如何設(shè)計(jì)問(wèn)卷 英文 Questionnaire Design

Howtodesignaquestionnaireasurveyinvolvesdirectlycollectinginformationfrompeople(orsometimesorganisations)whomweareinterestedin.Welldesignedquestionnairesarehighlystructuredtoallowthesametypesofinformationtobecollectedfromalargenumberofpeopleinthesamewayandfordatatobeanalysedquantitativelyandsystematically.Questionnairesarebestusedforcollectingfactualdataandappropriatequestionnairedesignisessentialtoensurethatweobtainvalidresponsestoourquestions.ObjectivesindesigningquestionnairesTherearetwomainobjectivesindesigningaquestionnaire:Tomaximizetheproportionofsubjectsansweringourquestionnaire—thatis,theresponserate.Toobtainaccuraterelevantinformationforoursurvey.

DecidingwhattoaskTherearethreepotentialtypesofinformation:Informationweareprimarilyinterestedin—thatis,dependentvariables.Informationwhichmightexplainthedependentvariables—thatis,independentvariables.Otherfactorsrelatedtobothdependentandindependentfactorswhichmaydistorttheresultsandhavetobeadjustedfor—thatis,confoundingvariablesWordingofindividualquestionsThewayquestionsarephrasedisimportantandtherearesomegeneralrulesforconstructinggoodquestionsinquestionnaireUseshortandsimplesentencesShort,simplesentencesaregenerallylessconfusingandambiguousthanlong,complexones.Asaruleofthumb,mostsentencesshouldcontainoneortwoclauses.Sentenceswithmorethanthreeclausesshouldberephrased.

AskforonlyonepieceofinformationatatimeForexample,“Pleaseratethelectureintermsofitscontentandpresentation”asksfortwopiecesofinformationatthesametime.Itshouldbedividedintotwoparts:“Pleaseratethelectureintermsof(a)itscontent,(b)itspresentation.”AvoidnegativesifpossibleNegativesshouldbeusedonlysparingly（謹(jǐn)慎地）.Forexample,insteadofaskingstudentswhethertheyagreewiththestatement,“Smallgroupteachingshouldnotbeabolished,”thestatementshouldberephrasedas,“Smallgroupteachingshouldcontinue.”Doublenegativesshouldalwaysbeavoided.AskprecisequestionsQuestionsmaybeambiguousbecauseawordortermmayhaveadifferentmeaning.Forexample,ifweaskstudentstoratetheirinterestin“medicine,”thistermmightmean“generalmedicine”(asopposedtogeneralsurgery)tosome,butinclusiveofallclinicalspecialties(asopposedtoprofessionsoutsidemedicine)toothers.Anothersourceofambiguityisafailuretospecifyaframeofreference.Forexample,inthequestion,“Howoftendidyouborrowbooksfromyourlibrary?”thetimereferenceismissing.Itmightberephrasedas,“Howmanybookshaveyouborrowedfromthelibrarywithinthepastsixmonthsaltogether?”EnsurethoseyouaskhavethenecessaryknowledgeForexample,inasurveyofuniversitylecturersonrecentchangesinhighereducation,thequestion,“DoyouagreewiththerecommendationsintheDearingreportonhighereducation?”isunsatisfactoryforseveralreasons.Notonlydoesitaskforseveralpiecesofinformationatthesametimeasthereareseveralrecommendationsinthereport,thequestionalsoassumesthatalllecturersknowabouttherelevantrecommendations.SensitiveissuesItisoftendifficulttoobtaintruthfulanswerstosensitivequestions.Clearly,thequestion,“Haveyouevercopiedotherstudents’answersinadegreeexam?”islikelytoproduceeithernoresponseornegativeresponses.Lessdirectapproacheshavebeensuggested.Firstly,thecasualapproach:“Bytheway,doyouhappentohavecopiedotherstudents’answersinadegreeexam?”maybeusedasalastpartofanotherdecoyquestion.

MinimisebiasPeopletendtoanswerquestionsinawaytheyperceivetobesociallydesiredorexpectedbythequestionerandtheyoftenlookforcluesinthequestions.Manyapparentlyneutralquestionscanpotentiallyleadtobias.Forexample,inthequestion,“Withinthepastmonth,howmanylectureshaveyoumissedduetoyoureveningjob?”studentsmayperceivethedesiredresponsestobe“never”tothefirstquestion.Thisquestioncouldberephrasedas,“Withinthepastmonth,howmanytimesdidyoureveningjobcommitmentclashwithlectures?Howmanytimesdidyougiveprioritytoyoureveningjob?”Takeanotherexample.Thequestion,“Pleaseratehowusefulthefollowingtext-booksare.Pleasealsostatewhethertheyareincludedinyourlecturer’srecommendedreadinglist?”Thereisariskthatthestudentsmayperceivethattheyshouldratebooksrecommendedbylecturersmorefavourablythanthosenotrecommendedbytheirlecturers.Thisriskmaybeminimisedbyputtingthesecondquestionlateroninthequestionnaire.FormatofresponsesTheresponsescanbeinopenorclosedformats.Box1:AdvantagesofopenorclosedformatOpenformatlAllowsexplorationoftherangeofpossiblethemesarisingfromanissuelCanbeusedevenifacomprehensiverangeofalternativechoicescannotbecompiledClosed—thatis,forcedchoice—formatlEasyandquicktofillinlMinimisediscriminationagainstthelessliterate(inselfadministeredquestionnaire)orthelessarticulate(ininterviewquestionnaire)lEasytocode,record,andanalyseresultsquantitativelylEasytoreportresultsLengthofquestionnaireTherearenouniversalagreementsabouttheoptimallengthofquestionnaires.Itprobablydependsonthetypeofrespondents.However,shortsimplequestionnairesusuallyattracthigherresponseratesthanlongcomplexones.ArrangingthequestionsTheorderofthequestionsisalsoimportant.Somegeneralrulesare:lGofromgeneraltoparticular.lGofromeasytodifficult.lGofromfactualtoabstract.lStartwithclosedformatquestions.lStartwithquestionsrelevanttothemainsubject.lDonotstartwithdemographicandpersonalquestions.

Introduction,personalisedletter,andendingItseemsagoodideatohaveeitherapersonalisedcoveringletteroratleastanintroductionexplainingbrieflythepurposeofthesurvey,theimportanceoftherespondents’participation,whoi