I. Survey Methodology

Size: px

Start display at page:

Download "I. Survey Methodology"

Hortense Andrews
5 years ago
Views:

I.SurveyMethodology TheElonUniversityPollisconductedusingastratifiedrandomsampleofhouseholdswithtelephonesin thepopulationofinterest inthiscase,citizensinnorthandsouthcarolina.

1 I.SurveyMethodology TheElonUniversityPollisconductedusingastratifiedrandomsampleofhouseholdswithtelephonesin thepopulationofinterest inthiscase,citizensinnorthandsouthcarolina.thesampleoftelephone numbersforthesurveyisobtainedfromsurveysamplinginternational,llc. SelectionofHouseholds Toequalizetheprobabilityoftelephoneselection,sampletelephonenumbersaresystematically stratifiedaccordingtosubpopulationstrata(e.g.,azipcode,acounty,astate,etc.),whichyieldsa samplefromtelephoneexchangesinproportiontoeachexchange'sshareoftelephonehouseholdsin thepopulationofinterest.estimatesoftelephonehouseholdsinthepopulationofinterestare generallyobtainedfromseveraldatabases.samplesoftelephonenumbersaredistributedacrossall eligibleblocksofnumbersinproportiontothedensityoflistedhouseholdsassignedinthepopulationof interestaccordingtoaspecifiedsubpopulationstratum.upondeterminingtheprojected(orpreferred) samplesize,asamplingintervaliscalculatedbysummingthenumberoflistedresidentialnumbersin eacheligibleblockwithinthepopulationofinterestanddividingthatsumbythenumberofsampling pointsassignedtothepopulation.fromarandomstartbetweenzeroandthesamplinginterval,blocks aresystematicallyselectedinproportiontothedensityoflistedhousehold"workingblocks."ablock (alsoknownasabank)isasetofcontiguousnumbersidentifiedbythefirsttwodigitsofthelastfour digitsofatelephonenumber.aworkingblockcontainsthreeormoreworkingtelephonenumbers. Exchangesareassignedtoapopulationonthebasisofalleligibleblocksinproportiontothedensityof workingtelephonehouseholds.onceeachpopulation'sproportionoftelephonehouseholdsis determined,thenasamplinginterval,basedonthatproportion,iscalculatedandspecificexchangesand numbersarerandomlyselected.becauseexchangesandnumbersarerandomlyselectedbythe computer,unlistedaswellaslistedtelephonenumbersareincludedinthesample.thus,thesampleof telephonenumbersgeneratedforthepopulationofinterestconstitutesarandomsampleoftelephone householdsofthepopulation,stratifiedbyexchange.

2 ProceduresUsedforConductingthePoll ThesurveywasconductedSunday,April19 th throughthursday,april23 rd of2009.duringthistimecalls weremadefrom5:00pmto9:00pmest,mondaythroughthursday,andfrom1:00pmto6:00pmest, onsunday.theelonuniversitypollusescatisystemsoftware(computerassistedtelephone interviewing)intheadministrationofsurveys.foreachworkingtelephonenumberinthesample, severalattemptsweremadetoreachthehousehold.onlyindividualsinhouseholds18yearsorolder wereinterviewed;thosereachedatbusinessorworknumberswerenotinterviewed.withineach household,oneadultisgenerallyselectedbasedonwhethers/heistheoldestoryoungestadultinthe home.interviews,whichareconductedbyliveinterviewers,arecompletedwithadultsfrom householdsinthetargetpopulationasspecified.interviewsforthissurveywerecompletedwith662 adultsfromhouseholdsinnorthcarolinaandsouthcarolina.forasamplesizeof662,thereisa95 percentprobabilitythatoursurveyresultsarewithinplusorminus3.9percentagepoints(themarginof samplingerror)oftheactualpopulationdistributionforanygivenquestion.forsub samples(a subgroupselectedfromtheoverallsample),themarginoferrorishigherdependingonthesizeofthe subsample.whenweuseasubsample,weidentifytheseresultsasbeingfromasubsampleandprovide thetotalnumberofrespondentsandmarginoferrorforthatsubsample.subsamplesareusedinthis surveyandarethestatesofnorthcarolinaandsouthcarolina.whendiscussingresultsfromthese statesindependentoftheoverallsample,thesubsampleinformationisasfollows:fornorthcarolina, interviewswerecompletedwith356adultsfromhouseholdsinnorthcarolina,and,forthissubsample, thereisa95percentprobabilitythatoursurveyresultsarewithinplusorminus5.3percentagepoints (themarginofsamplingerror)oftheactualpopulationdistributionforanygivenquestion;forsouth Carolina,interviewswerecompletedwith305adultsfromhouseholdsinSouthCarolina,and,forthis subsample,thereisa95percentprobabilitythatoursurveyresultsarewithinplusorminus5.7 percentagepoints(themarginofsamplingerror)oftheactualpopulationdistributionforanygiven question.inreportingourresults,wenoteanyuseofasubsamplewhereapplicable.becauseour surveysarebasedonprobabilitysampling,thereareavarietyoffactorsthatpreventtheseresultsfrom beingperfect,completedepictionsofthepopulation;theforemostexampleisthatofmarginof samplingerror(asnotedabove).withallprobabilitysamples,therearetheoreticalandpractical difficultiesestimatingpopulationcharacteristics(orparameters).thus,whileeffortsaremadeto reduceorlessensuchthreats,samplingerroraswellasothersourcesoferror whilenotallinclusive, examplesofothererroreffectsarenon responserates,questionordereffects,questionwording effects,etc. arepresentinsurveysderivedfromprobabilitysamples. QuestionsandQuestionOrder TheElonUniversityPollprovidesthequestionsaswordedandtheorderinwhichthesequestionsare administered(torespondents).conspicuousinreviewingsomequestionsisthe bracketed information.informationcontainedwithinbrackets([])denotesresponseoptionsasprovidedinthe question;thisbracketedinformationisrotatedperquestiontoensurethatrespondentsdonotreceivea setorderofresponseoptionspresentedtothem,whichalsomaintainsquestionconstructionintegrity byavoidingrespondentacquiescencebasedonquestioncomposition.similarly,toprotectagainst

3 questionordereffects(wherespecificquestionsmay prime orinfluencearespondent sanswersto subsequentquestions),questionorderisrandomizedduringadministration.forgroupsofquestionsin whichastandard,setorderofadministrationcaninfluenceresponsestosubsequentquestions, questionsareadministeredrandomlyduringthesurvey;thistechniqueprotectsagainstquestionorder effects.somequestionsusedaprobemaneuvertodeterminearespondent sintensityofperspective. Probetechniquesusedinthisquestionnairemainlyconsistofaskingarespondentiftheirresponseis moreintensethaninitiallyprovided.forexample,uponindicatingwhethers/heissatisfiedor dissatisfied,weaskedtherespondent wouldyousayyouarevery satisfied / dissatisfied? This techniqueisemployedinsomequestionsasopposedtospecifyingthefullrangeofchoicesinthe question.thoughspecifyingthefullrangeofoptionsinquestionsisacommonlyacceptedpracticein surveyresearch,wesometimespreferthattherespondentdeterminewhethertheirperspectiveis strongerormoreintenseforwhichtheprobetechniqueused.anothermethodforacquiring informationfromrespondentsistoaskan open ended question.theopen endedquestionisa questionforwhichnoresponseoptionsareprovided,i.e.,itisentirelyuptotherespondenttoprovide theresponseinformation. TheElonUniversityPoll TheElonUniversityPollisconductedundertheauspicesoftheCenterforPublicOpinionPolling(Hunter Bacot,Director&MileahKromer,AssistantDirector),whichisaconstituentpartoftheInstitutefor PoliticsandPublicAffairs(GeorgeTaylor,Director);boththeseorganizationsarehousedinthe departmentofpoliticalscienceatelonuniversity.theseacademicunitsarepartofeloncollege,the CollegeofArtsandSciencesatElonUniversity,whichisunderthedirectionofDr.StevenHouse(Dean). TheElonUniversityadministration,ledbyDr.LeoLambert,Presidentoftheuniversity,fullysupportthe ElonUniversityPollaspartofitsservicecommitmenttostate,regional,andnationalconstituents.Dr. HunterBacot,aprofessorinthedepartmentofpoliticalscience,directstheElonUniversityPoll.Elon UniversitystudentsadministerthesurveyaspartoftheUniversity scommitmenttoexperiential learningwhere studentslearnthroughdoing.

4 II.SurveyInstrumentandPercentDistributionsbyQuestion Interviewswerecompletedwith662adultsfromhouseholdsintheNorthandSouthCarolina.Fora samplesizeof662,thereisa95percentprobabilitythatoursurveyresultsarewithinplusorminus 3.9percentagepoints(themarginofsamplingerror)oftheactualpopulationdistributionforany givenquestion.foreachstate,thesamplesizeandmarginofsamplingerrorare:northcarolina 356, +/ 5.3;SouthCarolina 305,+/ 5.7.Questionsarepresentedbytopic,notintheorderas administered;dataareweightedtoreflecttheadultpopulationintermsofgender. AbouttheCodesappearinginQuestionsandResponses ResponseOptionsnotoffered v=volunteeredresponse p=probedresponse Responseoptionsarenotofferedtothepersontakingthe survey(respondent),butareincludedinthequestionas asked(andusuallydenotedbybrackets,[]).response optionsaregenerallyofferedonlyfordemographicquestions (backgroundcharacteristic,e.g.,age,education,income, etc.). Respondentsvolunteerresponseoption.Asresponseoptions arenotofferedtothosetakingthesurvey,somerespondents offerorvolunteerresponseoptions.thoughnotall volunteeredoptionscanbeanticipated,themorecommon optionsarenoted. Respondentsself placeinthisoptionorcategory.aprobe maneuverisusedinquestionstoallowtherespondentto indicatewhetherher/hisresponseismoreintensethan initiallyprovidedforinthechoicesappearinginthequestion. Forexample,onprobequestionstheinterviewer,upona respondentindicatingthatshe/heissatisfied(ordissatisfied), isinstructedtoaskhim/her Wouldyousayyouare very satisfied?

5 First, I would like to know what you think is the most important issue facing your state? ECONOMY JOBS & UNEMPLOYMENT ELEMENTARY & SECONDARY EDUCATION HEALTH CARE TAXES CRIME & DRUGS IMMIGRATION FAMILY VALUES & MORALS UNIVERSITY SYSTEM TRAFFIC & ROAD CONSTRUCTION PUBLIC ASSISTAE/MEDICAID OR MEDICARE OTHER DON T KNOW REFUSED

6 Now, thinking about things in the country, do you feel things in this country [have gotten off on the wrong track, or are going in the right direction]? STRONGLY GOTTEN OFF ON THE WRONG TRACK (p) GOTTEN OFF ON THE WRONG TRACK GOING IN THE RIGHT DIRECTION STRONGLY GOING IN THE RIGHT DIRECTION (p) DON T KNOW (v) REFUSED (v) Now I m going to ask you some questions about the President.. Do you [approve or disapprove] of the way Barack Obama is handling his job as President? STRONGLY DISAPPROVE (p) DISAPPROVE APPROVE STRONGLY APPROVE (p) DON T KNOW (v) REFUSED (v).5.6.7

7 Now, I'd like to know how you think President Obama is handling some issues...i'm going to read you a list and first I want you to tell me if you [approve or disapprove]of how he is handling each issue. Then tell me if you think the President is paying [too little, about the right amount, or too much] attention to it. DISAPPROVE APPROVE TOO LITTLE ABOUT RIGHT TOO MUCH EDUCATION IMMIGRATION HEALTH CARE ECONOMY MORTGAGE CRISIS IRAQ WAR TAXES AFGHANISTAN BUDGET DEFICIT AUTO INDUSTRY FINAIAL CRISIS Total (; ) **Excludes Don t Know DISAPPROVE NORTH CAROLINA APPROVE TOO LITTLE ABOUT RIGHT TOO MUCH EDUCATION IMMIGRATION HEALTH CARE ECONOMY MORTGAGE CRISIS IRAQ WAR TAXES AFGHANISTAN BUDGET DEFICIT AUTO INDUSTRY FINAIAL CRISIS Total (; )

8 DISAPPROVE SOUTH CAROLINA APPROVE TOO LITTLE ABOUT RIGHT TOO MUCH EDUCATION IMMIGRATION HEALTH CARE ECONOMY MORTGAGE CRISIS IRAQ WAR TAXES AFGHANISTAN BUDGET DEFICIT AUTO INDUSTRY FINAIAL CRISIS Total (; ) Do you [approve or disapprove] of the way President Obama is handling the economy? STRONGLY DISAPPROVE (p) DISAPPROVE APPROVE STRONGLY APPROVE (p) DON T KNOW (v)

9 Do you think President Obama has proposed spending [too much money, about the right amount of money, or not enough money] to address the economic situation? NOT ENOUGH ABOUT THE RIGHT AMOUNT TOO MUCH DON T KNOW (v) REFUSED (v) Since taking office, have President Obama's economic policies made economic conditions [worse, better,] or have these not had an effect yet? WORSE NOT HAD AN EFFECT YET BETTER DON T KNOW (v) REFUSED (v)

10 So far, do you think President Obama is: [-trying to address too many issues at once, -focusing on too few issues, or -focusing on the right number of issues]? FOCUSING ON TOO FEW ISSUES FOCUSING ON THE RIGHT NUMBER OF ISSUES TRYING TO ADDRESS TOO MANY ISSUES AT OE DON'T KNOW (v) REFUSED (v).4.6.3

11 Now, changing topics a bit, I'm going to ask you about the level of Confidence you have in people or institutions in charge of running things in the United States... I'm going to read you a list, and I'd like for you to tell me if you have [a great deal of confidence, some confidence, or no confidence] in them... NO CONFIDEE SOME CONFIDEE A GREAT DEAL OF CONFIDEE DON T KNOW THE MILITARY WALL STREET COLLEGES & UNIVERSITIES ORGANIZED RELIGION TELEVISION NEWS CONGRESS BANKS LAW FIRMS UNITED SUPREME COURT LABOR UNIONS PUBLIC HOOLS MEDICAL DOCTORS SMALL BUSINESSES THE WHITE HOUSE Total (N=620; +/- 4.0%)

12 NO CONFIDEE SOME CONFIDEE A GREAT DEAL OF CONFIDEE DON T KNOW THE MILITARY WALL STREET COLLEGES & UNIVERSITIES ORGANIZED RELIGION TELEVISION NEWS CONGRESS BANKS LAW FIRMS UNITED SUPREME COURT LABOR UNIONS PUBLIC HOOLS MEDICAL DOCTORS SMALL BUSINESSES THE WHITE HOUSE Total (; )

13 NO CONFIDEE SOME CONFIDEE A GREAT DEAL OF CONFIDEE DON T KNOW THE MILITARY WALL STREET COLLEGES & UNIVERSITIES ORGANIZED RELIGION TELEVISION NEWS CONGRESS BANKS LAW FIRMS UNITED SUPREME COURT LABOR UNIONS PUBLIC HOOLS MEDICAL DOCTORS SMALL BUSINESSES THE WHITE HOUSE Total (; )

14 Do you think the [Republican party or the Democratic party] is doing a better job of managing the economy? DEMOCRATIC PARTY REPUBLICAN PARTY NOT DECIDED/UNSURE (v) NEITHER PARTY (v) DON T KNOW (v) REFUSED (v) Do you expect the national economy to [get better, stay about the same, or get worse] by the end of this year? GET WORSE STAY ABOUT THE SAME GET BETTER DON'T KNOW (v) Do you expect the state economy [to get better, stay about the same, or get worse] by the end of this year? GET WORSE STAY ABOUT THE SAME GET BETTER DON'T KNOW (v)

15 Thinking about the economy overall, do you think that... [-it has started to improve, -that the worst is yet to come, or -that it has leveled-off but has not started to improve yet.] THE WORST IS YET TO COME HAS LEVELED OFF, BUT HAS NOT STARTED TO IMPROVE YET STARTED TO IMPROVE DON'T KNOW (v) REFUSED (v)

I. Survey Methodology

I. Survey Methodology I.SurveyMethodology TheElonUniversityPollisconductedusingastratifiedrandomsampleofhouseholdswithtelephonesin thepopulationofinterest inthiscase,citizensinnorthcarolina.thesampleoftelephonenumbers forthesurveyisobtainedfromsurveysamplinginternational,llc.