2.3 Inductive or Deductive? Two Different Approaches

A deductive approach to research is the one that people typically associate with scientific investigation. The researcher studies what others have done, reads ... PreviousSection TableofContents NextSection 2.3InductiveorDeductive?TwoDifferentApproaches LearningObjectives Describetheinductiveapproachtoresearch,andprovideexamplesofinductiveresearch. Describethedeductiveapproachtoresearch,andprovideexamplesofdeductiveresearch. Describethewaysthatinductiveanddeductiveapproachesmaybecomplementary. Theoriesstructureandinformsociologicalresearch.So,too,doesresearchstructureandinformtheory.Thereciprocalrelationshipbetweentheoryandresearchoftenbecomesevidenttostudentsnewtothesetopicswhentheyconsidertherelationshipsbetweentheoryandresearchininductiveanddeductiveapproachestoresearch.Inbothcases,theoryiscrucial.Buttherelationshipbetweentheoryandresearchdiffersforeachapproach.Inductiveanddeductiveapproachestoresearcharequitedifferent,buttheycanalsobecomplementary.Let’sstartbylookingateachoneandhowtheydifferfromoneanother.Thenwe’llmoveontothinkingabouthowtheycomplementoneanother. InductiveApproachesandSomeExamples InaninductiveapproachCollectdata,analyzepatternsinthedata,andthentheorizefromthedata.toresearch,aresearcherbeginsbycollectingdatathatisrelevanttohisorhertopicofinterest.Onceasubstantialamountofdatahavebeencollected,theresearcherwillthentakeabreatherfromdatacollection,steppingbacktogetabird’seyeviewofherdata.Atthisstage,theresearcherlooksforpatternsinthedata,workingtodevelopatheorythatcouldexplainthosepatterns.Thuswhenresearcherstakeaninductiveapproach,theystartwithasetofobservationsandthentheymovefromthoseparticularexperiencestoamoregeneralsetofpropositionsaboutthoseexperiences.Inotherwords,theymovefromdatatotheory,orfromthespecifictothegeneral.Figure2.5"InductiveResearch"outlinesthestepsinvolvedwithaninductiveapproachtoresearch. Figure2.5InductiveResearch Therearemanygoodexamplesofinductiveresearch,butwe’lllookatjustafewhere.OnefascinatingrecentstudyinwhichtheresearcherstookaninductiveapproachwasKatherineAllen,ChristineKaestle,andAbbieGoldberg’sstudy(2011)Allen,K.R.,Kaestle,C.E.,&Goldberg,A.E.(2011).Morethanjustapunctuationmark:Howboysandyoungmenlearnaboutmenstruation.JournalofFamilyIssues,32,129–156.ofhowboysandyoungmenlearnaboutmenstruation.Tounderstandthisprocess,Allenandhercolleaguesanalyzedthewrittennarrativesof23youngmeninwhichthemendescribedhowtheylearnedaboutmenstruation,whattheythoughtofitwhentheyfirstlearnedaboutit,andwhattheythinkofitnow.Bylookingforpatternsacrossall23men’snarratives,theresearcherswereabletodevelopageneraltheoryofhowboysandyoungmenlearnaboutthisaspectofgirls’andwomen’sbiology.Theyconcludethatsistersplayanimportantroleinboys’earlyunderstandingofmenstruation,thatmenstruationmakesboysfeelsomewhatseparatedfromgirls,andthatastheyenteryoungadulthoodandformromanticrelationships,youngmendevelopmorematureattitudesaboutmenstruation. Inanotherinductivestudy,KristinFergusonandcolleagues(Ferguson,Kim,&McCoy,2011)Ferguson,K.M.,Kim,M.A.,&McCoy,S.(2011).Enhancingempowermentandleadershipamonghomelessyouthinagencyandcommunitysettings:Agroundedtheoryapproach.ChildandAdolescentSocialWorkJournal,28,1–22.analyzedempiricaldatatobetterunderstandhowbesttomeettheneedsofyoungpeoplewhoarehomeless.Theauthorsanalyzeddatafromfocusgroupswith20youngpeopleatahomelessshelter.Fromthesedatatheydevelopedasetofrecommendationsforthoseinterestedinappliedinterventionsthatservehomelessyouth.Theresearchersalsodevelopedhypothesesforpeoplewhomightwishtoconductfurtherinvestigationofthetopic.ThoughFergusonandhercolleaguesdidnottestthehypothesesthattheydevelopedfromtheiranalysis,theirstudyendswheremostdeductiveinvestigationsbegin:withasetoftestablehypotheses. DeductiveApproachesandSomeExamples ResearcherstakingadeductiveapproachDevelophypothesesbasedonsometheoryortheories,collectdatathatcanbeusedtotestthehypotheses,andassesswhetherthedatacollectedsupportthehypotheses.takethestepsdescribedearlierforinductiveresearchandreversetheirorder.Theystartwithasocialtheorythattheyfindcompellingandthentestitsimplicationswithdata.Thatis,theymovefromamoregeneralleveltoamorespecificone.Adeductiveapproachtoresearchistheonethatpeopletypicallyassociatewithscientificinvestigation.Theresearcherstudieswhatothershavedone,readsexistingtheoriesofwhateverphenomenonheorsheisstudying,andthentestshypothesesthatemergefromthosetheories.Figure2.6"DeductiveResearch"outlinesthestepsinvolvedwithadeductiveapproachtoresearch. Figure2.6DeductiveResearch Whilenotallresearchersfollowadeductiveapproach,asyouhaveseenintheprecedingdiscussion,manydo,andthereareanumberofexcellentrecentexamplesofdeductiveresearch.We’lltakealookatacoupleofthosenext. InastudyofUSlawenforcementresponsestohatecrimes,RyanKingandcolleagues(King,Messner,&Baller,2009)King,R.D.,Messner,S.F.,&Baller,R.D.(2009).Contemporaryhatecrimes,lawenforcement,andthelegacyofracialviolence.AmericanSociologicalReview,74,291–315.hypothesizedthatlawenforcement’sresponsewouldbelessvigorousinareasofthecountrythathadastrongerhistoryofracialviolence.Theauthorsdevelopedtheirhypothesisfromtheirreadingofpriorresearchandtheoriesonthetopic.Next,theytestedthehypothesisbyanalyzingdataonstates’lynchinghistoriesandhatecrimeresponses.Overall,theauthorsfoundsupportfortheirhypothesis. Inanotherrecentdeductivestudy,MelissaMilkieandCatharineWarner(2011)Milkie,M.A.,&Warner,C.H.(2011).Classroomlearningenvironmentsandthementalhealthoffirstgradechildren.JournalofHealthandSocialBehavior,52,4–22.studiedtheeffectsofdifferentclassroomenvironmentsonfirstgraders’mentalhealth.Basedonpriorresearchandtheory,MilkieandWarnerhypothesizedthatnegativeclassroomfeatures,suchasalackofbasicsuppliesandevenheat,wouldbeassociatedwithemotionalandbehavioralproblemsinchildren.Theresearchersfoundsupportfortheirhypothesis,demonstratingthatpolicymakersshouldprobablybepayingmoreattentiontothementalhealthoutcomesofchildren’sschoolexperiences,justastheytrackacademicoutcomes(AmericanSociologicalAssociation,2011).TheAmericanSociologicalAssociationwroteapressreleaseonMilkieandWarner’sfindings:AmericanSociologicalAssociation.(2011).Study:Negativeclassroomenvironmentadverselyaffectschildren’smentalhealth.Retrievedfromhttp://asanet.org/press/Negative_Classroom_Environment_Adversely_Affects_Childs_Mental_Health.cfm ComplementaryApproaches? Whileinductiveanddeductiveapproachestoresearchseemquitedifferent,theycanactuallyberathercomplementary.Insomecases,researcherswillplanfortheirresearchtoincludemultiplecomponents,oneinductiveandtheotherdeductive.Inothercases,aresearchermightbeginastudywiththeplantoonlyconducteitherinductiveordeductiveresearch,butthenheorshediscoversalongthewaythattheotherapproachisneededtohelpilluminatefindings.Hereisanexampleofeachsuchcase. Inthecaseofmycollaborativeresearchonsexualharassment,webeganthestudyknowingthatwewouldliketotakebothadeductiveandaninductiveapproachinourwork.Wethereforeadministeredaquantitativesurvey,theresponsestowhichwecouldanalyzeinordertotesthypotheses,andalsoconductedqualitativeinterviewswithanumberofthesurveyparticipants.Thesurveydatawerewellsuitedtoadeductiveapproach;wecouldanalyzethosedatatotesthypothesesthatweregeneratedbasedontheoriesofharassment.Theinterviewdatawerewellsuitedtoaninductiveapproach;welookedforpatternsacrosstheinterviewsandthentriedtomakesenseofthosepatternsbytheorizingaboutthem. Foronepaper(Uggen&Blackstone,2004),Uggen,C.,&Blackstone,A.(2004).Sexualharassmentasagenderedexpressionofpower.AmericanSociologicalReview,69,64–92.webeganwithaprominentfeministtheoryofthesexualharassmentofadultwomenanddevelopedasetofhypothesesoutlininghowweexpectedthetheorytoapplyinthecaseofyoungerwomen’sandmen’sharassmentexperiences.Wethentestedourhypothesesbyanalyzingthesurveydata.Ingeneral,wefoundsupportforthetheorythatpositedthatthecurrentgendersystem,inwhichheteronormativemenwieldthemostpowerintheworkplace,explainedworkplacesexualharassment—notjustofadultwomenbutofyoungerwomenandmenaswell.Inamorerecentpaper(Blackstone,Houle,&Uggen,2006),Blackstone,A.,Houle,J.,&Uggen,C.“AtthetimeIthoughtitwasgreat”:Age,experience,andworkers’perceptionsofsexualharassment.Presentedatthe2006meetingsoftheAmericanSociologicalAssociation.Currentlyunderreview.wedidnothypothesizeaboutwhatwemightfindbutinsteadinductivelyanalyzedtheinterviewdata,lookingforpatternsthatmighttellussomethingabouthoworwhetherworkers’perceptionsofharassmentchangeastheyageandgainworkplaceexperience.Fromthisanalysis,wedeterminedthatworkers’perceptionsofharassmentdidindeedshiftastheygainedexperienceandthattheirlaterdefinitionsofharassmentweremorestringentthanthosetheyheldduringadolescence.Overall,ourdesiretounderstandyoungworkers’harassmentexperiencesfully—intermsoftheirobjectiveworkplaceexperiences,theirperceptionsofthoseexperiences,andtheirstoriesoftheirexperiences—ledustoadoptbothdeductiveandinductiveapproachesinthework. Researchersmaynotalwayssetouttoemploybothapproachesintheirworkbutsometimesfindthattheiruseofoneapproachleadsthemtotheother.OnesuchexampleisdescribedeloquentlyinRussellSchutt’sInvestigatingtheSocialWorld(2006).Schutt,R.K.(2006).Investigatingthesocialworld:Theprocessandpracticeofresearch.ThousandOaks,CA:PineForgePress.AsSchuttdescribes,researchersLawrenceShermanandRichardBerk(1984)Sherman,L.W.,&Berk,R.A.(1984).Thespecificdeterrenteffectsofarrestfordomesticassault.AmericanSociologicalReview,49,261–272.conductedanexperimenttotesttwocompetingtheoriesoftheeffectsofpunishmentondeterringdeviance(inthiscase,domesticviolence).Specifically,ShermanandBerkhypothesizedthatdeterrencetheorywouldprovideabetterexplanationoftheeffectsofarrestingaccusedbatterersthanlabelingtheory.Deterrencetheorypredictsthatarrestinganaccusedspousebattererwillreducefutureincidentsofviolence.Conversely,labelingtheorypredictsthatarrestingaccusedspousebattererswillincreasefutureincidents.Figure2.7"PredictingtheEffectsofArrestonFutureSpouseBattery"summarizesthetwocompetingtheoriesandthepredictionsthatShermanandBerksetouttotest. Figure2.7PredictingtheEffectsofArrestonFutureSpouseBattery ShermanandBerkfound,afterconductinganexperimentwiththehelpoflocalpoliceinonecity,thatarrestdidinfactdeterfutureincidentsofviolence,thussupportingtheirhypothesisthatdeterrencetheorywouldbetterpredicttheeffectofarrest.Afterconductingthisresearch,theyandotherresearcherswentontoconductsimilarexperimentsTheresearchersdidwhat’scalledreplication.We’lllearnmoreaboutreplicationinChapter3"ResearchEthics".insixadditionalcities(Berk,Campbell,Klap,&Western,1992;Pate&Hamilton,1992;Sherman&Smith,1992).Berk,R.,Campbell,A.,Klap,R.,&Western,B.(1992).Thedeterrenteffectofarrestinincidentsofdomesticviolence:ABayesiananalysisoffourfieldexperiments.AmericanSociologicalReview,57,698–708;Pate,A.,&Hamilton,E.(1992).Formalandinformaldeterrentstodomesticviolence:TheDadecountyspouseassaultexperiment.AmericanSociologicalReview,57,691–697;Sherman,L.,&Smith,D.(1992).Crime,punishment,andstakeinconformity:Legalandinformalcontrolofdomesticviolence.AmericanSociologicalReview,57,680–690.Resultsfromthesefollow-upstudiesweremixed.Insomecases,arrestdeterredfutureincidentsofviolence.Inothercases,itdidnot.Thislefttheresearcherswithnewdatathattheyneededtoexplain.Theresearchersthereforetookaninductiveapproachinanefforttomakesenseoftheirlatestempiricalobservations.Thenewstudiesrevealedthatarrestseemedtohaveadeterrenteffectforthosewhoweremarriedandemployedbutthatitledtoincreasedoffensesforthosewhowereunmarriedandunemployed.Researchersthusturnedtocontroltheory,whichpredictsthathavingsomestakeinconformitythroughthesocialtiesprovidedbymarriageandemployment,asthebetterexplanation. Figure2.8PredictingtheEffectsofArrestonFutureSpouseBattery:ANewTheory WhattheShermanandBerkresearch,alongwiththefollow-upstudies,showsusisthatwemightstartwithadeductiveapproachtoresearch,butthen,ifconfrontedbynewdatathatwemustmakesenseof,wemaymovetoaninductiveapproach.RussellSchuttdepictsthisprocessquitenicelyinhistext,andI’veadaptedhisdepictionhere,inFigure2.9"TheResearchProcess:MovingFromDeductivetoInductiveinaStudyofDomesticViolenceRecidivism". KeyTakeaways Theinductiveapproachinvolvesbeginningwithasetofempiricalobservations,seekingpatternsinthoseobservations,andthentheorizingaboutthosepatterns. Thedeductiveapproachinvolvesbeginningwithatheory,developinghypothesesfromthattheory,andthencollectingandanalyzingdatatotestthosehypotheses. Inductiveanddeductiveapproachestoresearchcanbeemployedtogetherforamorecompleteunderstandingofthetopicthataresearcherisstudying. Thoughresearchersdon’talwayssetouttousebothinductiveanddeductivestrategiesintheirwork,theysometimesfindthatnewquestionsariseinthecourseofaninvestigationthatcanbestbeansweredbyemployingbothapproaches. Exercises Forahilariousexampleoflogicgoneawry,checkoutthefollowingclipfrom MontyPythonandHolyGrail: (clicktoseevideo) Dothetownspeopletakeaninductiveordeductiveapproachtodeterminewhetherthewomaninquestionisawitch?Whataresomeofthedifferentsourcesofknowledge(recallChapter1"Introduction")theyrelyon? Thinkabouthowyoucouldapproachastudyoftherelationshipbetweengenderanddrivingoverthespeedlimit.Howcouldyoulearnaboutthisrelationshipusinganinductiveapproach?Whatwouldastudyofthesamerelationshiplooklikeifexaminedusingadeductiveapproach?Trythesamethingwithanytopicofyourchoice.Howmightyoustudythetopicinductively?Deductively? PreviousSection TableofContents NextSection