Sie sind auf Seite 1von 20

TheEffectsofLanguageBiasonEvaluationofPoetryforPersonal

Relevance
GraceHartley,ErikDant,IsabellaMori

Abstract
Overthepastmonthandahalf,weasked:howdoeslanguagebiasaffectperceptionof
personalrelevanceinatextwritteninanonnativelanguage?Wewantedtoknowiftelling
someoneapieceoftexthadbeentranslatedfromanotherlanguagetheywereunfamiliarwith
wouldhaveanyeffectontheirperceptionofit,asopposedtolettingthemassumeitwaswritten
intheirnativelanguage.Forourtext,weusedliteraturethatisintendedtobeemotionally
meaningful:poetry.Wemeasuredtheirresponseofperceivedmeaningfulnessthroughabasic
15scale,1beingmeantnothingtomeand5meaningitwasverymeaningfultome.Our
surveyaskedthemmultiplequestionsdesignedtogaugetheirresponsetothetext,alongwitha
fewotherquestionstotrytoeliminatebiasesandconfoundingvariables.
Ourmaininterestwastoobserveifpeopleclosethemselvesofftowrittencommunication
becauseofananticipatedbarrierwhilelanguagebarrierscertainlydoexist,wewantedto
observehowobstructivepeopleperceivethemtobe.Wewereinterestedtofindiftherewas
evidencetoshowthatthetextwasfarlessmeaningfultosampledpeopleoncetheyhadbeentold
itwastranslated,wewouldhavebeenabletoinferthatalanguagebiaswouldpreventpeople
fromacknowledgingthevalueoftextsbecauseoftheculturetiedtothem.

Toconductthesesurveys,weusedawebsite(SurveyMonkey)andsharedthelinkstoour
surveysonPSUstudentwebpages,andourownpersonalfeedsoverFacebook.Wehadtwo
surveys,acontrolsurveywithquestionsaboutapoemwithoutatranslationlabel,andasurvey
withapoemlabeledastranslatedfromGermanfortheexplanatoryvariable.Ourdatawas
inconclusive,anddidnotsupportthepresenceofaclearlanguagebias,thoughtheinformality
andsmallscaleofthestudycouldhaveaffectedthedataasfarasstatisticalsignificancegoes.

Introduction
Itwasimportanttodefinemeaningfulnessforourhypothesis,asitcanbedifficultto
quantify,anditisimportanttoexplainwhytranslationmayaffectapoems(oranytexts)
perceivedvalue.Firstly,emotionalresonanceormeaningisthe
intention
oftheworkwecould
havechosenaninformationaltextandaskedthemhowinformativeorcredibletheyfoundit,ora
satiricaltextandaskedthemhowhumoroustheyfoundit.Wechosepoetry,asitisintendedto
resonatewiththereader,tomeasuremeaningfulnessintheformofliteraryexperience.As
Diltheyputsit,...Whatweseekfrompoetryispowerfulstimuli...thepurposeofpoetryis
toexpandthescopeofourlivedexperience...(1985,p.6).Thus,inthiscontext,theperceived
meaningfulness,personalsignificance,emotionalresonance,etc...isthepointoftheworkwe
attemptedtoseehowtranslationwillinfluencetheefficacyoftheliterature.Thisexplainshow
perceivedsignificancefunctionscontextuallyinthestudy.Thesecondimportantaspectis
contentrelatedmorespecificallyhowdowequantifysignificance?
Inpoetry,meaningcanbeasomewhatabstractconcept.Diltheydescribesitasthe
recognitionofhowthingsconnectandrelatetooneandanother,stating...themostimportant

part...theprocessofcompletionwherebytheimaginationdrawsonthefullnessandrichnessof
thisoverallnexus...weobtainfromimagesandconnectionswhatisessentialaboutastateof
affairs:whatgivesititsmeaninginthenexusofreality,(Dilthey,1985,p.8).Thisbeingsaid,
howthepieceresonateswiththereaderandtheirpersonalinterpretationofitisnotsomethingwe
canmeasureorquantifyweareonlyconcernedwiththepresenceofmeaning,notwhatthe
poemmeanstothem.Wemeasurethissignificancethroughtwoquestionsonthesurvey
(answeredona15scaleindicatinghowsignificanttheyfoundit),whichweaverageouttofind
acomparativemeaningfulnessscore.
Ofcourse,thereareafewissueswiththisapproachandwithsuchsubjectivedata.Itis
possiblethesubjectcouldbesomeonewhodoesntvaluepoetryatall,andthetext,regardlessof
translation,doesntgarneranymeaningforthem.Orperhapsthepersondoesgenerallyvalue
poetrybutthatparticulardaytheyarentinareceptiveplaceforliteraryexperience.The
biggestissueisthatthecontextinwhichtheyreadthepoemisntconducivetoanemotional
response.OnFacebook,wherewedistributedoursurvey,therearenumerouspoliticalandsocial
activistsandpeopleintransitionwhoneedassistance,whichsaturatesthemediumwithsurveys.
Thepresentationofthispoemisinthisform,whichisunremarkableintheculturalcontextsoas
toprovelessswaying.However,thesevariablesthataffectperceivedmeaningapplytoboththe
translatedanduntranslatedpoem,meaningthedataisconsistentasfarastheexplanatory
variableisconcerned.
Itisalsoimportanttogivetheexpectationsofourhypothesis:
Wepredictthatapersons
perceptionofatextchangesifthetexthasbeentranslated,orisbelievedtobeforeign,making
themlessreceptivetothetextanditsconcepts.
Forthishypothesis,wewantedtoarguethatthe

mereimplicationofanideahavingforeignoriginswouldinduceareadertobelessreceptivetoit
becauseoftheirsociolinguisticidentity.Gumperzputsforth,(1982)thatidentityandperceived
identityareshapedbyourcommunication,andthevitalrolelanguageplaysnotonlyinsociety
butinideologicaldevelopment.Thisbookisbasedoffofseveralobservationalcasestudies
aidedbytaperecordersandphysicalobservance.Avarietyofsituationsdealingwiththe
interactionofsocialideologywiththelanguagecommunicatingitwereanalyzed.Thoughitis
nottheaimofthebook,theinseparablerelationshipoflanguageandideologyisheavily
dependedon,andlanguagenotonlyasawayofcommunicatingbutasawayofthinking.We
arguethatitisthisbondoflanguagetoideologywhichwould,onsomelevel,causesomeoneto
rejectanideabecauseitwasnotconceivedintheirpersonallanguage,thatthedifferent
languagesimplydifferentsocialidentitiesandwaysofthinking,whichpossiblyobstruct
relatability.Basically,webaseourhypothesisoffthepossibilitythatifanideawasconceivedin
aculturewedontrelateto,bynature,wedontrelatetotheideaaswelleither.Andoneonly
needstothinkthereisaforeignelementtoquestionit,likelyonasubconsciousleveland
sometimesonaconsciouslevel.However,thisisnttosayanideawillbecompletelyrejected
ourhypothesisismoreaccountableforaminorreadjustmentinevaluationasopposedtoa
refusaltoevaluate.
Anotherstudy

(Maass,AnneSalvi,DanielaArcuri,LucianoSemin,GnR.,1989),
functioningmorethroughperceivedidentityandstereotypes,exploressimilarthemes.There
weretwoexperimentsconductedinwhichtheresearchersattemptedtofindevidencethatpeople
weremorelikelytoperpetuateundesirabletraitsofstereotypesthangoodones.Whatmakesthis
relevantisthemodeofperpetuationresearcherswereconcernedwiththepoweroflanguage,

andhowitinfluencedthespeakersandlistenersideaofthemselves.Thisstudyofseveralsmall
socialgroupsandinteractionsdiscoveredthatlanguagecreatesaselfperpetuatingcycleof
stereotypesitisthisrelationshipoflanguagetoidentitythatweareconcernedwith,andthese
tenuousideaswillbetested.
Anothersource,moreintroductorytoourproject,isfromtheInternationalJournalof
Epidemiology.Thisretrospectiveobservation(Juni,P.,Holenstein,F.,&Sterne,J.,2002),has
manyparallelideaswithourownstudy.Theseresearcherslookedintoamedicaldatabaseof
studiespublishedinEnglish,thenstudiespublishedinotherlanguages,toseehowtheir
translationsaffectedtheiruse.Itisaverybriefarticlesummarizingbackground,methods,
results,andconclusionsoftheirobservation.NotonlydidtheconceptsofeffectsofEnglishvs.
nonEnglishsourcesgiveusgoodreference,butthestatisticaltoolsemployedfamiliarizedus
withtheimportantstatisticalconceptsweusedinourstudy.
Themotivation,asstatedabove,isconcernedwiththepossibledenialofliterary
experienceduetolanguagebias.Becausesubjectivelyandobjectivelyvaluableliteratureis
producedbynearlyeverycultureinalmostalllanguages,thosewhospeakonlyoneortwo
languages(amajorityoftheworld)couldbeclosingthemselvesofftoallkindsofideas,
information,orwaysofknowing.Evenonceatextistranslatedtoalanguagetheyunderstand,
theycoulddiscountitsworthbecauseofthetranslationprocessorthefactitcamefromanother
culturetheymaynotunderstand.Perhapstheyoverestimatehowmuchislostintranslationor
thereissomesortorskepticismtowardsthesource.Insomecases,thismakessensethereare
manywordsthatcaptureuniqueculturalconceptswedonthaveEnglishwordsfor,andone
shouldoftenbecriticaloftheirsourcesofinformation.However,thisdoesntjustifyalanguage

bias,aspeoplewilljudgeethnocentricallyorwitharacistviewattimes.Unfortunately,ourstudy
hasntbeenrefinedtodistinguishthesefoundedbiasesfromunfoundedbiasesweareonly
lookingforthepresenceofabias.Consideringthewealthofknowledgeavailabletous,itis
importanttoaskifwecloseourselvesfrompossiblyenrichingmaterialsbecauseofaninherent
languagebias.
Anexampleofanunfounded,unreasonable,dangerouslanguagebiasisexhibitedin
anotherexperiment(CookieW.StephanandWalterG.Stephan,1986).Theresearchers
observedracistjudgementsandlanguagebiasinjuriesbasedonthenativelanguageofthe
defendant.Thesetwostudieswereperformedinasimilarway:whenadefendantpresentedin
eitherSpanishorThai(aidedbyaninterpreter)toagroupofcollegestudentsactingasjury,
EnglishspeakingstudentsconsideredthemguiltierthaniftheirtestimonywereinEnglish.When
toldtoignoretheinterpretation,thejurybecamelessbiased.Thisstudyisaveryimportant
exampleinhowdangerousalanguagebiascanbeforourquestion,themostthatisatstakeis
personalenrichment,whereastherecouldbepeopleinsocietyservingunderservedtimeina
correctionalfacility.
Afinalstudy(Egger,1997)wasconductedandshowedthatthereisagreaternumberof
statisticalrandomised,controlledtrialarticlesinEnglishthanthereareinGerman(62%
comparedto35%),despitethefactthattheoriginaltrialswereconductedinGermanspeaking
partsofEurope.ThesenumberswerefoundbyidentifyingthenumberofRandomisedControl
Trialswithlowpvaluespublishedineitherlanguage.Thearticlewentontoshowthat
significantfindingsweremorelikelytobepostedinEnglishthaninGerman,thoughtheywere
conductedanddistributedinWestern,GermanicEurope.Thestudyconcludedthat...
English

languagebiasmay,therefore,beintroducedinreviewsandmetaanalyses...(Egger,1997),
though,unfortunately,itdidnotincludehowthismayaffecttheconcernedfields.However,in
thecontextofourhypothesis,inwhichweonlysearchforthepresenceoflanguagebias,this
missingdetaildoesnotnegativelyinfluencethestudy.
Theaimofourworkistodiscoverthepresenceandpossiblestrengthoflanguagebias
towardanativelanguagewhenevaluatingtext.
Wepredictthattheperceivedpresenceofa
translationinatextaffectsmeaningfulnessandpersonalrelevanceforthereaderwhen
evaluatingpoetry.Seeinghowdrasticallyormoderatelyitaffectstheevaluationofthetext(ifit
affectsitatall)willhelpgaugehowbigaroletranslation,andinextension,theperceptionof
foreigninfluenceonanidea,affectsonesevaluationofit,andifthereisaclearpreferencefor
ideasconceivedinthesubjectsnativelanguage.

Methods
Weareconductinganexperimentalstudy,becausewearentlookingintothepastfor
causaldata.Oursubjectswillbeobservedbyusinthepresent.IfourhypothesisisHowdoes,if
atall,theperceptionofanativeversustranslatedtextaffectsmeaningfulness?thenournull
hypothesisisThereisnotarelationshipbetweenwhetherornotatextistranslated,andits
interpretationofvalue.Ourexplanatoryvariablesaretheimplementationofdifferentlanguages
fortheoriginaltexttellingthesubjectitwastranslatedfromGerman.Ourresponsevariable
willbethelevelofpersonalidentificationandrespectforthetextthesubjectrespondswith.
Ourstudypopulationforthisexperimentwasthe
Facebookcommunity,andoursample
wastheFacebookusersassociatedwithourpersonalFacebookaccountswhochosetotakeour

surveys.Wechosethispopulationpartiallyoutofconvenience,andpartiallytoseehowthisdata
willrelevantlyaffectus,theresearchers,consideringwefunctionwithinthispopulation.To
surveyourtargetedpopulation,wecreatedanonlinesurvey(surveymonkey.com)anddistributed
itoverourFacebookpages,postingcontrolsurveysfeaturinganuntranslatedpoemonone
researcher,ErikDantspage.Wepostedanexplanatorysurveyfeaturingapoemlabeledas
translatedfromGermanonanotherresearcher,GraceHartleyspage.Wepostedthemonlineto
reachthesamplemorequicklyandeasily,andtoavoidtarnishingsubjectsopinionsofthepiece
byconfrontingthemwithaclipboard,astheiropinionsmighthavebeenifwehadhandedout
physicalsurveys,whichwouldhavealsobeeninfluencedbywherewehandedthemout,when,
whowechosetohandthemto,ourappearanceaswehandedthemout,theappearanceofthe
survey,etc.,whileonlinesurveysareopentomorepeopleandarefarmoreanonymous.Ouraim
wasasampleof50responsesforeachsurveywegot22responsesforthecontrolsurveyand26
fortheexperimentalatthetimewebegancalculatingdata.
Thesesurveys,includedintheappendix,wereformattedwithapoematthetopfollowed
by34questions.Bothsurveyshadtwoquestionsintendedtomeasureperceivedmeaning:1)
Howwelldoesthistextresonatewithyou?
and2)
Howlikelyareyoutomakedecisionsbasedon
theseconcepts?
Thesequestionshadascaleof15forthemeasuredresponse,1beingdidnot
resonateatalland5beingwasverymeaningful.Eachpoemalsohadaquestionmeanttodeal
withthepossibilityofpreviousexperienceandattachmenttothepoem(whichmayinfluencethe
perceivedmeaningfulnessorsignificance):
Haveyoureadthispiecebefore?
Thecontrolsurvey
hadapoemwithnolabel,followedbythesefourquestions.Theexplanatorysurveyhadthe
samepoemwiththeaddedprefacethatithadbeentranslatedfromGermanandanadded

questionattheendmeanttodealwiththevariableofpreviousexperiencewithGerman(which
wouldlikelyinfluencetheirviewofthetranslation).Bothsurveyshadthesamepoem,

Ifthere
blewawind,/Icouldhoistasail./Weretherenosail,/Iwouldmakeoneofsticksandcloth
(Brecht),sothattherewouldbenoconfoundingvariableofinterpretingeachrespectivepoem
differentlywewantedthetranslationalonetobethedifference,notthetext.
Forthetwoquestionsintendedtogaugeperceivedsignificance,thenumericscaleforthe
responsewaschosensomeaningfulnesscouldbeeasilyquantifiedforthestudy.Tocreatea
roughmeaningfulnessscore,theresponsestothetwoquestionswereaveraged.Forexample,if
someonerespondedwitha3toQuestion1(moderatepersonalsignificance),anda1toQuestion
2(lowpersonalsignificance),wewouldaddthesenumberstogetherforasumof4anddivideit
bythenumberset(thereare2numbersinthissetofdata)forameanof2,whichisa
lowmoderatemeaningfulnessscore.Thesescoreswerethenusedtogenerateour5number
summariesforaboxplot,mean,standarddeviation,standarderror,pvalue,andconfidence
intervalsinMicrosoftExcel.

Results
Forourcontrolsurveyweobservedameanscore(theaverageoftheaveraged
meaningfulnessscores)of3.14withastandarddeviation(generaldistanceanddistributionof
dataawayfromthemean)of0.6004.Theexplanatorysurveyhadameanof2.9733anda
standarddeviationof0.51258,whichmadeitabitlowerand,becauseofitssmallerstandard
deviation,moreconcentrated.TheseaveragesareshowninFigure1.Thefivenumber
summaries,illustratingthedistributionofthedata,wasusedtomakeavisualrepresentation:box

andwhiskerplots.TheshapeofthedataforthecontrolsurveycanbeseeninFigure2,andthe
shapeofthedatafortheexplanatorysurveycanbeseeninFigure3.
Fig.1

Fig.2


Fig.3

Thesetablesshowthedistributionofthemeaningfulnessscorestheyshowroughlythe
rangeoftheanswersabouthowmeaningfuleachsubjectperceivedthepoemtobe.Thereare
onlyveryslightdifferencesbetweenthetables,whichmeanstheanswerswereclosetothesame
overallbetweenthetwosurveys.Thebox(showingquartile1andquartile2separatedbythe
median)issmallerinFig.3becausethedataismorecondensed,andbecausewehavethesame
numberofresponsesforeachsurvey,weknowthatmeanspeopleansweredwithinasmaller
rangeofnumbersmorefrequentlyintheexplanatorysurveythaninthecontrolsurvey.The
medianineachboxplotisequaltotheupperquartilerange,soitdoesnotappearontheboxand
whiskerplotthisisbecausetheunitshavebeenaveragedoutandcondensed.Overall,this
meansthesecondfiguregotmoreconsistentnumbersthattendedtobeclosertothemedianand
slightlyloweroverallscores.Whetherornotthescoresaredifferentenoughtoshowthe
presenceoflanguagebiasisshownbythepvalueandconfidenceintervals.
Theconfidenceintervalistheareaaroundthesamplemeanthatmostaccuratelyreflects
themeanofthepopulation.Forexample,themarginoferrorofthecontrolsurveydatais
0.12800whichmeansthatbecausewegot3.14asthemeanofoursample,theactualmeanof

thepopulationisanywherefrom3.012to3.268thisistheconfidenceinterval.Themarginof
errorfortheexplanatorysurveydatawas0.10052whichputtherangeforthetruepopulation
meanfrom2.872to3.073.ThesevaluesarealsoillustratedinFigure1.Thepvalueisconcerned
withhowmuchtheseoverlapthelowerthepvalue,thelesstheyoverlap,andthelessthatthe
populationaveragesoverlap,thebiggerthedifferenceisbetweenthedatayoucollected.Thisis
calledstatisticalsignificance.Forexample,ifthecontrolsurveyhadapopulationmeanrangeof
4.35,thatmeansitwouldhavehadveryhighmeaningfulnessscores,andiftheexplanatory
surveyhadgottenapopulationmeanrangeof1.62.3,thiswouldhavemeantitwouldhavehad
verylowmeaningfulnessscorestheywouldhaveshownacleardifferenceandthiswouldbe
statisticallysignificant.However,ouractualscoreswereveryclose,withzscoreof
1.02175,
andalowpvalueof
0.16(alowpvalueislessthan0.5),meaningtherewasnostatistically
significantdifferencebetweenthem.

Discussion
Asstatedabove,ourpvaluewasquitehighsoournullhypothesisisverypossible,and
ourdataisnotstatisticallysignificant.Thoughtherewereminordifferencesinthedatasets,they
werenotenoughtosupportourhypothesistheaveragevaluesoftheexplanatorysurveydata
wasnotlowenoughcomparedtothecontrolsurveytosayanythingforcertain.Thismeansthat,
asfarasthisstudyisconcerned,thereisnoclearlanguagebiasthatwouldcausepeopleto
perceiveatranslatedpoemasanylessmeaningfulthanapoemwrittenintheirnativelanguage.
Thetranslationdidnotsignificantlyaffecttheirperceptionofthetext.

Theresultsofthisspecificstudy,however,maynotaccuratelyreflectthepossible
languagebiasoftheFacebookcommunity.Thoughwewerethoroughinoursearchfor
confoundingvariables,andotherissues,therearecertainlyunidentifiedfactorswemaynothave
consideredthatwouldskewourresultsandpreventthemfrombeingaccuratetothepopulation.
Themostobviousissueinourstudyisthesmallsamplesizewehad40participants.Hadwe
gottenalargersample,wecouldhavetakenmoredatain,whichwouldevenmoreaccurately
reflectthebiasofthecommunity.
Anothersignificantflawmayhavebeenintheformulationofourhypothesisperhaps
weshouldhaveclarifiedthebiasinquestion.Forexample,wecouldhavetestedtoseeifpeople
thoughtEnglishwasthebestlanguageforacademicwriting,orifFrenchwasthebestforpoetry.
StephanandStephan

(1986)

revealedthatpeopleassociatedSpanishandSpanishspeakerswith
criminality,assumingEnglish(andthosewhospokeit)wasmorehonestandmoral.
Theseissuescouldbeimprovedbylargersamplesizesandmorerefinedcorequestions
overalongertimeperiod.However,wecouldhavetestedthesamehypothesiswithmorediverse
examplesandsurveyswecouldhaveaddedasurveywithinformationaltextandaskedthem
howinformativeorcredibletheyfoundit,andanotherwithsatiricaltextandaskedthemhow
humoroustheyfoundit.Thisway,wewouldhavemultiplemethodsandtextsformeasuringthe
ideathattranslationcouldaffecttheperceivedefficacyofliteratureoverallasopposedtojust
poetry.Otherfactorstoconsidermaybethepopulationduetoitsdiversity,theFacebook
community,namelythegroupsassociatedwithus,maybelesslikelytohavealanguagebias,
andwemayhavehaddifferentresultsifwewereinconservative,homogenousgroups.Andof
coursethelanguageitselfwillhaveaneffectthestereotypesassociatedwiththelanguageand

howthestereotypesinteractwiththetextareimportant,liketheconceptofcriminalityin
Habla
Ingles?TheEffectsofLanguageTranslationonSimulatedJurorDecisions
(Stephan,C.and
Stephan,W.G,1986),ortheideathatEnglishismoreprestigiousorbetterforintellectual
reportsthanGermanin
LanguagebiasinrandomisedcontrolledtrialspublishedinEnglishand
German
(Egger,M.,ZellwegerZhner,T.,Schneider,M.,Junker,C.,Lengeler,C.,&Antes,G.,
1997).Thesechangestocontentofthetext,contextofthelanguageandpopulation,andthe
lengthandthoroughnessofthestudycouldpossiblybringustomorestatisticallysignificant,
reliable,accurateresultsinfutureattempts.

Conclusion
Thekeyresultsherewerethatthedatawasnotstatisticallysignificantanddidnot
supporttheideatherewasalanguagebiaswhenevaluatingpoetry.Ouronlinesurveyfoundthat
theperceivedmeaningfulnessoftranslatedvs.untranslatedpoetrycollecteddatathatwastoo
ambiguoustoclaimthetranslationmadeadifference.Ourintentioninfindingwhetherornot
peopledeniedthemselvespotentiallyenrichingliteraryexperiencessimplybecauseitwasnot
writtenintheirownlanguagerevealedtherewasnosignificantbias.Atthistime,ourstudy
supportstheideathatthoseintheFacebookcommunityareopenornonreceptivetothese
experiencesregardlessofwhatnativelanguageitwaswrittenin.Becauseoftheflaws,narrow
scope,andprobablecalculationerrors,thisstudyunfortunatelyprovidesverylittlenew
informationinthelanguagebiasfield.Despiteitslackofhelpfuldata,themainbenefitofthe
studywastheexperienceinresearchandstatisticalanalysisitprovidedforus.

WorksCited

Brecht,E.B.(n.d.).BuckowerElegien.RetrievedMay12,2016,from

http://www.brechtweigelhaus.de/elegien.htm

Dilthey,W.,Makkreel,R.A.,Rodi,F.,Dilthey,W.,&Dilthey,W.(1985).
Poetryand
experience

(Vol.5).Pinceton,NJ:PrincetonUniversityPress
Egger,M.,ZellwegerZhner,T.,Schneider,M.,Junker,C.,Lengeler,C.,&Antes,G.(1997).
LanguagebiasinrandomisedcontrolledtrialspublishedinEnglishandGerman.The
Lancet,350(9074),326329.doi:10.1016/s01406736(97)024197
Gumperz,J.J.(1982).
Languageandsocialidentity
.Cambridge:CambridgeUniversityPress.

Juni,P.,Holenstein,F.,&Sterne,J.(2002).InternationalJournalofEpidemiology.Retrieved
April19,2016,fromhttp://ije.oxfordjournals.org/content/31/1/115.short
Maass,AnneSalvi,DanielaArcuri,LucianoSemin,GnR.
Languageuseinintergroup
contexts:Thelinguisticintergroupbias.
JournalofPersonalityandSocialPsychology,
Vol57(6),Dec1989,981993.
Stephan,C.andStephan,W.G.(1986),HablaIngles?TheEffectsofLanguageTranslationon
SimulatedJurorDecisions.JournalofAppliedSocialPsychology,
16:577589.doi:10.1111/j.15591816.1986.tb01160.x

Appendix
FormatforControlSurvey
Pleasereadthefollowingtext,thenanswerthequestions.

Ifthereblewawind,

Icouldhoistasail.

Weretherenosail,

Iwouldmakeoneoutofsticksandcloth.

*
1.Howwelldoesthispieceresonatewithyou?

1It
doesn't
resonate

1It
doesn't
resonate

5It
completel
y
resonates

5It
completely
resonates

*
2.Howlikelyareyoutomakeadecisionbasedontheseconcepts?

1Not
likelyatall

1Not
likelyatall

*
3.Haveyoureadthispiecebefore?

Yes

No

5Very
likely

5Very
likely

FormatforExplanatorySurvey
Pleasereadthefollowingtext,whichwastranslatedfromGerman,thenanswer
thequestions.

Ifthereblewawind,

Icouldhoistasail.

Weretherenosail,

Iwouldmakeoneoutofsticksandcloth.

*
1.Howwelldoesthispieceresonatewithyou?

1It
doesn't
resonate

1It
doesn't
resonate

5It
completel
y
resonates

5It
completely
resonates

*
2.Howlikelyareyoutomakeadecisionbasedontheseconcepts?

1Not
likelyatall

5Very
likely

1Not
likelyatall

*
3.Haveyoureadthispiecebefore?

Yes

No

*
4.DoyouhavesignificantexperiencewiththeGermanlanguage?

ReferencedDataSets
Controlgroup
5numbersum:min1.5,Q12.5,med3,Q23,max4
Mean:3.14
Stdev:0.6004
ConfidenceInterval:

0.12800
Rangeforthetruepopulationmean:

3.012to3.268
ExperimentalGroup
5numbersum:min1,Q11.5,med3,Q23,max4
Mean:2.9733
Stdev:0.51258
ConfidenceInterval:

0.10052

5Very
likely

Rangeforthetruepopulationmean:

2.872to3.073
Thetwotailed
Pvalue
equals
1.02175
Byconventionalcriteria,thisdifferenceisconsideredtobenotstatisticallysignificant.

Das könnte Ihnen auch gefallen