client satisfaction. To avoid visedcomputer programs that pro- likerandom bias, she choosesan SRS ofsize 5. duceoutput that looks digits.These are called"pseudo- Step 1: Label. Give each client a numerical label, randomnumbers," and that'swhat using as few digits as possible. TWo digits are needed TableA contains.Pseudo-random numberswork fine for statistical to label 30 clients, so we use labels randomizing,but they havehidden nonrandompatterns that can mess up morerefined uses. - 26 CHAPTER2 Samples, Good and Bad

It is also correct to use labels 00 to 29 or even another choice of B0 two- digit labels. Here is the list of clients, with labels attached:

01 A-l Plumbing 16 JL Records 02 Accent Printing 17 JohnsonCommodities 03 Action Sport Shop 18 Keiser Construction 04 Anderson Construction 19 Liu's ChineseRestaurant 05 Bailey Tfucking 20 MagicTan 06 Balloons,Inc. 2I PeerlessMachine 07 Bennett Hardware 22 Photo Arts 08 Best's Camera Shop 23 River City Books 09 Blue Print Specialties 24 Riverside Tavern 10 Central Ttee Service 25 Rustic Boutique 11 ClassicFlowers 26 Satellite Services 12 Computer Answers 27 ScotchWash 13 Darlene'sDolls 28 Sewer'sCenter 14 Fleisch Realty 29 Tire Specialties 15 HernandezElectronics 30 Von's Video Store

step 2: Table. Enter Table A anywhere and read two-digit groups. s.rp- pose we enter at line 130, which is

69051 64877 87774 09517 84534 06489 87207 97245

The first 10 two-digit groups in this line are

69 05 16 48 r7 87 77 40 95 77

Each two-digit group in Table A is equally rikely to be any of the 100 possiblegroups, 00, 01, 02, ..., 9g. so two-digit groups choosetwo-digit labels at random. That's just what we want. Joan used only labels 01 to 30, so we ignore all other two-digit groups. The first 5 labels between 01 and B0 that we encounter in the table choose our sample. of the first 10 labels in line 180, we ignore 5 becausethey are too high (over30). The others are 0b, 16, L7,17, and 17.The clients labeled 05, 16, and 17 go into the sample. Ignore the secondand third 1zs because that client is already in the sample. Now run your finger across line 180 (and continue to line 131 if needed) until b clients are chosen. The sample is the clients labeled 0b, 16, rr,20,1g. These are Bailey T[ucking, JL Becords, Johnson commodities, MagicTan, and Liu's Chinese Restaurant.

using the table of random digits is much quicker than drawing names from a hat. As Example 3 shows, choosing an SRS has two steps. Statistics in Summary 29 that invite bias. We have more confidence in results from an SRS, because it uses impersonal chance to avoid bias. The first question to ask of any sample is whether it was chosen at random. Opinion polls and other sample surveys carried out by people who know what they are doing use random sampling.

EXAMPLE5 AGallupPoll A Gallup PolI on sex and violence on television be- gan with the question "In your view, does the entertainment industry need to make a serious effort to significantly reduce the amount of sex and violence in its movies, television shows, and music, or don't you think they need to do this?" The press release reported that the poll found "75Yoof Americans say- Golfingat random Randomdrawings give all the same ing the entertainment industry should make a seri- chanceto be chosen,so they of- ous effort to reduce the amount of sex and violence fer a fair way to decidewho gets in its movies, television shows, and music, whtle 24% a scarcegood-like a roundof golf. say this isn't necessary." Can we trust this claim? Lots of golferswant to play the fa- Ask first how Gallup selectedits sample. Later in the mous Old Courseat St. Andrews. Scotland.A few can reservein ad- press release we read this: "These results are based vance.Most must hope that chance on telephone interviews with a randomly selectedna- favorsthem in the daily random tional sample of 1008 adults, 18 years and older, con- drawingfor teetimes. At the height ducted February 6-8, 2004." of the summerseason, only I in 6 This is a good start toward gaining our confidence. winsthe right to pay .C13O(about Gallup tells us what population it has in mind (people $260)for a round. at least 18 years old living anywhere in the United States). We know that the sample from this population was of size 1008 and, most important, that it was chosen at random. There is more to consider in assessing a poll, and we will soon discuss this, but we have at least heard the comforting words "randomly selected."

STATISTICSIN SUMMARY

We select a sample in order to get information about some population. How can we choose a sample that fairly represents the population? Convenience samples and voluntary response samples are common but do not produce trustworthy data. These sampling methods are usually biased. That is, they systematically favor some parts of the population over others in choosing the sample. The deliberate use ofchance in producing data is one ofthe big ideas of statistics. Random samples use chance to choosea sample, thus avoiding bias due to personal choice.The basic type of random sample is the simple 30 CHAPTER2 Samples, Good and Bad

random sample, which gives all samples of the same size the same chance to be the sample we actually choose.To choosean SRS by hand, use a table of random digits such as Table A in the back of the book.

CASE STUDY To participatein the jelly bean poll, describedin the Case EVALUATED Studythat openedthe chapter,one mustpurchase jelly beans at the storeor online.Use what you havelearned in thischapter to assesswhether the datacollected in the jelly beanpoll aregood or bad.Your assessment should be written so that someone who knows no statisticswill understandvour reasoning.r

CHAPTER2 EXERCISES

For Erercise 2.1, seepage 24; for Exercise 2.2, seepage 27. 2.3 Letters to the editor. You work for a local newspaperthat has recently reported on a bill that would make it easier to create charter schoolsin the state. You report to the editor that 2Ol letters have been receivedon the issue, of which 171 opposethe legislation. "I'm surprised that most of our readers opposethe bill. I thought it would be quite popular," says the editor. Are you convinced that a majority of the readers opposethe bill? How would you explain the statistical issue to the editor? 2.4 Instant opinion. The BusinessWeekonline poll can be found at the Web site indicated in the "Notes and Data Sources,"on page 34. The latest question appearson the screen,and visitors to the site can simply click buttons to vote. On March 29,2007,the questionwas "Do you think Googleis too powerful?"In all, 1336 (35.9o/o)said 'Yes," 2O5I (55.L%)said "No," and 335 (9.0%)said "I'm not sure." (a) What is the sample size for this poll? (b) At the Web site, BusinessWeekincludes the following statement about its online poll. "Note: Theseare surveys,not scientificpolls." Explain why the poll may give unreliable information. (c) Just abovethe poII question was the following statement: "Google'saccel- erating lead in searchand its movesinto software and traditional advertising are sparking a backlash among rivals." How might this statement affect the poll results? 2.5 Ann Landers takes a sample. Advice columnist Ann Landers onceasked her divorcedreaders whether they regretted their decisionto divorce. She received approximately 30,000 responses,about 23,000 of which came from women. Nearly 75% said they were glad they divorced, and most of them said they wished they had done it sooner.Explain why this sample is certainly biased.What is the likely direction of the bias?That is, is 75o/o probably higher or Iower than the truth about the population of aII adults who have been divorced? 2.6 We don't like one-way streets. Highway planners decided to make a main street in WestLafayette, Indiana, a one-waystreet. The LafayetteJournal Chapter2 Exercises 31

and Courier took a one-daypoll by inviting readersto call a telephonenumber to record their comments.The next day, the paper reported: Journal and courier readersouerwhelmingly prefer two-way traffic flow in west Lafayette'svllage area to one-waystreets. By nearrya T-1margin, callers to the newspdper'sExpress Yourselfopinion rine on wednesdiy complainedabout the one-waystreets that haue beenin place sinceMay. Of the 98 commentsreceiued, all but 14 said no to one-way. (a) what population do you think the newspaperwants information about? (b) Is the proportion of this population who favor one-way streets almost certainly larger or smaller than the proportion 14198in the sample?why? 2.7 Design your own bad sample. Your college wants to gather student opinion about parking for students on campus. It isn't practical to contact all students. (a) Give an exampleof a way to choosea sampleof stud.entsthat is poor practice becauseit dependson voluntary response. (b) Give an exampleof a bad way to choosea samplethat doesn'tuse voluntarv response. 2.8 .lK A call-in opinion poll. In 200b the son FranciscoBay Tlmes '),i) reported on a poll in New zealand that found that New Zealanders r',ENiewsopposed the nation's new gay-inclusivecivil-unions law by a 3-1 ratio. This poll was a call-in poll that cost g1 to participate in. The San FranciscoBay Timesarticle alsoreported that a scientificpolling organizationfound. that New Zealandersfavor the law by a margin of b6.4%to Bg.Bo/o.Explain to someone who knows no statistics why the two polls can give such widely differing results and which poll is likely to be more reliable. 2.9 Call-in versus random sample polls. A national survey of rv network news viewers found that 48%said they would believea phone-inpoll of 800,000 personsrather than a random sampleof 1000persons. Ofthe viewers, 42%said, they would believethe random sample poll. Explain to someonewho knows no statistics why the opinions of only 1000 randomly chosenrespondents are a better guide to what all peoplethink than the opinions of 800,000callers. 2.10 Choose an SRS. A firm wants to understand the attitudes of its minority managers toward its system for assessingmanagement performance. Below is a list of all the firm's managers who are members of minority groups. use TableA at line 137to choose6 to be interviewed in detail about the plrformance appraisal system.

Agarwal Dewald Huang Puri Aikens Edwaids Johnson Reynolds Alfonseca Fleming Kim Richards Baxter Fonseca Liao Rodriguez Bowman Gates Mourning Santiago Brown Goel Nunez Shen Cortez Gomez Peters Vega Cross Hernandez Pliego Watanabe 32 OHAPTER2 Samples, Good and Bad

2.11 Choose an SRS. Your classin ancient Ugaritic religion is poorly taught and the classmembers have decidedto complain to the dean.The classdecides to choose 4 of its members at random to carry the complaint. The class list appearsbelow. Choose an SRS of 4 using the table of random digits, beginning at line 140.

Burkett Howry Nixon Damon Jackson Ofiiz Embree Lowe Person Fossum Lyon Ramirez Fox Martinez Timlin Garciaparra Mendoza Varitek Giambi Millar Vaughan Gonzales Mirabelli Williams Hillenbrand Mueller Young

2.12 An election day sample. You want to choosean SRSof 20 of a city's 480 voting precincts for special voting-fraud surveillance on election day. (a) Explain clearly how you would label the 480 precincts. How many digits make up each of your labels? What is the gteatest number of precincts you could label using this number of digits? (b) Use Table A to choosethe SRS, and list the labels of the precincts you selected. Enter Table A at line 107. 2.13 Is this an SRS? A university has 30,000undergraduate and 10,000grad- uate students. A survey of student opinion concerning health care benefits for domesticpartners ofstudents selects300 ofthe 30,000undergraduate students at random and then separately selects100 ofthe 10,000graduate students at random. The 400 students chosenmake up the sample. (a) Explain why this sampling method gives each student an equal chance to be chosen. (b) Nonetheless,this is not an SRS.Why not? 2.14 How much do students pay for rent? A university's housing and residence office wants to know how much students pay per month for rent in off- campus housing. The university does not have enough on-campus housing for students and this information will be usedin a brochure about student housing. The population contains 12,304students who live in off-campushousing and have not yet graduated.The university will send a questionnaireto an SRS of 200 ofthese students, drawn from an alphabetizedlist. (a) Describehow you'would label the students in order to selectthe sample. (b) Use Table A, beginning at line t25, to select the first 5 students in the sample. 2.15 Apartment living. You are planning a report on apartment living in a college town. You decide to select three apartment complexesat random for in- depth interviews with residents. Use Table A, starting at line 117,to select a -

Chapter 2 Exercises 33 simple random sample of three of the following apartment complexes.

Abbington Village French Run Old Cape Colony Bayberry Place Glenchester Olentanry Terrace Blendon Square Grand Haven Peppermill Cabot's Mill Harvest GIen Pinecrest Carlin Manor Highland Terrace RaccoonCreek Cedar Grove Iuka Park Royal James Plaza Creekside Jefferson Chase Slate Ridge Dartmouth Place Kensington Commons Summerview Eaglecrest Lantern Square Timber Tlail Emerald Glen Laurel Lake Walnut Glen Fair Oaks Meadowbrook Williamsburg Fox and Hounds Millstream Village Yearling Green

2.16 How do random digits behave? Which of the following statementsare true of a table of random digits, and which are false? Explain your answers. (a) Each pair ofdigits has chance 1/100ofbeing 33. (b) There are exactly four 4s in each row of 40 digits. (c) The digits 99999 can never appear as a group, becausethis pattern is not random. 2.17 Ttrition for illegal immigrants. As part of a project for a political sci- ence class, a student decided to conduct an online poll asking the following question: "Do you think that illegal immigrants should be allowed to pay in- state tuition, even though they are not legal citizens of that state?" Possible responseswere 'Yes," "No," and "IJnsure." Of those who responded,81% said "No," 19%osaid'Yes," and 0% said "IJnsure." Explain why this opinion poll is almost certainly biased. 2.18 More randomization. Most sample surveys call residential telephone numbers at random. They do not, however, always ask their questions of the person who picks up the phone. Instead, they ask about the adults who live in the residence and chooseone at random to be in the sample. Why is this a good idea? 2.19 Racial profiling and traffic stops. The Denver Police Department wants to know if Hispanic residents of Denver believe that the police use racial profiling when making traffic stops. A sociologist prepares several questions about the police. The police department chooses an SRS of 200 mailing ad- dressesin predominantly Hispanic neighborhoods and sends a uniformed His- panic police officer to each address to ask the questions ofan adult living there. (a) What are the population and the sample? G) \ryhv are the results likely to be biased even though the sample is an SRS? 2.20 Random selection? Choosing at random is a "fair" way to decide who gets some scarce good,in the sensethat everyone has the same chance to win. Random choice isn't always a good idea-sometimes we don't want to treat everyonethe same, becausesome people have a better claim. In each of the f------_ I I 34 CHAPTER2 Samples,Good and Bad I { I following situations, would you support choosing at random? Give your rea- sons in each case. (a) The basketball arena has 4000 student seats, and 7000 etudents want tickets. Shall we choose4000 of the 7000 at random? (b) The list of people waiting for liver transplants is much larger than the number of available livers. Shall we let impersonal chance decide who gets a transplant? (c) During the Vietnam War, young men were chosen for army service at random, by a "drafb lottery." Is this the best way to decide who goesand who stays home?

2.21Web-based exercise. There are several voluntary response polls available on the Internet. Visit www.misterpoll.com and examine several of the current polls. What are the sample sizes in these polls? Who can vote? Is it possible to vote more than once?Do you think you can trust the results of the polls on www.misterpoll.com? Why? Z.2?Web-based exercise. Tb see how software speedsup choosing an SRS, go to the Research Randomizer at www.randomizer.org. Click on 'Bandomize Noq/'and fill in the boxes.You can even ask the Randomizer to arrange your sample in order.

NOTESAND DATASOURCES

Page 2l Case Study. A recent search Page 29 Example 5: 'Most Americans revealed no information about the Helen offended by sex and violence on television," Grace Chocolatesjelly bean poll since 2000. Gallup News Service press release by Jeffrey This may mean that the poll is no longer M. Jones, February t2, 2004; available on- conducted by Helen Grace Chocolates or line at www.gallup.corr/po1V10588/Most- it may be that the poll was inconect in Americans-Offended-Sex-Violence-Television 2004 and hence no longer a news item. .aspx. Helen Grace Chocolates has a Web site at Page 30 Exercise 2.4: The link (re- www.helengrace.com.You can visit the site quires a login) to the BusinessWeeft poll is to see if there is any information about the http://bwnt.businessweek.com/polls/bwlive- poll in 2008. results. asp?displayFormat=vote&poll jd Page 2l 'Acadian ambulance officials, =20O7OBt 423225446660. workers flood call-in poII," Baton Rouge Ad- Page 31 Exercise 2.8: "Clarffication," Son uocate,January 22, L999. Francisco Bay Tlmes, November 15, 2005.