alexander y knight - situated identities and social pshychological experimentation.pdf

Upload: whatever-forever

Post on 14-Apr-2018

240 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    1/19

    Situated Identities and Social Psychological ExperimentationAuthor(s): C. Norman Alexander, Jr. and Gordon W. KnightSource: Sociometry, Vol. 34, No. 1 (Mar., 1971), pp. 65-82Published by: American Sociological AssociationStable URL: http://www.jstor.org/stable/2786351 .

    Accessed: 02/01/2011 19:28

    Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

    you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

    may use content in the JSTOR archive only for your personal, non-commercial use.

    Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=asa. .

    Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

    page of such transmission.

    JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

    content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

    of scholarship. For more information about JSTOR, please contact [email protected].

    American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to

    Sociometry.

    http://www.jstor.org

    http://www.jstor.org/action/showPublisher?publisherCode=asahttp://www.jstor.org/stable/2786351?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=asahttp://www.jstor.org/action/showPublisher?publisherCode=asahttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/2786351?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=asa
  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    2/19

    Sociometry1971, Vol. 34, No. 1, 65-82

    SituatedIdentities nd SocialPsychologicalExperimentation*C. NORMAN ALEXANDER, JR. AND GORDON W. KNIGHTStanfordUniversity

    Thispaper ntroduces newapproach o theconceptualizationndmeasure-mentof experimentalonditionsnd to the assessment f themeaning fdependent-variableesponsesn a givencondition. y simulatingheresultsof a seriesof "insufficientustification"xperiments,ontradictorynd in-explicable indingsre accounted or.The implicationsf the situation ndofavailableresponse lternativesorthe "situated dentity" f thesubject-actorare related o subjectbehavior.Withthesimplehypothesishatsub-jectsattempt o create he mostfavorable ituated dentitiesn experimentalencounters,e are able to explainwhythey nswred hedependent-variablequestion s theydid and as observers stimated heywoulddo. Focusingon thesalienceof situated dentityttributesnables thisapproach, nlikeothers,o specify hedifferenceshatdistinguishne experimentalonditionfrom nother.We discuss he mplicationsfthesefindingsnd recommenda newtheoreticalnd methodologicalpproach o experimentationn socialpsychology.

    Behaviorbecomesmeaningfulocial action when t is perceptuallynte-grated nto a shared nterpretiveerspectiveHeider, 1958; Mead, 1934).Attributionheories uggest that these explanatory ramesof referenceimportantlynvolve mputationsbout the dispositional haracteristicsfactors (Jonesand Davis, 1965; Kelley,1967), whileothertheorists avestrongly mphasizedthat social identity ttributes rganizeand orientinteractionGoffman,959,1963; Strauss, 959). Thesevarious erspectivessuggest hepossibility f describingocial activityn terms f theidentityattributesfparticipants.In thispaper wepropose o conceptualize nd measure hedefinitionndmeaning f situations n terms f their situated dentity"mplicationsoran actor's behavior.The term"situated dentity" Alexanderand Weil,1969) designates hedispositionalmputationsbout an individual hatareconveyed y his actions n a particularocialcontext.When a person cts,

    * This research was supported by a grant from the National Science Foundation(GS-2759, "SituationalMeaning and Social PsychologicalExperimentation",C. NormanAlexander,Jr.,Principal Investigator).65

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    3/19

    66 SOCIOMETRYhe communicatesnformationbout the kind ofpersonhe presumes o beand obligesothers o regardhimas beingthatkindof person. n Mead'sterminology,ehaviors re "constituted"nto action,because theyhave im-plications for the creation, ffirmation,r transformationf an actor'ssituated dentity. he social situation s conceptualized,rom n actor'sper-spective, s a range of possible identities ntailedby the behavioral l-ternatives e confronts.For situations o be socially defined here mustbe relativeconsensusabout the meaningof actions. Within a population t is necessary hatthere e some greementbout the dispositionalimensionshat re relevantto describe n individual's onduct, nd about how a particular ction isto be evaluated along those dimensions.When theseconditions re met,thenwe can say that a situation as consensualmeaning r social reality.In essence,we will measure ocial situations y determininghe situatedidentitymplicationsonveyed y behaviorwithin hem.When a social situationxistsby thesecriteria,ehavior an be predictedif the situated dentity hat resultsfrom he choice of one action s moresociallydesirable hanthose ssociatedwith lternativections: ndividualsshouldchoose the most favorably valuatedsituated dentity. hus, ourperspectivean contributeo the understandingf social actionwhen be-havioral lternativesmplydifferentiallyesirable ituated dentities.Althoughwe do not presumethat people spontaneously erbalize orconsciously epresentheiractivity n theseterms,we do assume that asituated dentity nalysis dequatelydescribes heirnormative rientationsand interpretiveerspectivesoward onductn a given ocialcontext.Nar-rative ncidents rom verydayife (Goffman, 961, 1963), convince s oftheconceptualndpredictive tility f this pproach.Moreover, he specula-tionand evidence bout subject orientationsn social psychological xperi-ments Friedman, 967; Riecken,1962; Rosenthal, 966) make t impera-tivetodeal systematically ith ituationalmeanings.After elating hesituated dentity erspectiveo events n experimentalsettings, e willapply t to theanalysis nd interpretationf somepuzzlingand controversialindingsf "insufficientustification" tudies. Followingthepresentationf methodsnd results rom study estingurhypotheses,we will discuss their mplicationsorexperimentationn social psychology.The paper concludeswith recommendationsor future esearch trategiesand for shift n the focus f nvestigatoryfforts.Experimentalettingsnd Situated dentity

    In recentyears there has been a growing oncern mong social psy-chologistsbout understandinghe "meaning" o subjectsof experimental

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    4/19

    SITUATED IDENTITIES AND THE EXPERIMENT 67situations.he problem as become cutebecauseofthecompellingemon-strationshat ubjects re incrediblyensitive o subtledifferencesn experi-menterttitudes,ctions, nd expectationsRosenthal nd Rosnow,1969).Apartfromhe question fthe accuracy ftheir nticipations,roblemsreraisedby the mere fact that subjects ntuitpurposes nd interpret ro-cedures s theyencounteraboratory vents.Whenwe add apprehensionaboutpersonal valuations r impressionormationRosenberg, 965,1969)to thisheightenednterpersonalensitivity,t seems nevitable hatsubjectswill be aware of the situated dentitymplicationsf theirbehaviors ndattentiveo them.We suspectthatexperimentalettings re the kind of social situationsthatwe proposeto measure n terms f situated dentity oncepts.Bem's(1967) demonstrationhat"simulation" bserversre able to predict ub-jects' responsesn experimentalettingsncourages s to believe that thisis the case. We interpretuccessful interpersonalimulations"s showingthat there s someunderlyingnterpretiverameworkhat makessense ofexperimentalhoices venundernovelcircumstances.t seemsobviousthatobserversould not make consensual much ess "correct")estimationsfsubjectresponsesunless the action situationwere sociallystructured.fthe simulated ituations o clearlyconveys he differentialesirabilityfresponses o observers, hen it is likelythat the actual situationcom-municated imilar nformationo subjects.Indeed, Bem argues stronglythatsubjects nd observers ely on the samesituational ues to determinetheappropriateesponse.Although e does not speculate bout thenatureofthesocial meanings, e wantto specifyhem.To test our ideaswe selected studyfrom he dissonanceiteraturehathas spawnedconsiderableontroversynd a complicatederiesof replica-tions nd simulations.Festingernd Carlsmith's tudy (1959) required ubjectsto performdull, boringtask, and-when a research assistantwas "unexpectedly"absent-to informn ostensibly aive subjectthatthe taskwas interestingand enjoyable.For lyingto theotherperson, ubjectswere told thattheywouldbe paid either 1 or $20, afterwhich heyratedthe task. Thosepaid$20 to makethepositive ask-descriptionsere morenegative owardthetaskthanthosepaid $1 tosay they iked t.Dissonance (Brehm and Cohen, 1962; Festinger, 957) predictsthisresult, easoninghat negative ask-attitudes moredissonantwithmakinga positive askdescriptionor$1 thanwithmaking t for$20. Thus, thetheoryrgues hat ubjects n the$1,conditionhange heir ttitudes owardthe task to be moreconsonantwiththeirpositive haracterizationsf it.Subsequently,emreplicated hesetask-likingifferencessingtheestima-

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    5/19

    68 SOCIOMETRYtions of observers o whom experimentalventswere depictedon tape-recordings,1967).Because theywerecontraryo "common ense,"theresults f the originalstudy generated onsiderable ttention nd discussion.During the sub-sequent debate, interpretationf the resultshas shifted o questions ofsufficientersus nsufficientustification orbehavior. uch concerns re inclose accordwith ttributionheories Jones nd Davis, 1965; Kelly, 1967).These theorieswould suggestthat subjects sought a socially acceptablerationale or heir yingbehaviors. n highpayment onditions he externalincentivetself ffordeduch a rationale;while n low payment onditionssubjects esorted o internalxplanations.When the external ncentives ereabsent,subjects attempted o account for theirconductby increasing e-ported and perhaps felt") task-liking.Carlsmith, ollins, and Helmreich 1966) replicated nd extended heFestingernd Carlsmithtudy,withvariations hat re ofconsiderablenter-est. Half their ubjectswerepaid from .50 to $5.00 to tell a naive subjectthat the dull experimentalaskwas interesting,hus essentially uplicatingthe original xperimentalonditions. he otherhalf wrote an anonymousessay, imilarly escribinghetask,buttheessay wouldbe seenonlyby theexperimenter.Although heyreplicatedheFestingernd Carlsmithesults n the"faceto face" condition, ask-likingncreasedwith ncentiven theEssay condi-tions. They conclude that undersome conditionsncreasing ressureforcomplianceeads to smaller mounts f attitude hange,whileunderothersit producesmore hange. What remains nspecified,owever,s thecrucialdifferenceetween he role-play nd essay-writingonditions"Carlsmith,et al., 1966,p. 12). Despitea gooddeal ofspeculativenterpretationftheirresults as well as subsequentempiricalwork (Helmreichand Collins,1968) these kindsof crucial differencesemainunspecified.t seems un-likelyto us that furtherttempts o reconciledissonance nd incentivetheorynterpretationsf theseresultswill be successful.The crucial differencesetween xperimentalonditions s well as thechoicesmadeby subjects n each conditionan be explained y oursituatedidentity erspective.We hypothesizehat the response lternatives f thetask-liking uestionare associatedwith differentiallyvaluated situatedidentity ttributionsnd thatsubjects hose themostfavorable ne in eachcondition.

    1 For example, see the papers by J. Merrill Carlsmith, rving L. Janis, Barry E.Collins, and Milton J. Rosenberg, n Robert P. Abelson et al., 1968:801-833.

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    6/19

    SITUATED IDENTITIES AND THE EXPERIMENT 69METHOD

    We createdtape recordings f the fourbasic experimentalonditions osimulate he essential eatures f the "insufficientustification"tudies.Thetapes depict naive subject, Bob Downing", pending n hourperformingtwodull, boringtasks-stacking spools on a trayand turning egs on aboard.After oingthis,he is given minimal xplanation f the study: Heis told that he has been a subject n the experiment'sontrolgroup, nwhich ubjects re toldnothing bout the tasks ahead of time; n thetreat-mentgroups, owever, ubjects re initially ed to expect hat the tasks arefun.Treatment roup xpectationsre createdby a ResearchAssistantwhopretends o be a subject ust completing is participationn the study.Hedescribes he tasks as interesting,xciting, njoyable, nd fun.The allegedpurposeof the study s to investigate he effects f task expectations nsubjects'performances.Following this explanation, he experimenterequests Bob to give afavorable escription f the taskshe performed; ob agrees to do so, andprovides rather nthusiastic resentation. inally,he is asked to reporthow enjoyablehe actually foundthe tasks to be (on an 11-point cale:very dull and boring-very nterestingnd enjoyable). The manipulationsthat differentiatehe conditions oncern 1) how muchmoneyBob is paidtomake the favorable ask descriptions,nd (2) the nature f his presenta-tionalencounter.IncentiveManipulation:To replicate heCarlsmith,t al. (1966), levelsof payment nd avoid theextraordinarilyigh$20.00 payment tilized ntheoriginal estingernd Carlsmithxperiment1959), we had theexperi-menter ffer ob $.50 in oneset ofconditionsnd $5.00 in theothers.EncounterManipulation:n one set ofconditions,ob is askedto tellanostensibly aive subjectthatthe tasks are enjoyable. t is explained hatthis is necessarybecause the experimenter'sssistant (who usually setsexpectations)s unfortunatelybsent and a "treatment ondition" ubjectis waiting o participate. his condition,Face Encounter",s intended oreplicate he original xperimentnd the "role-playing"ondition f thereplication.n the "Essay Encounter"Bob is asked to writea favorabletaskdescriptionnd dictate t into a taperecorder rovided y theexperi-menter. t is explainedthat his favorable mpressionswill provide thescriptfor a research ssistantwho will later describe he tasks to naivesubjectsn thetreatmentonditions.Up to thispoint,our experiments essentiallyn Interpersonalimula-tion (Bem, 1967) of Carlsmith,t al., (1966). Observers our subjects)hear one of the tapes, including ob's favorable escription f the tasks,

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    7/19

    70 SOCIOMETRYthen hey stimate is response n thepost-experimentalask-likinguestion.However,we extended heexperimentalesignby informingbservers owthe "Bob" theyheard had actually nswered he task-likinguestion.Afterthis,we asked observers or their mpressionsf what Bob was like as aperson.Bob-LikingManipulation:Observerswere told Bob's responseto thefollowing uestion: "How enjoyablewere the tasks? Indicateyour feelingby checking n appropriateointon the inebelow:"-5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5/ / / / / / / / / / /VeryDull Neutral Very Interestingand Boring Feelings and EnjoyableAfter stimatingow Bob responded, bserversweretold that he actuallychecked ne (of five)position nthe cale,from 2 to + 2.Thus, our experimentontains20 cells in a 2x 2x 5 design: Incentiveby Encounter y Bob-Liking.Afterhearing ne of the tapes and learninghowBob checked hetask-likingcale,observers atedhimon 25 Bi-polarAdjective imensions. inally hey ircled hetenadjectives hey onsideredmost elevant ndimportantoformingheir verallmpressionfBob.Pre-TestData

    To resolve omemethodologicalroblems hat seemedcritical,we con-ducted n extensive re-test. irst,we wereconcernedhatwe might istortthe "situated dentity"valuations f Bob by an unintendedias in select-ing the adjectives.Second,we wereskeptical f the effectsf havingob-serversestimateBob's task-liking esponseprior to being told how he"actually" responded o that scale. Therefore, e presented ur tapes to150 high school subjects in classroom administrationsnd used theirresponsesoconstructhefinal ersion f theexperiment.By eliminatingrom he morethan 800 words ontainedn theAnderson(1968) and Gough nd Heilbrun 1965) lists thoseterms hatwereclearlyirrelevantr redundant, e narrowed hepre-testttributenventoryownto 100 adjectivesthatmightpotentially haracterize ob in any of thefour xperimentalonditions. re-test ubjectsheard one of the tapes andwereasked to crossout thoseadjectivesthattheyfound rrelevant; heycircled hetenadjectivesmost mportanto describeBob. Theirresponsesshowed hat sixteen djectivesweredescriptivelyelevant n all conditions,while ixmorewerehighly elevantn one or more ftheconditions.We in-cludedtwomore independentndwarm)tobring hetotal on thefinal ist

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    8/19

    SITUATED IDENTITIES AND THE EXPERIMENT 71to 25. The selected djectiveswere paired with their ntonymso form i-polar scales of 51 points.Secondly,we anticipated hat observersmight ttempt o justifytheirinitial stimate f Bob's task-likingesponseswhen ater confronted ith adiscrepant eport f "Bob-Liking". ince we desiredboth the estimates fBob's task-liking nd the identity eactions o the reportedBob-Liking,we wanted o gather oth f we could avoid contaminatinghe atter.There-forewe variedthe manner n whichfeedback n Bob-Liking either 3 or- 1) was presented.One-third f pre-test ubjectswere asked to estimateBob's task-liking,weretold his actual response,nd then selected djectivesrelevant o de-scribehim.Another hirdwere told to "think bout" (but notwritedown)how Bob wouldhave responded o thetask-likinguestion; then, heyweretoldhow he responded nd asked to pickrelevantdjectives. he final hirdwere mmediatelyold howBob "actually" responded, ithout avingesti-mated or thought bout his feelings.We did not finddifferencesn "rele-vance" selectionswhen comparing he responses f these three feedbackgroups cross the two Bob-Liking onditions.Furthermore,ithin he imited angeof Bob-Likings nvestigatedn thepretest -1, -3), the variation n selection f relevant djectivesby Bob-Likingconditionswas substantiallyess than the Incentive nd Encountermanipulationsroduced. hat is,we found hatBob's response o thetask-liking uestion idnotappreciablyffectheattributeshosen s importantto formingn overall mpressionf him.This tellsus thatBob's reportedresponses much ess importantor mpression ormationhan the circum-stancesunderwhichhe makesthatresponse. hus,we decided that askingexperimentalbservers or rior stimatesfBob's task-likingesponsewouldnot bias their ubsequent valuation fhis reported ehavior.Dependent-VariableifferencesnSociallyDesirableSituateddentities

    Table 1 presents hemeanestimationsf task-likingn each of thefourconditions efinedby Incentive nd Encountermanipulations. he dataclearly eproducehesignificantonditional ifferencesound n theoriginalexperimentaltudiesand the interpersonalimulation f FACE conditions(Bem, 1967). In FACE conditionshere s an inverse elationshipetweenincentivemagnitude nd estimatedask-likingp

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    9/19

    72 SOCIOMETRYTABLE 1

    Means of Estimated Task Liking-By Encounterarnd ncentiveConditionsEncounterIncentive* Face Essay

    $ .50 + .60(7a) - .54(75)$5.00 - 1.13(75) + 1.2 75)*Differencesby Incentivewithineach encounterconditionare significantt the .01level by one-tailedT tests,not assuming qual variances.

    highly ignificantifferencesithin ach condition s a functionf Bob'sreported ask-liking.omparinghese mpressioncoreswith thevalues inTable 1, t is evident hatBob receives hemostfavorablevaluations30.0or above) whenhe reportedlyhecksthetask-likingesponse losest o theestimatedmean. This is true n all conditions.We can geta betterdea of thecorrspondenceetweenBob's dependent-variableresponsesnd his situateddentityfwe calculate valuation coresforonlythe mostrelevant ttributionimensions.ince theadjectives hatobservers icked s critical o formingn overall mpressionfBob differedsubstantiallys a function f Incentive nd Encountermanipulations, ecalculatedBob's situateddentitycoreson theten mostfrequentlyhosenadjectiveswithin ach of thefourbasic conditions. he graphs f Figure1illustrate he resultswe obtained: The top row showsthepercentages fobservers hoestimated ob's task-likingesponse o be at one of thescalepositions rom 2 to +2, and thebottom owpresents hemeanof ob-servers' valuations fBob on the ten"Most conditionallyelevant"dentitydimensions.

    TABLE 2Mean Evaluations of "Bob Downing" on All Adjective Pairs-By Encounter,Incentive nd Bob-LikingConditions

    Bob's ReportedLikingResponse Significance fEncounter Incentive - 2 - 1 0 + 1 + 2 Differences*Face: $ .50 20.6 24.3 28.6 30.5 25.7 p < .01$5.00 24.0 33.6 25.8 23.0 24.3 p < .01Essay: $ .50 24.6 29.4 31.4 30.8 26.3 p < .01$5.00 17.8 23.9 29.4 30.7 28.5 p < .01* There are 15 cases per cell; significanceevels determinedby one-way analysis ofvarianceacrossBob-liking onditions,with (4,70) degreesof freedom.

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    10/19

    SITUATED IDENTITIES AND THE EXPERIMENT 73FACE ncounter: ESSAY ncounter:

    a) Estimated 100-1 100-1Liking of Tasks 60- 60-Incentive: 45- 45|

    $5.00-..-- 30- 3030JI 4-

    -2 -1 0 +1 +2 -2 -1 0 +1 +2Liking Likingb) Evaluations 51-i 51-i--based upon he 10MostConditionally 40- i 40-RelevantDimensions 3- , o

    'A30- --30-Incentive: %No r"$50 U-on 422 o 0-1 o 0-,-2 -1 0 +1 +2 -2 -1 0 +1 +2Liking Liking

    FIGURE 1Distribution f Response Estimations nd Identity valuationsThe criticalcomparisonnvolvesthe strikingly arallel curvesbetweenthetworows.Notethat nall four onditions heestimatedask-likingeaksat the same positionas the most favorable ituated dentity valuation.Judgesdistributestimationsf an individual's esponse o the situationandfavorable valuations f his situated dentityn much the same way.

    Evenstrongerhan nTable 2, One-wayAnalysis f Variance howedhighlysignificantifferencesnmean evaluationsp

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    11/19

    74 SOCIOMETRYTABLE 3

    Mean Evaluations of "Bob Downing" on Justification imension in Face Conditions-By Incentive nd Bob-LikingBob's ReportedLiking Response -Significance ofIncentive -2 -1 0 + 1 + 2 Differences*

    $ .50 14.6 23.2 30.1 32.8 34.3 p < .01$5.00 33.6 37.0 30.4 26.4 25.9 p < .05*There are 15 cases per cell; significanceevels determined y One-Way Analysis ofVariance acrossBob-Liking conditions,with (4,70) degreesof freedom.

    tive and Bob-Liking eedback.n the $.50 condition, ob is seen as morejustified o the extent hathe likes the task, although hemajor drop inhis justificationcore occursonly whenhe marks negative cale position.In the$5.00condition e ismost ustified henhe checks he-1 task-likingresponse nd least justifiedwhenhe has positive ask attitudes. hese rat-ings roughly orrespond o thedistributionf task-likingstimationsFig-ure 1), but they re less closely elated o them han the composite atingsofBob on theten mostrelevantdjectives.We prefer o regard ustifications one dimension f evaluative elevancethat characterizesction n these ituations; ut it is not theonly one,andit is not eventhe mostrelevant ne,as we will show ater. Explaining e-haviorsby reference o a singlevariable s just too simple.Betterpredic-tionsresultwhenthe total configurationf relevant, ituated-identityi-mensions s considered.Afterthe following iscussion,we shall furtherexplore he multidimensionalature f these xperimentalituations.Discussion

    Thus far,we have shown hatobserversre able to consensuallystimatethe responses f a stimulus ersonwho has undergone rather omplexseriesof experiencesn a novel, aboratoryetting. hereby,we have simu-latedthe ncentiveifferencesyFACE andESSAY encountershatpreviousstudieshave found.We also have shown that there s an impressiveor-respondence etween he distributionurves forestimated ask-likingndthosefor he situateddentityvaluations.We propose hatpeopleare ableto maketheseestimationsnd evaluations onsensuallyecause theysharecommonymbolic tructures.t is becausethepotential ask-likingsonveydifferentiallyvaluatedsituated dentitieshatindividuals re able to pre-dict theappropriate esponse;thatis, the one thatis associatedwiththemostfavorable valuation.Fromourperspective,hesimilarityetween he distributionsf identity

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    12/19

    SITUATED IDENTITIES AND THE EXPERIMENT 75evaluationsnd task-likingesponsess causal: Concern o maximize esire-able identity mpressions etermines he dependent-variable,ask-liking.A reversal f thiscausual orderwould nterprethe situated dentityvalua-tions n terms f theappropriatenessf behavior;Bob beingrated highlyfor n expected esponse nd unfavorablyharacterized ordeviance.How-ever,wemustreject his nterpretationecause t does notspecify nybasesuponwhich eopleform heir xpectation,or does t suggest wayto meas-ure them.Measuring hesituated dentities onveyed y behavorial lternativesn-ables us to determine hether here s a "social desirability" ias implicitin the selection f dependent-variableesponses.Our findingshow thatsubjects'actions can be subject to "social desirability"ffects,s are re-sponseson some personalitynventoriesnd attitude cales. From the sub-ject's standpoint,xperiments ayinclude face saving" or "image enhanc-ing" response lternativeshat re unintentionalonsequencesf the experi-mental esign.In the studieswe examined, valuative utcomeswere clearly ssociatedwith hedependent-variablelternatives.ome responses ecidedly ast"BobDowning" nto an extremelynflatteringight,while othersmade him amost appealingfellow.Since evaluativecharacterizationsre drawn fromthesubject'sresponse,nd he can expect hem o be drawn,he may notberespondingrimarilyo the direct uestion: "How muchdid you enjoythetasks?"butrather o the mplicit uestion: "What kind ofperson re you?Indicateby checking task-likingesponse."There are good reasonsforadopting n explanation f the "insufficientjustification"tudies n terms f participants'ituated dentities. ur per-spective xplainsthe estimation ata in all simulation onditions,nd noother heory eally ccountsfor nyof it. In order o extend hisexplana-tion to thepredictionfresults rom he aboratory,t is only necessary oassumethatpeoplesee themselvess others ee them. t seems reasonablethatpersonallynvolved, valuatively pprehensiveubjectswould be evenmore conscious hanobservers f the situated dentities mpliedby theirbehaviors. hus, from he sameperspective e can parismoniouslyxplainboththe simulation nd theexperimentalata. Furthermore,e can dealnot onlywithconditionalmeans, ut also withthe distributionf responses(see Figure1).Situated dentitylso provides methodologyorapproachinghemeas-urement f social situations, perationalizingerms hat symbolicnterac-tionists ave used intuitively ordecades. It quantifies he "meaning"ofsituated ctivity nd social action.The approachyieldsdata to replacethespeculationbout whatexperimentalettings,manipulations,nd measure-

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    13/19

    76 SOCIOMETRYmentsmeantosubjectswho experiencehem; nd, consequently,owexperi-menters ught o interprethem.This sectionhas shown hatdifferentiallyvaluated ituated dentitiesreconveyed y the actionalternatives hat people choose. The next sectionwill demonstratehatchanges n theconditions f theexperimentltertherelevant nd important imensionsf evaluation.Thus, we turnattentionfrom heway situated dentities etermine ependent-variableesponses otherelationshipetween ituated dentities nd independent ariables.Independent ariableDifferencesn Situated dentity ets

    In addition o the evaluative haracterizationssituated dentities) hatresult rom subject's ctions, he social context f thesituation stablishestherelative alience fattributionimensionsor mpressionormation.hisimplies hatthebasic contoursfa social situationanbe describedn termsof theconfigurationf situated dentity imensionshatare important orshaping n actor's mage.And t raisestheprospect hatwe may be able tomeasurethe points of comparabilitynd discrepancy mongexperimentalconditions.Table 4 presents hedifferentialelevance fadjectives o formingverallimpressionsfBob-by Encounterncentive. he left ide ofthe tableshowsthepercent f subjects n each of thefourbasic conditions ho circled nadjective s among heten mostrelevant. he right ide of the tableshowsthesignificancef differencesetween onditionsn the selection f revelantadjectives.The orderingnd inclusion f adjectiveswas determinedy thesignificanceithwhich heydifferentiatedp < .05 orsmaller)between n-counter r Incentive onditions.It is evident hat there re pronounced ifferencesn theidentity imen-sionsrelevant o impression ormation,ach of the fourconditions rovid-ing a distinctiveet. Some of the relevantdimensions onfirmcommonsense"notions bouthowtheexperimentalituations iffer;nd it is com-fortingo have evidence o substantiatehese peculationsboutsituationalmeaning. he attributesffriendliness,armth,leasantness,nd likeabilityaremorerelevant o Bob's situateddentity ormationn FACE encounters,whereas ntelligences more mportantn ESSAY. Similarly,Materialismand Justificationave greater alience with thehigherncentive. romanintuitive tandpointt is not surprisinghat thesefactorsdistinguishheexperimentalituations.What s surprisingnd disturbings thenature ndnumber fdimensionsthat significantlyifferentiatencentive onditions.When differentncen-tive evels lterbasic social mputationsike ntelligence,leasantness,ports-manship, incerity, ositiveness, ooperativeness,lexibility,ndependence,

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    14/19

    SITUATED IDENTITIES AND THE EXPERIMENT 77TABLE 4

    Percent of Subjects SelectingDimensionsAmong the "Ten Most Relevant"to CharacterizeBob-By Encounter and IncentiveConditionsPercent %) of Subjects Selecting Significance fDifferencesDimension as Relevant Between Conditions yChiSquare

    Face Essay Encounter: Incentive:Ditmellsion* $ .50 $5.00 $ .50 $5.00 Face vs. Essay $ .50 vs. $5.00Friendly-Unfriendly 6O% 75% 44% 31% p < .001 nsWarm-Cool 51 51 24 35 p < .001 nsMature-Immature 49 28 21 24 p < .01 nsActive-Passive 20 45 53 51 p < .01 nsLikeable-Disagreeable 66 63 52 53 p < .0.5 nsMaterialistic-Idealistic 15 61 44 68 p < .01 p < .001Intelligent-Simple 41 25 71 31 p < .01 p < .001Pleasant-Unpleasant 68 57 59 36 p < .02 p < .01Sportsmanly-Unsportsmnanly5 20 48 16 ns p < .001Sincere-Insincere 51 47 60 44 ns p < .01Positive-Negative 48 32 45 25 ns p < .01Cooperative-Competitive 48 56 47 75 ns p < .01Flexible-Rigid 55 37 52 40 ns p < .02Justified-Unjustified 37 59 35 40 ns p < .05Independent-Depend-nt 41 43 33 56 ns p < .05Self Confident-Self-Doubting 29 31 21 45 ns p < .05

    * Relevance selectionsweremade by 75 subjects n each of the fourconditions.n addi-tion to the dimensions resentedn the table, two were selectedamong the "Ten MostRelevant" in at least one condition,but were not chosenwith significantlyifferentre-quencies: Dependable-Undependable nd Responsible-Irresponsible.and self-confidence,e are forced oquestion ur ntuitivempressionsboutthe effectshat ncreased ayment should"have. In doing o, some funda-mental assumptions f experimentationre challenged:the comparabilityof conditions nd theirdifferentiationlong a singlevariableof manipula-tive nterest.Discussion

    To assigncausal influenceo differentialeward payment)requires hatall else be heldconstant xceptfor he rewardmagnitudetself. ur identityimputation ata show that changing ne "element" n the social contextchanges hemeaning f the situation nd of actions taken n it. That is,univariate hanges n the stimulus alue (from he experimenter'sointofview) effectmulti-variatehangesn stimulation alues (from ubjects'per-spectives).The "meaning" f subjects'responses hangeswith conditional

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    15/19

    78 SOCIOMETRYmanipulations,ecause of changes n the"kindof person"theybecomebychoosing mong available response lternatives.When themeaning f thesituation nd situated dentity mplicationsfresponses ary n conjunction ithchanges n "the" independentariable,thennot only s thecomparabilityfconditionsuestionable; uttheabilityto assign causal influenceo the intendedmanipulations also weakened.In one sense, xperimentsreatedifferenceopulations f "persons" o par-ticipaten each of theexperimentalonditions y creating ifferentituatedidentity sets".Conventionalchecks"on theeffectivenessfmanipulationsare of littlevalue,because-while assessingwhether t least the intendedeffectccurred-they o notdetermine hethermultiple hangeshave beeneffected.There s a distinct ossibilityhatmany aboratorytudies n socialpsy-chologymaynotbe "experiments"t all: in the sensethatwe cannot at-tribute ausal influence o a singlemanipulated ariablenor specify hatconditions re comparablen all but one respect. f "money"has suchex-tensive ffects,t seems ikely hatotherforms fdifferentialeward socialapproval, lectrichock, tc.) wouldgenerateven morewidespreadhanges.On the otherhand, t is possiblethat thesemultivariateffectsccuronlyin certain ypesof experimentalituations. or sometasks and responses,the situateddentityetsofsubjectsmaynot affect redictionsr outcomes.The possibilities eserve eriousconsiderationnd empirical xplorationnreplicationsnd theplanning ffuturetudies.The issuesof comparabilitynd causalitythat we have raisedare notirresolvable,utto confronthem irectlyequires revisionfexperimentalpractices.Ratherthanassuming hatwe, as experimenters,now the stim-ulation alue andmeaning f ourmanipulationso subjects, ata ought obegathered n the effects uch manipulations ave. We may thenfindthatapparently nrelatedmanipulationsroducethe same resultsbecause theyeffecthangesn the same dentity imensions.Forexample, emightwanttoknow heeffectsf ncreasedjustification"for ounter-attitudinaltatementsn task-likingesponses. he present atado not allowus to separatethose effects rom ncreasedmaterialism,e-creased incerity,nd so on.However,we could ustify ob's behaviorwith-out involving ay: He couldbe "required" o describe he task favorably;he couldbe askedtohelptheexperimenterut of the bind forfree;he couldbe told thatthewaiting ubject ouldnotparticipatendwouldnotbe paidif someonedidn'tgivehimtheproper xpectational et. Probably, ny ofthesewill ustifyhis favorable ask-descriptionnd have additional ffectson the relevance fsituateddentityttributes.ut a series fsuchmanipu-

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    16/19

    SITUATED IDENTITIES AND THE EXPERIMENT 79lations hould ffect he salience fdifferentttributes,eaving hehighrele-vance of justification airly onstant.Such a research trategywould show eitherthat justification ad con-stant effects, espitewide variations n the salience of other ttributes,rthat ts effects ere contingentn otherdimensions eing salient. n anycase, this approachwould provide systematicmeans for measuringitua-tional context nd assessingthe total effects f the independent ariablemanipulations.t would also direct perational ttention o situated dentityvariables hat re usually reated iscursively hen nterpretingxperimentalresults.

    SUMMARY AND CONCLUSIONSThe situated dentity erspective onceives f individuals s structuringcomplex ctivities nd stimuli equences n terms f their mplication orparticipants' ispositional ttributes.When members f a givenpopulationconsensuallynvoke he same dimensions o characterizectors, he attribu-tional rocess reates ocial ituations. ehavior s predictedf their s a "mostfavorable"ituated dentity o be gainedfrom ne of the behavorial lterna-tives hat repossibleunder he circumstances.eople are expected ochoosetobecome he kindof person hatwill be mosthighly alued.To test these deas we conducted n interpersonalimulation, atheringdata not only on estimated esponses ut also on the situated dentitiescreatedby the stimulus-personhose responseswere observed. he estima-tionresults eplicated he relationshipsoundna series f "insufficientusti-fication" tudies nd werepredictable rom valuative mpressions. ttemptsto reconcile hese resultswith dissonance nd incentive xplanations avebeenunsatisfactorynd inelegant.We propose that subjects nd observersattach the samesituationalmeaning o experimentalonditions nd behav-iors n them, nd this enables us to accountforthe entire eriesof resultsfrom single erspective.We conceptualizehepredictivelymportantariables t a differentevelfrom revious heories: n the social situation reatedby each conditionof an experimentur concern s withthe consensual ttributions.n the ex-periments e considered he dependent-variableesponseswere associatedwithdifferentiallyesirable ituated dentities. hus, the resultsobtainedunder heseconditionsouldbe explained y the "social desirability" f re-sponse lternatives ithout eferenceo thevariables nd hypothesesorm-erly mployed. o the extent hat desirable ituated dentities re associatedwith heresponses erived rom particular heory,xperimentalata shouldshowthe "right" predicted) resultsfor the wrongreasons.The apparent

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    17/19

    80 SOCIOMETRYconfirmationf hypothesesmay be producedby subjects' concerns boutself-presentation.At present,we can only wonderhow many studies n the experimentalliterature f social psychology ave achieved theirresultsbecause of thesituated dentities t stake. Some hints of how widespread hese effectsmight e come from ecent tudies n such established reas of investigationas game playing Alexander nd Weil, 1969), groupdecisionmaking Ma-daras and Bem,1968), and role-conflictesolutionAlexander nd Epstein,1969). Furthermore,iventhe evaluative pprehensionnd "good subject"orientation ith which most subjects enter the laboratory,t seems quitelikely hatprevious esearch n manyother reas of social psychologicaln-vestigations implicated.2The situated dentity pproach rovides meansfordetecting hese nflu-ences n experimentalituations. rior simulations f intended xperimentscould determinehe nfluence f situated dentitiesn the various onditionsand gatheringhiskindof data as part of the experimentalrocedurewouldyieldvaluable nformationbout subjects'definitionsf the situation.Asidefrom ny corrective unctionshat a searchfor ituated dentitymight ervein the interpretationf experimentalesults, he approachmay providescientificallyatisfactory eanstomeasure he situational conditions nderwhich"responses re obtained.Ifwe arecorrectnassuminghat ituated dentitiesefine hemeaning fsocial settingsnd describe he orientations f actorsto conduct n them,thenthepredictiveealmof theperspectivextends arbeyond he abora-tory.What we need to understandre the processesby which dentitiesare created nd transformedo that t willbecomepossibleto specify owparticular ocial identitiesnfluence ctors' behaviors ndervarying itua-tional onditions. he conceptualndmethodologicalpproach hatwe havepresentedn thispaper shouldprovide he means to proceedwiththis task.

    2 The findingswe present are even more distressing ecause the dependent variableresponses hat produce such dramatic changes n identity valuationare relatively non-impactful" pencil-and-paperbehaviors. If an ostensibly"trivial" report of task-likingon a post-experimentaluestionnaire as such substantial mpact upon the kindof persona subject becomes,we can only guess at the effects f elicitingmore impactful nd in-volving responses, s Aronson nd Carlsmith 1968) recommend.REFERENCES

    Abelson,Robert P., Elliot Aronson,WilliamJ. McGuire,TheodoreM. Newcomb, MiltonJ. Rosenberg nd PercyH. Tannenbaum eds.)1968 Theories of Cognitive Consistency:A Sourcebook. Chicago: Rand McNally.Alexander,C. Norman, Jr. and HarrisonG. Weil

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    18/19

    SITUATED IDENTITIES AND THE EXPERIMENT 811969 "Players, persons, and purposes: situational meaning and the prisoner'sdilemma game". SociometryVol. 32 (June):121-144.Alexander, . Norman, Jr. and Joyce Epstein1969 "Problems of dispositional nference n person perception research". Soci-ometryVol. 32 (December) 381-395.Anderson, orman H.1968 "Likableness ratingsof 555 personality-trait ords". Journal of Personalityand Social Psychology (July):272-279.Aronson, lliot and J. MerrillCarlsmith1968 "Experimentation n social psychology".Pp. 1-79 in Gardner Lindzey andElliot Aronson (eds.), The Handbook of Social Psychology. Volume II.Reading,Mass.: Addison-Wesley.Bem,Daryl J.1967 "Self-perception: an alternative interpretationof cognitive dissonancephenomena". Psychological Review 74 (May): 183-200.Brehm,Jack W. and ArthurR. Cohen1962 Explorations n Cognitive Dissonance. New York: John Wiley & Sons.Carlsmith,J Merrill,Barry E. Collins and RobertL. Helmreich1966 "Studies in forced compliance: I. the effect f pressurefor complianceonattitude change produced by face-to-facerole playing and anonymousessay writing".Journal of Personality nd Social Psychology4 (January):1-13.Festinger, eon1957 A Theory of Cognitive Dissonance. Evanston, Illinois: Row, Peterson,&Co.Festinger, eon and J. MerrillCarlsmith1959 "Cognitive consequences of forced compliance". Journalof AbnormalandSocial Psychology 8 (January) 203-210.Friedman,Neil1967 The Social Nature of PsychologicalResearch.New York: Basic Books.Goffman, rving1959 The Presentation of Self in Everyday Life. Garden City, New York:Doubleday.1961 Encounters: Two Studies in the Sociology of Interaction. Indianapolis:Bobbs-Merrill.1963 Behavior in Public Places. New York: The Free Press.Gough,HarrisonG. and Alfred . Heilbrun,Jr.1965 The Adjective Check List Manual. Palo Alto, California: ConsultingPsychologists ress.ILeider, ritz1958 The Psychology of InterpersonalRelations. New York: John Wiley &Sons.Helmreich, obert and BarryE. Collins.1968 "Studies in forcedcompliance: commitmentnd magnitudeof inducementto comply as determinants f opinion change". Journalof Personality ndSocial Psychology10 (September):75-81.Jones, dward E. and Keith E. Davis1965 "From acts to dispositions:the attributionprocess in person perception".Pp. 220-266 in Leonard Berkowitz (ed.), Advances in Experimental ocialPsychology.VolumeII. New York: AcademicPress.

  • 7/30/2019 Alexander y Knight - Situated Identities and Social Pshychological Experimentation.pdf

    19/19

    82 SOCIOMETRYKelley,Harold H.1967 "Attribution heory in social psychology".Pp. 192-238 in David Levine(ed.), Nebraska Symposiumon Motivation. Lincoln, Nebraska: University

    ofNebraska Press.Madaras, GeorgeR. and Daryl J. Bern1968 "Risk and conservatismn group decision-making". ournal of ExperimentalSocial Psychology4 (July) 350-365.Mead, GeorgeHerbert1934 Mind, Self, and Society. Chicago: University f Chicago Press.Riecken,HenryW.1962 "A programfor researchon experimentsn social psychology".Pp. 25-41in Norman Washburne (ed.), Decisions, Values, and Groups. Volume II.New York: PergamonPress.Rosenberg,MiltonJ.1965 "When dissonance fails: on eliminating pprehensionfrom attitude mea-surement". Journal of Personality and Social Psychology 1 (January):28-42.1969 "The conditions and consequencesof evaluation apprehension".Pp. 280-350 in Robert Rosenthal and Ralph L. Rosnow (eds.), Artifact n Behav-ioral Research.New York: AcademicPress.Rosenthal,Robert1966 ExperimenterEffects n Behavioral Research. New York: Appleton-Cen-tury-Crofts.Rosenthal,Robertand Ralph L. Rosnow (eds.)1969 Artifact n Behavioral Research. New York: Academic Press.Strauss,Anselm1959 Mirrors nd Masks. Glencoe: The Free Press.