Using the Scrambled Sentences Test to Examine Relationships Between Cognitive Bias, Thought Suppression and Borderline Personality Features

Transcription

1 University of Kentucky UKnowledge Theses and Dissertations--Psychology Psychology 2012 Using the Scrambled Sentences Test to Examine Relationships Between Cognitive Bias, Thought Suppression and Borderline Personality Features Paul Jefferson Geiger University of Kentucky, Click here to let us know how access to this document benefits you. Recommended Citation Geiger, Paul Jefferson, Using the Scrambled Sentences Test to Examine Relationships Between Cognitive Bias, Thought Suppression and Borderline Personality Features (2012). Theses and Dissertations--Psychology This Master's Thesis is brought to you for free and open access by the Psychology at UKnowledge. It has been accepted for inclusion in Theses and Dissertations--Psychology by an authorized administrator of UKnowledge. For more information, please contact

2 STUDENT AGREEMENT: I represent that my thesis or dissertation and abstract are my original work. Proper attribution has been given to all outside sources. I understand that I am solely responsible for obtaining any needed copyright permissions. I have obtained and attached hereto needed written permission statements(s) from the owner(s) of each third-party copyrighted matter to be included in my work, allowing electronic distribution (if such use is not permitted by the fair use doctrine). I hereby grant to The University of Kentucky and its agents the non-exclusive license to archive and make accessible my work in whole or in part in all forms of media, now or hereafter known. I agree that the document mentioned above may be made available immediately for worldwide access unless a preapproved embargo applies. I retain all other ownership rights to the copyright of my work. I also retain the right to use in future works (such as articles or books) all or part of my work. I understand that I am free to register the copyright to my work. REVIEW, APPROVAL AND ACCEPTANCE The document mentioned above has been reviewed and accepted by the student s advisor, on behalf of the advisory committee, and by the Director of Graduate Studies (DGS), on behalf of the program; we verify that this is the final, approved version of the student s dissertation including all changes required by the advisory committee. The undersigned agree to abide by the statements above. Paul Jefferson Geiger, Student Dr. Ruth A. Baer, Major Professor Dr. David T. R. Berry, Director of Graduate Studies

3 USINGTHESCRAMBLEDSENTENCESTESTTOEXAMINERELATIONSHIPSBETWEEN COGNITIVEBIAS,THOUGHTSUPPRESSIONANDBORDERLINEPERSONALITYFEATURES THESIS Athesissubmittedinpartialfulfillmentofthe requirementsforthedegreeofmasterofsciencesinthe CollegeofArtsandSciences attheuniversityofkentucky By PaulJ.Geiger Lexington,Kentucky Director:Dr.RuthA.Baer,ProfessorofPsychology Lexington,Kentucky 2012 Copyright PaulJ.Geiger2012

4 ABSTRACTOFTHESIS USINGTHESCRAMBLEDSENTENCESTESTTOEXAMINERELATIONSHIPSBETWEEN COGNTIVEBIAS,THOUGHTSUPPRESSION,ANDBORDERLINEPERSONALITYFEATURES Cognitivebiasandthoughtsuppressionaretwomaladaptivepatternsofthinkingthat havebeenassociatedwithborderlinepersonalitydisorder(bpd).negativecognitive biasesrelatedtobpdincludethoughtsthattheyarebad,powerless,orvulnerableand thattheworldisdangerous.thoughtsuppressionisamaladaptiveemotionregulation strategywhereunwantedthoughtsareintentionallypushedoutofone sconsciousness. However,previousresearchhasconnectedthoughtsuppressionandcognitivebiasesto BPDonlyviaself\reportmeasures.Thepresentstudyexaminedwhetheralaboratory taskmeanttomeasurecognitivebiasandthoughtsuppression(scrambledsentences Test)wouldpredictBPDfeaturesoverandaboveselfreportmeasuresofcognitivebias andthoughtsuppression.asampleof153undergraduatescompletedself\report measuresofbpdfeatures,thoughtsuppression,andnegativecognitivebiases,aswell asthescrambledsentencestest(sst).resultsshowedthatwhilethesstwasagood predictorofcognitivebiases,itdidnotpredictthoughtsuppressionwhenselfreport measureswereincluded.recognizingtheimportanceofnegativecognitivebiasinbpd maybeusefulincontinuedtreatmentdevelopment.furtherresearchintootherwaysof measuringthoughtsuppressionandcognitivebiasesinthelabmaybewarranted. KEYWORDS:BorderlinePersonalityDisorder,ThoughtSuppression,CognitiveBias PaulJ.Geiger September6,2012

5 USINGTHESCRAMBLEDSENTENCESTESTTOEXAMINERELATIONSHIPSBETWEEN COGNITIVEBIAS,THOUGHTSUPPRESSIONANDBORDERLINEPERSONALITYFEATURES By PaulJ.Geiger RuthA.Baer DirectorofThesis DavidT.R.Berry DirectorofGraduateStudies September6,2012

6 TABLEOFCONTENTS ListofTables...iv ListofFigures...v ChapterOne:Introduction...1 BiosocialTheoryofBPD...1 BPDandCognitiveProcessingBiases...2 CognitiveDistortionsinBPD...3 ThoughtSuppressioninBPD...4 AssessmentofCognitiveDistortionsandThoughtSuppression...7 ResearchUsingtheScrambledSentencesTest...8 CurrentStudy...11 ChapterTwo:Methods...14 Participants...14 Self\ReportMeasures...14 PersonalityAssessmentInventory\BorderlineFeaturesScale...14 WhiteBearSuppressionInventory...14 PersonalityBeliefsQuestionnaire\BPDSubscale...15 EmotionalVulnerabilityinChildhood...15 PositiveAffectNegativeAffectSchedule...15 SocializationofEmotionScale...16 ScrambledSentencesTest...16 Procedure...17 ChapterThree:Results...19 PreliminaryAnalyses...19 Hypothesis Hypothesis Hypothesis Hypothesis Hypothesis Hypothesis Hypothesis Hypothesis IncrementalvalidityoftheSSToverbothselfreportmeasuresofcognitive functioning...28 ChapterFour:Discussion...33 References...37 iii

7 Vita...40 iv

8 LISTOFTABLES Table3.1,BivariateCorrelationsandInternalConsistenciesofStudyVariables...20 Table3.2,MeanProportionofPositiveSentencesCreatedontheSSTforLoadandNo\ LoadConditions...22 Table3.3,SummaryofRegressionAnalysesPredictingBPDSymptomSeverityfrom MeasuresofThoughtSuppression.24 Table3.4,RegressionAnalysesPredictingBPDSymptomSeverityfromCognitiveBias..26 Table3.5,SummaryofRegressionAnalysesPredictingBPDSymptomSeverityfromSelf\ ReportMeasuresofThoughtSuppressionandCognitiveBiasandaLabTask AssessingCognitiveBias.. 31 v

9 LISTOFFIGURES Figure3.1,TestofMediationbySST\LoadScoresoftheRelationshipBetweenEV\Child TotalScoresandPAI\BORTotalScores...29 vi

10 ChapterOne:Introduction BorderlinePersonalityDisorderisdefinedas apervasivepatternofinstabilityof interpersonalrelationships,self\image,andaffects,andmarkedimpulsivitybeginningby earlyadulthoodandpresentinavarietyofcontexts (DSM\IV;AmericanPsychiatric Association,1994).IndividualswithBorderlinePersonalityDisorder(BPD)oftenhave unstableinterpersonalrelationships,severeexternalizingbehavior(e.g.suicide,self\ injury,druguse),cognitivedifficultiesincludingdissociation,andheightenedemotional reactivity(apa,1994).bpdisaseverementalillnessthatisestimatedtooccurinless than2%ofthepopulation,yetoccursinasmuchas20%ofinpatientsintheunited States(Kroll,Sines,Martin,1981).Upto40%ofthosewhorepeatedlyseekinpatient treatmentarediagnosedwithbpd. Biosocial(Theory(of(BPD( AwellestablishedmodelofBPDisLinehan sbiosocialtheory,whichpurports thatbpdisaresultofaninvalidatingenvironmentandabiologicallybasedemotional vulnerability,whichinteractovertimetoproduceemotiondysregulation(difficulties regulatingone semotions)andrelatedmaladaptivebehaviors(linehan,1993).an invalidatingenvironmenttypicallyoccursinchildhoodandischaracterizedbycaregivers whofrequentlycriticize,minimize,anderraticallyreinforcecommunicationofinternal experiencesincludingthoughtsandemotions.inaddition,parentalfiguresoftenfailto teachproperproblemsolvingskillsduetooversimplificationofproblemsolving.an invalidatingenvironmentcanalsoincludesexual,physical,andemotionalabuse (WagnerLinehan,1997).Emotionalvulnerabilityisabiologicallybasedpredisposition 1

11 forheightenedemotionalsensitivityandreactivitytostimuli,aswellasaslowreturnto anemotionallystablebaselineafteranemotionalepisode.studieshaveshownthat individualswithbpdhavemoreintenseandvariableemotionalexperienceswhen comparedtonon\bpdcontrols(e.g.koenigsbergetal.,2002).incombination, emotionalvulnerabilityandaninvalidatingenvironmentleadtodeficitsintheskills requiredtoregulateemotionsinareasonableway.thelackofadaptiveskillsfor managingemotionsleadstothebehavioraldysregulationtypicallyseeninbpd, includingself\harm,substanceabuse,bingeeating,andimpulsivespending.these behaviorsareconceptualizedasmaladaptivewaysinwhichpeoplewithbpdattemptto reduceoravoidtheirintensenegativeaffect.arecentlongitudinalstudyfoundthat affectiveinstabilitywasthestrongestandmostconsistentpredictorofbpdsymptoms overtime(tragesser,solhan,schwartz\mette,trull,2007),lendingsupporttothe ideathatemotiondysregulationisthecentralfeatureofbpdandthattheother symptomsfollowfromthiscoredysfunction. BPD(and(Cognitive(Processing(Biases( Recentresearchsuggeststhatcognitiveprocessingbiasescontributein importantwaystothedevelopment,maintenance,orexacerbationofemotional disorders(mathewsandmacleod,2005).cognitivebiasesshowntoberelatedto emotionaldysfunctionincludeselectiveattentiontothreateningstimuli,selective memoryfornegativeinformation,anddistortionsinthecontentofbeliefs,assumptions, andinterpretations(wilson,macleod,campbell,2007).repeatednegativeideation, suchasworryandrumination,andattemptstoinhibitnegativeideationthrough 2

12 thoughtsuppression,hasalsobeenextensivelystudied.mostofthisliteratureconcerns AxisIdisorders.Forexample,panicdisorderisassociatedwithcatastrophic misinterpretationofbodilysensations,whereassocialanxietyinvolvesaself\focused attentionalbiasanddepressionisassociatedwithaself\blamingattributionalstyleand rumination(seemathewsmacleod,2005,forarecentreview).morerecentresearch hasextendedthislineofinquirytobpd.twoformsofcognitiveprocessingbiashave particularlystrongsupportforrelationshipswithbpd:negativedistortionsincognitive contentandthoughtsuppression.( Cognitive(distortions(in(BPD( SeveralmodelsofbiasedbeliefsinBPDhavebeenproposed.BeckFreeman (1990)suggestedthatpeoplewithBPDarelikelytoendorseawiderangeofnegative beliefs,includingbeliefsthattheyarebad,powerless,andvulnerable,andthatthe worldisdangerous.similarly,pretzer(1990)proposedthatasetofthreecorebeliefs underlieborderlinepathology:theworldandotherpeoplearedangerousand malevolent,theselfispowerlessandvulnerable,andtheselfisunacceptable, unlovable,anddeservingofpunishment.empiricalstudiessuggestthatpeoplewithbpd endorsemanynegativebeliefs,includingthosetypicalofmanyotherdisorders(arntz, Dietzel,Dreeson,1999).However,particularbeliefshavebeenshowntobeespecially commoninbpd.butler,brown,beck,grisham(2002)reportedthatpatientswithbpd weremorelikelythanthosewithotherpersonalitydisorderstoendorsebeliefsof dependency,helplessness,distrustofothers,andfearsoflosingemotionalcontrol. FactoranalysisofthePersonalityBeliefsQuestionnaire(PBQ)inasampleofpatients 3

13 diagnosedwithbpdsuggestedthreefactors:distrustofothers,dependencyand neediness,andneedforself\protectioninrelationships(bhar,beck,andbrown,2008). ThePersonalityDisorderBeliefsQuestionnaire(PDBQ;DreesonArntz,1995)measures beliefsassociatedwithseveralpersonalitydisorders,includingbpd(arntzetal.,1999). TheBPDsubscalereflectsthemesofloneliness,unlovability,rejectionbyothers,lackof self\control,andtheselfasbadanddeservingofpunishment.patientswithbpdshowed elevationonallofthesubscales,butscoredhigherthanthosewithotherpdsonthe BPDsubscale(Arntzetal.,1999).Overall,thisliteraturesuggeststhatBPDsymptoms arestronglyassociatedwithnegativebeliefs. Thought(suppression(in(BPD( Thoughtsuppressionistheintentionalattempttopushunpleasantorunwanted cognitionsoutofone sconsciousness.itisconceptualizedasamaladaptiveemotion regulationstrategy.manystudiesshowthatthoughtsuppressionparadoxically increasesthefrequencyoftheunwantedthoughts(seeabramowitz,tolin,street, 2001,forareview).Thiseffect,knownastherebound(effect,(hasbeenexplainedby ironicprocesstheorywhichsuggeststhatthoughtsuppressionrequirestwomental processes:aconscioussearchfordistractersthatisultimatelycognitivelydraining,and anunconscioussearchfortheunwantedthoughtthatisnotcognitivelydraining (Wegner,1992).Thesetwoprocessesoftenworkintandem:highsensitivityand vigilancetounwantedthoughtsactivatethemoreconsciousprocessofthought suppressionanytimeunwantedthoughtsarefound.inessence,inordertosuppressa thought,itisnecessarytomonitorwhethertheunwantedthoughtispresent.ina 4

14 famousstudybywegnerzanakos(1994),participantswereaskedtosuppress thoughtsofawhitebear.however,instreamofconsciousnessreports,subjects showeddifficultysuppressingthatthought,andevenmentionedthewhitebear numeroustimeswhenthiswouldlikelyhaveneverbeenthoughtof,hadtheinstructed thoughtsuppressiontasknotbeenimplemented. Anumberofstudieshavelinkedthoughtsuppressiontoborderlinepersonality disorderandsuggestittobeakeycontributortothemanifestationsofbpd symptomatologyandseverity.chapmanetal.(2005)reportedthattheself\reported tendencytosuppressthoughtswasassociatedwithgreaterfrequencyofunwanted thoughtsandtriggersforself\harm.cheavensetal.(2005),foundthatthought suppression,measuredbythewhitebearsuppressioninventory(wbsi;wegner Zanakos,1994),fullymediatedtherelationshipbetweennegativeaffect intensity/reactivityandacompositescoreofbpdfeatures(includingmeasuresof impulsivity,aggression,andinterpersonalsensitivity).inaddition,thoughtsuppression partiallymediatedtherelationshipbetweenperceivedparentalcriticismandbpd features.findingssuggestthatindividualswhoaremoreemotionallyvulnerableand haveahistoryofunsupportiveparentingmaybemorepronetodevelopingbpd featureswhenthoughtsuppressionisusedasanemotionregulationstrategy.the authorsconcludethat,becausechildhoodexperiencesandbiologicalpredispositionsare notreadilysubjecttochange,therapeuticstrategiesthatencouragereductionsin thoughtsuppressionandteachmoreadaptiveemotionregulationstrategiesmaybe beneficial(cheavensetal.,2005). 5

15 Afollow\upstudybyRosenthaletal.(2005)replicatedthesefindingsusinga structuredinterview(scid\ii)ratherthanself\reportmeasurestoassessbpdfeatures. Chronicthoughtsuppressionfullymediatedtherelationshipbetweennegative affectivityandbpdsymptoms,evenaftercontrollingforvariancerelatedtochildhood sexualabuse.rosenthaletal.notethatthoughtsuppressionprobablypersistsbecause itrelievesnegativeaffectintheshortterm.overthelongterm,however,therepeated useofthoughtsuppressionasanemotionregulationstrategymayleadtoincreased frequencyoftheunwantedthoughtsandmoreintensenegativeemotions,requiring moresevereemotionregulationstrategiessuchasparasuicidalbehavioranddruguse (Rosenthaletal.,2005). SauerandBaer(2009)expandedonthisliteraturebyusingmeasuresthatmore specificallycapturechildhoodemotionalvulnerabilityandinvalidatingenvironment. Resultsshowedthatthoughtsuppressionfullymediatedtherelationshipbetween invalidatingchildhoodenvironmentandsymptomsofbpd.theauthorsnotedthatan invalidatingchildhoodenvironmentmayleadthechildtobelievethathisorher emotionsareharmfulorbad,andthatlosingcontroloftheseemotionscouldleadto severeconsequences.inlinewiththishypothesis,thestudyshowedthatfearof emotionspartiallymediatedtherelationshipbetweenemotionalvulnerabilityand thoughtsuppression,andcompletelymediatedtherelationshipbetweeninvalidating childhoodenvironmentandthoughtsuppression.findingssuggestthatindividualswho growupfearingtheirownemotionsarelikelytousethoughtsuppressionasastrategy foravoidingorescapingthem(sauerandbaer2009). 6

16 Onbalance,thisbodyofliteraturestronglysuggeststhatthoughtsuppressionis acommonlyusedbutmaladaptiveemotionregulationstrategyinpeoplewithbpd.itis associatedwithbpdfeaturesandmaymediaterelationshipsbetweenchildhood precursorstobpdandcurrentseverityofbpdsymptoms. Assessment(of(Cognitive(Distortions(and(Thought(Suppression( Cognitivedistortionsandthoughtsuppressionaremostcommonlyassessed usingself\reportmethods.forbeliefsassociatedwithbpd,themostcommonlyused measuresarethepersonalitybeliefsquestionnaire\bpdsubscale(pbq\bpd;butleret al.,2002)andthepersonalitydisorderbeliefsquestionnaire(arntzetal.,1995),in whichrespondentsratetheextenttowhichtheybelievestatementssuchas Iam needyandweak and Icannottrustotherpeople. Thoughtsuppressionismostoften measuredusingthewbsi(wegnerzanakos,1994),aself\reportinstrumentwith15 itemssuchas TherearethingsIprefernottothinkabout and Ialwaystrytoput thingsoutofmymind. Toprovideanalternativetoself\reportmethods,whichmaybe susceptibletoself\presentationstrategiesordemandcharacteristics,someauthors haveexploredtheuseofthescrambledsentencestest(sst;wenzlaffbates,1998).in thistask,participantsaregivenstringsofsixwordsandaskedtorearrangethemto create5\wordsentences(leavingonewordout).thesix\wordstringsareconstructedso thatthevalenceoftheunscrambledsentencecanbepositiveornegative,dependingon whichwordisomitted.forexample, looksfuturemybrightverydismal canbe my futurelooksverybright or myfuturelooksverydismal. Thetaskistimedand participantsareencouragedtoworkasquicklyaspossible.initsoriginalform,thetask 7

17 isconsideredameasureofnegativecognitivebias;accordingly,peoplewithdepressive disordershavebeenshowntocreatemorenegativesentencesthandonondepressed controls.inavariationonthistaskdesignedtoassessthoughtsuppression,participants areaskedtounscrambletwosetsofsentences:onesetwhilerememberingasix\digit number(cognitiveloadcondition)andanothersetwithoutacognitiveload.because rememberingthesix\digitnumberiseffortful,itreducesthecognitiveresources availablefordistractingattentionfromunwantedthoughts.ithasthereforebeen hypothesizedthatpeoplewhoareattemptingtosuppressnegativethoughtswillfindit moredifficulttodosointhecognitiveloadcondition.asaresult,theywillcreatemore negativesentencesundercognitiveloadthanintheno\loadcondition.thusthe differencebetweenthetwoconditionsintheproportionofnegativesentencescreated isconsideredameasureofthoughtsuppression(wenzlaffbates,1998). Research(using(the(Scrambled(Sentences(Test( TheSSThasbeenusedinseveralstudies,mostlywithdepressedpopulations. WenzlaffBates(1998),whodevelopedthistask,comparedcurrentlydepressed patients,previouslydepressedpatientsinremission,andnever\depressedpersons. Undertheno\loadcondition,thecurrentlydepressedparticipantscreatedmore negativesentencesthanthenever\depressedandthepreviouslydepressedgroups, whichdidnotdifferfromeachother.thisfindingsuggestedalackofcognitivebiasin theremittedgroupandwasconsistentwithmanyself\reportstudiesshowingthat peoplewhohaverecoveredfromdepressiondonotshowevidenceofdepressive cognitivedistortions(e.g.gotlibcane,1987;hollonetal.,1986).however,underthe 8

18 cognitiveloadcondition,theremittedgroupproducedsignificantlymorenegative sentencesthantheyhadproducedwithnoload,andsignificantlymorethanthe nondepressedgroup.incontrast,thenondepressedparticipantsproducedequallylow numbersofnegativesentencesinboththeloadandno\loadconditions.resultswere interpretedtomeanthattheremittedgrouphadnegativecognitivebiasesthatthey weresuppressing,andthatthesuppressioneffortsfailedunderthecognitiveload condition. Rudeetal.(2002)usedtheSSTtostudytherelationshipbetweennegative processingbiasesandfuturedepressiveepisodesinanundergraduatesample. ParticipantscompletedtheSST,theBeckDepressionInventoryandWhiteBear SuppressionInventory.ResultsshowedthathighscoresontheSST(whichindicatea negativeprocessingbias)predicteddepressivesymptomsthatweremeasured4to6 weekslater.thesstloadconditionpredictedfuturedepressiveepisodesinbothmen andwomen,whiletheno\loadconditiononlypredictedfuturedepressiveepisodesin women.interestingly,thedifferencescores(differencebetweenloadandno\load conditions)predictedfuturedepressioninmenonly.additionally,thecombinationof self\reportmeasuresofthoughtsuppression(wbsi)withsstdifferencescoresorload conditionscoreswasanevenstrongerindicatoroffuturedepressioninmen.itwas hypothesizedthatthesegenderdifferenceswereduetodifferentself\presentation strategiesinself\reportmeasures. Afollow\upstudybyRudeetal(2003)utilizedtheSSTwithalarge undergraduatestudentsample.resultsshowedthatthenumberofnegativesentences 9

19 producedintheloadconditionpredicteddepressioninan18\28monthfollow\up assessment(evenaftercontrollingfordepressivesymptomsattheinitialmeeting). Negativesentencesproducedintheno\loadconditionwerenotpredictiveoffuture depression.thedifferencescore(loadvs.no\load)wasamarginallysignificant predictor. Rudeetal(2010)utilizedtheSSTalongsidetheDysfunctionalAttitudesScale (DAS;WeissmanBeck,1978)topredictfutureepisodesofmajordepressivedisorder (MDD)inacommunity\basedsampleofcurrentlynon\depressedwomen.Thisstudy foundthat,whenanalyzedseparately,boththesst(withcognitiveload)andthedas weresignificantpredictorsoffuturedepression.whenexaminedtogether,after controllingfortime1depressionscores,thesst(loadcondition)wasasignificant predictorofmddstatus,whereasthedaswasamarginallysignificantpredictor.sst scoreswithnocognitiveloadwerenotpredictiveofmdd.becausemanyofthe participantshadbeendepressedinthepast,thesefindingswereconsistentwith WenzlaffBates(1998)previouswork,whichsuggestedthatthoughtsuppressionisa strategyusedtocontrolmoodinpeopleatriskforfuturedepression. AlthoughpreviousstudieshaveusedtheSSTprimarilyinthestudyof depression,thesstmayalsobeusefulinthestudyofbpdfeatures.thought suppressionisassociatedwithbothdisorders,andthecognitivedistortionstypicalof depressionappeartobeverysimilartothoseseeninbpd.thatis,bothpopulations endorsebeliefsthattheyareweak,helpless,vulnerable,unlovable,andunworthy.thus, useofataskthatdoesnotrelyonself\reportofgeneraltendenciesmaycontributeto 10

20 knowledgeoftherelationshipsbetweentheseimportantcognitivebiasesandbpd features. Current(Study( Thepurposeofthecurrentstudyistoextendpreviousfindingsoncognitivebias andthoughtsuppressionasmeasuredbythessttothestudyofborderlinepersonality features.thecurrentliteraturestronglysuggeststhatindividualswithbpdendorse manycognitivedistortionsandengageinthoughtsuppressionasaformofemotion regulation,whichismaladaptiveinthelongrunandexacerbatessymptomseverity. However,thesestudieshaveassessedcognitivebiasandthoughtsuppressiononlywith self\reportmethods.nostudyhasexaminedwhetherthetendencytosuppress thoughtsisweakenedundercognitiveloadinthispopulation.theproposedstudyused astudentsamplethathadbeenscreenedandselectedtoincludeawiderangeofbpd features,includingmanyscoringaboveapreviouslyestablishedthresholdforclinically significantbpdsymptoms.theprimarygoalwastotestwhethersstscoresarerelated inexpectedwaystoseverityofbpdfeatures.severalhypothesesweretested. Hypothesis1:Underbothloadandno\loadconditions,participantswithhigher levelsofbpdfeatureswillcreatefewerpositivesentencesonthesst.ifthishypothesis issupported,findingswillbeconsistentwithpreviousresearchsuggestingnegative cognitivecontentbiasesinbpd. Hypothesis2:Differencesbetweentheloadandno\loadconditioninnumberof positivesentencescreated(sstdifferencescore)willbesignificantlynegatively correlatedwithseverityofbpdfeatures.high\bpdparticipantsareexpectedtobemore 11

21 consistentlytryingtosuppressdysfunctionalthoughtsandtohavemoredifficultydoing soundercognitiveloadconditions.incontrast,low\bpdparticipantsarenotexpectedto beengaginginhighratesofthoughtsuppressionduetotheabsenceofcognitivebiases andthereforeshouldproduceroughlyequivalentnumbersofpositivesentencesunder bothconditions. Hypothesis3:ThoughtsuppressionasmeasuredbytheWBSIwillbesignificantly correlatedwiththoughtsuppressionasmeasuredbythesstdifferencescore. Hypothesis4:Self\reporteddysfunctionalbeliefs(PersonalityBeliefs Questionnaire BPDsubscale)willbesignificantlynegativelycorrelatedwiththeSST score(bothloadandno\loadconditions). Hypothesis5:ConsistentwiththefindingsofRudeetal.(2010)inthestudyof depression,bothwbsiandsstdifferencescorewillaccountforsignificantvariancein BPDsymptomseveritywhenbothareenteredintoaregressionmodel.Ifthishypothesis issupported,findingswillsuggestthatself\reportedthoughtsuppressionandthe behavioralmeasureofthoughtsuppressioncapturevarianceinbpdfeaturesthatisnot entirelyoverlapping. Hypothesis6:BothdysfunctionalbeliefsandSSTscorewillaccountforsignificant varianceinbpdseveritywhenbothareenteredintoregressionmodels.aswiththe previoushypothesis,thisfindingwillsuggestthatself\reportedcognitivebiasandthe behavioralmeasureofcognitivebiascapturenon\overlappingvarianceinbpdsymptom severity. 12

22 Asecondarygoaloftheproposedstudyistoreplicatepreviousresearch (Rosenthaletal.,2005;Cheavensetal.,2005;SauerBaer,2009)showingthatthought suppressionmediatestherelationshipbetweenlinehan schildhoodprecursorstobpd (emotionalvulnerabilityandaninvalidatingenvironment)andseverityofbpd symptoms.theproposedstudywillexpandonthispreviousliteraturebyusingthesst differencescoretoassessthoughtsuppression.thefollowinghypotheseswillbetested: Hypothesis7:Measuresofchildhoodemotionalvulnerabilityandaninvalidating environmentwillbepositivelycorrelatedwiththesstdifferencescoreandwithseverity ofbpdfeatures. Hypothesis8:Usingregressionanalysis,betacoefficientsforthechildhood precursormeasuresinpredictingbpdfeatureswillbesignificantlyreducedwhenthe SSTdifferencescoreisincludedinthemodel. Copyright PaulJ.Geiger

23 ChapterTwo:Methods Participants Participantsforthisstudyincluded153undergraduatestudentsrecruitedfrom theintroductiontopsychologypoolattheuniversityofkentucky.inamassscreening procedureearlyinthesemester,studentsfilledoutthepersonalityassessment Inventory BorderlineFeaturesScale(PAI\BOR;Morey,1991)aspartofalarger questionnairepacket.individualswithscoresof38orhigher(t>70)wereconsidered tohavehighbpdfeatures(trull,1995),andwereoversampled.usingundergraduate samplestostudybpdprovidesusefulinformation,asawiderangeofclinically significantbpdfeaturescanbefoundincollegestudents.additionally,itishelpfulto studybpdsymptomsinearlyadulthoodtobetterunderstandhowthedisorder develops(trull,1995).powercalculationsrevealedthat95participantswouldyield acceptablepowerfordetectingasmalltomediumeffectsize(α=.05and1 β=.80). Self?(Report(Measures( Personality(Assessment(Inventory Borderline(Features(Scale(PAI\BOR;Morey, 1991).PAI\BORincludes24itemsratedona4\pointscale(false,slightlytrue,mainly true,andverytrue).itprovidesatotalscoreandsubscalescoresforfourcorefeatures ofbpdsymptomologyincludingselfharm,difficultrelationships,identityproblems,and affectiveinstability.thepai\boriswidelyusedandhasshownexcellentpsychometric properties(morey,1991). White(Bear(Suppression(Inventory((WBSI)(WegnerZanakos,1994).TheWBSI isusedtoassessthegeneraltendencytosuppressthoughts.thismeasureconsistsof 14

24 15items,ratedona5pointLikert\typescale(1=stronglydisagree,5=stronglyagree). Scoresmayrangefrom15to75,withhigherscoresindicatingagreaterproclivityto suppressunwantedthoughts.theauthorsofthisscalereportedgoodinternal consistency(α=.89)andtest\retestreliability(r=.80). Personality(Beliefs(Questionnaire( (BPD(subscale((Arntzetal.,1999).ThePBQ\ BPDsubscaleisusedtomeasurebeliefsmostcommonlyassociatedwithBPD.The measureconsistsof14\items,inwhichrespondentsratehowmuchtheybelieveeach statement.thesestatementsinclude, Iamneedyandweak and Icannottrustother people. Emotional(Vulnerability(in(Childhood((EV?Child).Thismeasurewascreatedby SauerBaer(2010)byadaptingtheAffectIntensityMeasure(Bryant,Yarnold, Grimm,1996;LarsonDiener,1987),whichassessescurrentemotionalintensityin adultrespondents.theev\childadaptedtheitemsandinstructionssothatrespondents ratetheirownemotionalvulnerabilityduringtheirchildhoodyears.researchwiththe EV\Child(SauerBaer,2009;2010)showshighinternalconsistency(α=.92),aclear single\factorstructure,andsignificantcorrelationswithcurrentbpdfeaturesand relatedvariables,aftercontrollingforgeneraldistress.significantagreementbetween studentandparentreportsalsowasnoted. ThePositive(Affect(Negative(Affect(Schedule((PANAS;Watson,Clark,Tellegen, 1988).ThePANASincludes20moodadjectives:tenforpositiveaffect(e.g.,happy)and tenfornegativeaffect(e.g.,(irritated).toprovideastate\levelindicationofnegative affect,participantsratedhowmuchtheywerefeelingeachoftheseitems rightnow 15

25 oratthepresentmoment.onlythenegativeaffectscorewasincludedinthepresent study. ( Socialization(of(Emotion(Scale((SES)(SauerBaer,2010).TheSESisusedto measurechildhoodinvalidationbyaskingrespondentstoreportretrospectivelyontheir parents responsestotheirchildhoodnegativeaffect.themeasurepresentsseveral commonplacechildhoodsituations(e.g.,losingaprizedpossessionandbecomingupset) andasksrespondentstoratetheextenttowhichtheirparentrespondedinvarious ways,someofwhicharevalidatingorsupportive(helpingtolookfortheitem)whereas othersareinvalidating(sayingthatthechildisover\reacting).atotalof33itemsare summedtocreatetwoscores:validationandinvalidation.eachitemisansweredtwice, sothatratingsareobtainedforboththemotherandfather.sauerbaer(2010) reportedstronginternalconsistenciesforbothscalesforreportsofbothparents(alphas rangingfrom.88to.95),acleartwo\factorstructure(validationandinvalidation),and significantcorrelationswithseverityofbpdsymptoms. The(Scrambled(Sentences(Task((SST).((TheSST(Wenzlaff,1998;1993)isa laboratorytaskusedtoassessnegativecognitivebiasesorthesuppressionofunwanted thoughts.thesstasksrespondentstounscramblesentencesfromascrambledphrase (e.g. usuallylikepeoplenotmedo ).Thisphrasecanbeunscrambledinoneoftwo ways:eitherwithapositivevalence( Usuallypeopledolikeme )oranegativevalence ( Peopledonotlikeme ).Respondentsareinstructedtowriteanumber(1\5)above fiveofthesixwordstoindicatetheirsequence.participantsarepresentedwithtwo blocksof25scrambledsentences,andaregiven3.5minutestocompleteeachblock. 16

26 Theyareaskedtocompleteasmanysentencesaspossibleinthetimeprovidedandto refrainfromcorrectingerrors.inaddition,acognitiveloadisrandomlyassignedtoone ofthetwoblocks.inthecognitiveloadcondition,participantsareaskedtoremembera sixdigitnumberwhilecompletingtheblockofsentences. BecausetheSSTwasdevelopedforthestudyofdepression,thesentencesare consistentwithdepression\relatedcognitivedistortions.forthepresentstudy,all60of theoriginalsentenceswereratedforhowmuchtheyresemblethecognitivedistortions typicalofbpdusinga4\pointlikertscale(1=notatall;4=verymuch).threeadvanced graduatestudentswithclinicaltrainingandexperiencetreatingpeoplewithbpd featuresratedeachsentence.thesentencewaskeptforthisstudyifalltherapistsrated thesentenceeithera3or4.thisresultedin17outof60sentencesbeingremoveddue todepression\specificcognitivedistortions.thepithencreated7newsentencesusing thepbq\bpdasaguide. Procedure( Individualswhoobtainedarawscoreover37onthePAI\BORinthemass screeningsessionwerecontactedviaphoneore\mailandaskedtoparticipateinthe study.thestudywasalsoavailabletoallstudentsintheparticipantpoolthroughsona, theonlineregistrationsystemforexperiments.thosewhosignedupthroughsona wereexpectedtofallprimarilywithintheaveragerangeofbpdfeatures.thescreening andinvitationprocedurewasdesignedtoinsurethattheupperendofthedistribution wasadequatelyrepresentedinthesample.studentswhosignedupthroughsonaor acceptedaninvitationtoparticipateweredirectedtoasmallgroupsessionwith 17

27 approximately10\15students.atthebeginningofthesession,theexperimenterbriefly explainedthestudy,questionswereanswered,andtheinformedconsentdocument wascompleted.eachparticipantwasgivenapacketofmaterialsthatincludeda demographicsquestionnaire,twosetsof25scrambledsentencesincounterbalanced order,andabatteryofquestionnaires.thefirstthingcompletedwasthescrambled sentencetest.participantswereinstructedtounscrambleeachsentencetocreatea grammaticalsentencethathasfivewords(leavingonewordout)andtocompleteas manyastheycaninthe3.5minutesallottedbyworkingasquicklyaspossible,creating whatevergrammaticalsentencecametomindfirst,andnotcorrectingerrorsor changingtheirresponses.theywereaskedtowritethenumbers1through5abovefive ofthesixwordstodenotetheorderoftheirunscrambledsentences,asinthefollowing example hasgreenchildtheeyesblue Theorderofthecognitiveloadandno\loadconditionswascounterbalancedbygroup session(eachgroupwasrandomlyassignedtoeitherloadconditionfirstorno\load conditionfirst).inthecognitiveloadcondition,allsubjectswereshownthesamesix\ digitnumberfor30secondsandaskedtocommitittomemory.theywereinstructed torememberthisnumberthroughoutthetask,astheywereaskedtowriteitdown whenthetaskiscompleted. Copyright PaulJ.Geiger

28 ChapterThree:Results Preliminary(Analyses( Alldatawerescreenedforoutliers,missingdata,andsignificantlynon\normal distributions.ofthe153participants,9wereidentifiedasoutliers(+/\3standard deviationsfromthemean)andwereremoved.inaddition,2participantshadmissing demographicdataand1participantdidnotcompleteanentirequestionnaireandwere excluded.thefinalsamplesizewas141.ofthese,30hadpai\borrawscoresover37, suggestingclinicallysignificantbpdfeatures.skewnessandkurtosiswereexamined;no variablesweresignificantlynon\normal.therefore,non\transformedmeanscoresforall variableswereusedfortestingthisstudy shypotheses.forallanalyses,alphawassetat <.05. Zero\ordercorrelationswereexaminedbetweendemographiccharacteristics andallotherstudyvariables.therewasasmallbutstatisticallysignificantcorrelation betweengenderandthedifferencebetweenloadandno\loadscoresonthesst(r(=\.19,p(<.05),showingthatmaleparticipantsshowedslightlygreaterdifferencesthan femaleparticipantsbetweenloadandno\loadconditions.nootherdemographic characteristicsweresignificantlycorrelatedwithanyotherstudyvariables.therefore, remaininganalyseswereconductedwithoutcontrollingfordemographiccharacteristics. Hypothesis(1(( Thefirsthypothesiswasthatunderbothloadandno\loadconditions,participants withhigherlevelsofbpdfeatureswouldcreatefewerpositivesentencesonthesst. 19

29 Table3.1.Bivariate(Correlations(and(internal(consistencies(for(Study(Variables((N(=(141)( Variable PAIBORTot (.89) 2.PAIBORAI.90 (.85) 3.PAIBORIP (.69) 4.PAIBORNR (.73) 5.PAIBORSH (.77) 6.SSTNoLoad E.30 E.25 E.29 E.19 E.22 7.SSTLoad E.44 E.32 E.44 E.37 E SSTDiff E.46 9.PBQEBPD E.37 E (.81) 10.WBSI E.06 E (.90) 11.EVTotal E.14 E (.92) 12.SES(Invalid) E.10 E (.87) 13.PANASENA E.26 E (.72) p<.05,p<.01 Note:Internalconsistenciesforscalespresentedonthediagonal.

30 Correlationalanalysessupportedthishypothesis,asPAI5BORtotalscorewas significantlynegativelycorrelatedwithsstloadscore(r=5.44,p<.001)andsstno5load score(r=5.30,p<.001).additionally,allpai5borsubscalescoresweresignificantly negativelycorrelatedwithbothsstloadscoreandsstno5loadscore.findingscanbe seenintable3.1(rows6and7). Hypothesis2 ThesecondhypothesiswasthattheSSTdifferencescorewouldbesignificantly correlatedwithseverityofbpdfeaturesasmeasuredbypai5bortotalandsubscale scores.thishypothesiswaspartiallysupported.pai5bornegativerelationships subscalescoreshowedasmallbutstatisticallysignificantlycorrelationwithsst differencescore(r=.20,p<.05).allotherpai5borscoreswerenotsignificantly correlatedwiththesstdifferencescore. Differencesbetweenloadandno5loadconditionswerefurtherexaminedby comparingmeanproportionofpositivesentencesinthesubsampleswithandwithout elevatedpai5borscores(rawscore>37).findingsareshownintable3.2.although participantswithelevatedpai5borscorescreatedfewerpositivesentencesintheload conditionthanintheno5loadcondition,thisdifferencewasnotstatisticallysignificant. ParticipantswithPAI5BORscoresbelowtheclinicallysignificantrangeshowedno significantdifferencebetweenloadandno5loadconditions. Hypothesis3 ThethirdhypothesiswasthatthoughtsuppressionasmeasuredbytheWBSI wouldbesignificantlycorrelatedwiththoughtsuppressionasmeasuredbysst 21

31 Table3.2.MeanproportionofpositivesentencecreatedontheSSTforloadandno8 loadconditions Noload Load M SD M SD t PAI1BOR< PAI1BOR>

32 differencescore.thishypothesiswassupported:thewbsiandsstdifferencescores weresignificantlycorrelated(r=.18,p<.05).however,thecorrelationwassmall. Hypothesis4 ThefourthhypothesiswasthatnegativebeliefsasmeasuredbythePBQ5BPD wouldbesignificantlynegativelycorrelatedwithsstscoresunderbothloadandno5load conditions.thishypothesiswassupported.thepbq5bpdwassignificantlynegatively correlatedwithbothload(r=5.43,p<.001)andno5load(r=5.37,p<.001)scores. Hypothesis5 Thefifthhypothesiswasthatbothmethodsofmeasuringthoughtsuppression (WBSIandSSTdifferencescores)wouldaccountforsignificantvarianceinBPDsymptom severity(pai5bortotalandsubscalescores)aftercontrollingforstate5levelnegative affect.thiswastestedusingonehierarchicalregressionanalysiswiththenegative Relationshipssubscalescoreasthedependentvariable(theremainingPAI5BORscores werenotsignificantlycorrelatedwithsstdifferencescores,andthereforenotexamined withregressionanalyses).tocontrolforstate5levelnegativeaffect,panas5nascores wereenteredinstep1ofthemodel.bothmeasuresofthoughtsuppression(wbsiand SSTdifferencescore)wereenteredinStep2.FindingsareshowninTable3.3.Negative affectwasasignificantpredictorofnegativerelationshipsymptoms.theadditionofthe thoughtsuppressionmeasuresinstep2ledtoasignificantincreaseinr 2.Overall,this modelwasstatisticallysignificantr 2 =.23,F(3,137)=13.74,p<.001.Inthefinalmodel, bothpanas5na(β=.19,p<.001)andwbsiscore(β=.36,p<.001)weresignificant 23

33 Table3.3.SummaryofRegressionAnalysesPredictingBPDsymptomseverityfrom measuresofthoughtsuppression DV Predictor R 2 beta PAI5BOR Step1.08 Negative PANAS5NA.29 Relationships Step2.14 PANAS5NA.19 WBSI.36 SSTDiff.12 TotalR 2.23 Note.p<.01 24

34 predictorsofpai5bortotalscores.however,thesstdifferencescorewasnota significantpredictor(β=.12,p=.11). Hypothesis6: Thesixthhypothesiswasthatbothmethodsofmeasuringnegativecognitivebias (PBQ5BPDandSSTscores)wouldaccountforsignificantvarianceinBPDsymptom severity.thishypothesiswastestedusingfivehierarchicalregressionanalyses:onefor thepai5bortotalscoreandoneforeachofthepai5borsubscalescores.preliminary correlationalanalysesshowedthatthesstloadandno5loadscoresweresignificantly correlatedwitheachother(r=.56,p<.01),andthatthesstloadscorewasconsistently morestronglycorrelatedwithpai5borscoresthanwasthesstno5loadscore.a preliminaryregressionanalysisshowedthatwhenloadandno5loadscoreswereboth includedinamodelpredictingpai5bortotalscore,onlytheloadscorewassignificant. Therefore,toavoidpotentialproblemswithhighlycorrelatedpredictors,onlySSTload scoreswereusedfortheseanalyses.tocontrolforstate5levelnegativeaffect,panas5 NAscoreswereenteredinStep1.Measuresofnegativecognitivebiaswereenteredin Step2.FindingsareshowninTable3.4.Forallanalyses,state5levelnegativeaffect (PANAS5NA)wasasignificantpredictorofBPDfeaturesatStep1,andthetwocognitive biasmeasures(pdqbandsst5load)accountedforsignificantadditionalvarianceinstep 2. Inthefinalmodelforthefirstanalysis,allthreevariablesweresignificant independentpredictorsofpai5bortotalscoresincludingpanas5na(β=.16,p<.05), PBQ5BPDscore(β=.53,p<.01),andSSTloadscore(β=5.16,p<.05).Inthesecond 25

35 Table3.4.RegressionAnalysesPredictingBPDsymptomseverityfromCognitiveBias DV Predictor R 2 beta PAI5BOR Step1.19 Totalscore PANAS5NA.44 Step2.30 PANAS5NA.16 PBQ5BPD.53 SSTLoad 5.16 TotalR 2.49 PAI5BOR Step1.13 AffectiveInstability PANAS5NA.37 Step2.23 PANAS5NA.13 PBQ5BPD.50 SSTLoad 5.06 TotalR 2.36 PAI5BOR Step1.21 IdentityProblems PANAS5NA.46 Step2.19 PANAS5NA.23 PBQ5BPD.36 SSTLoad 5.21 TotalR 2.40 PAI5BOR Step1.08 Negative PANAS5NA.29 Relationships Step2.24 PANAS5NA.04 PBQ5BPD.46 SSTLoad 5.16 TotalR 2.33 PAI5BOR Step1.06 Self5Harm PANAS5NA.25 Step2.10 PANAS5NA.09 PBQ5BPD.31 SSTLoad 5.09 TotalR 2.16 Note.p<.01,p<

36 analysispredictingaffectiveinstability,onlythepbq5bpdwasasignificantpredictorin thefinalmodel.inthethirdanalysispredictingidentityproblems,allpredictorswere significantinthefinalmodel,includingpanas5na(β=.23,p<.01),pbq5bpd(β=.36,p <.01),andSSTloadscore(β=5.21,p<.01).InthefourthanalysispredictingNegative Relationships,thePBQ5BPD(β=.46,p<.01)andSSTloadscore(β=5.16,p<.05)were significantinthefinalmodel,whereasthepanas5nawasnotsignificant.inthefifth analysispredictingself5harm,onlythepbq5bpd(β=.31,p<.01)wasasignificant predictorinthefinalmodel. Hypothesis7: TheseventhhypothesiswasthatchildhoodprecursorsofBPD(emotional vulnerabilityandaninvalidatingenvironment)wouldbecorrelatedwithpai5bortotal andsubscalescores,andwiththoughtsuppressionasmeasuredbythesstdifference score.thishypothesiswaspartiallysupported.emotionalvulnerabilityasmeasuredby EV5ChildwassignificantlycorrelatedwithallPAI5BORmeasuresandSSTdifference scores(seetable3.1).aninvalidatingenvironmentasmeasuredbythesesinvalidation totalscorewassignificantlycorrelatedwithallpai5bormeasuresexceptforthe NegativeRelationshipssubscale.However,theSEStotalscorewasnotsignificantly correlatedwithsstdifferencescores. Hypothesis8 Theeighthhypothesiswasthatbetacoefficientsforthechildhoodprecursor measuresinpredictingbpdfeatureswouldbesignificantlyreducedwhenthesst differencescorewasincludedinaregressionmodel.becausethemeasureof 27

37 invalidationinchildhood(ses)wasnotsignificantlycorrelatedwiththesstdifference score,thishypothesiswastestedonlyforemotionalvulnerability(evchild).a mediationalmodelutilizingbootstrapping(preacherhayes,2008)wasusedtotest theindirecteffectofthoughtsuppression(sstdifferencescore)intherelationship betweenemotionalvulnerabilityandbpdfeatures.95%bias5correctedconfidence intervalsweregeneratedusing1,000bootstrapsamples.theindirecteffectofthought suppressionmeasuredbysstdifferencescorewasnotsignificant(p=.33)andthebeta coefficientsfortheevtotalscorewerenotsignificantlyreduced.biascorrected confidenceintervalswere5.01to.05.thisfindingsuggeststhatthoughtsuppressionas measuredbysstdifferencescoredoesnotmediatetherelationshipbetweenemotional vulnerabilityinchildhoodandcurrentbpdfeatures. TheSST5loadscorewasmuchmorestronglycorrelatedwiththePAI5BORthan wasthesst5differencescore;therefore,thisanalysiswasrepeatedwithsst5loadscore asthepotentialmediator.childhoodemotionalvulnerability(ev5child)significantly predictedpai5bortotalscoresandsst5loadscore.whenthesst5loadscorewas includedinthemodel,thestrengthoftherelationshipbetweenev5childandpai5bor scoreswasreduced(betadecreasedfrom.51to.40,p<.001).bootstrappingresults suggestthisindirecteffectisstatisticallysignificant,basedonthe95%confidence interval(.03,.19)(seefigure3.1). IncrementalvalidityoftheSSToverbothselfRreportmeasuresofcognitivefunctioning Posthocregressionanalyseswerecompletedasamorestringenttestofthe utilityofthelaboratorytaskinaccountingforvarianceinbpdfeaturesaftercontrolling 28

38 Figure3.1.TestofmediationbySST5LoadScoresoftherelationshipbetweenEV5Child TotalScoresandPAI5BORTotalScores SST-Load Score -.31 EV-Child Total Score.51 (.40) PAI-BOR Total Score Note.Allvaluesarebetacoefficients.Thevalueinparenthesesrepresentsthebeta valuewhenthemediatorisincludedinthemodel. 29

39 forthetwoself5reportmeasuresofthoughtsuppressionandcognitivedistortions(wbsi andpbq5bpd).theutilityofthelabtaskwastestedusingfivehierarchicalregression analyses:oneforthepai5bortotalscoreandoneforeachofthepai5borsubscale scores.tocontrolforstate5levelnegativeaffect,panas5nascoreswereenteredinstep 1.Thetwoself5reportmeasuresofthoughtsuppression(WBSI)andnegativecognitive bias(pdqb)wereenteredinstep2.thelabtask(sstloadscore)wasenteredinstep3. FindingsareshowninTable3.5.Forallanalyses,state5levelnegativeaffect(PANAS5NA) wasasignificantpredictorofbpdfeaturesatstep1,andtheself5reportmeasures(wbsi andpdqb)accountedforsignificantadditionalvarianceinstep2.thelabtask(sst5 Load)accountedforsignificantadditionalvarianceinStep3forPAI5BORtotalscoreand IdentityProblemssubscalescore.Inthefinalmodels,SST5Loadwasasignificant independentpredictorforthepai5bortotalscoreandtheidentityproblemssubscale. 30

40 Table5.SummaryofRegressionAnalysesPredictingBPDsymptomseverityfromSelfR ReportMeasuresofThoughtSuppressionandCognitiveBiasandaLabTaskAssessing CognitiveBias. DV Predictor R 2 beta PAI5BOR Step1.19 Totalscore PANAS5NA.44 Step2.35 PANAS5NA.15 PBQ5BPD.47 WBSI.30 Step3.02 PANAS5NA.13 PBQ5BPD.42 WBSI.29 SST5Load 5.15 TotalR 2.56 PAI5BOR Step1.13 AffectInstability PANAS5NA.37 Step2.27 PANAS5NA.12 PBQ5BPD.44 WBSI.23 Step3.00 PANAS5NA.11 PBQ5BPD.42 WBSI.23 SST5Load 5.05 TotalR 2.39 PAI5BOR Step1.21 IdentityProblems PANAS5NA.46 Step2.22 PANAS5NA.24 PBQ5BPD.33 WBSI.28 Step3.03 PANAS5NA.21 PBQ5BPD.26 31

41 WBSI.28 SST5Load 5.20 TotalR 2.46 PAI5BOR Step1.08 Negative PANAS5NA.29 Relationships Step2.26 PANAS5NA.04 PBQ5BPD.43 WBSI.23 Step3.02 PANAS5NA.02 PBQ5BPD.38 WBSI.22 SST5Load 5.15 TotalR 2.37 PAI5BOR Step1.06 Self5Harm PANAS5NA.25 Step2.12 PANAS5NA.09 PBQ5BPD.27 WBSI.18 Step3.01 PANAS5NA.08 PBQ5BPD.24 WBSI.18 SST5Load 5.08 TotalR 2.19 Note.p<.01;p<.05. Copyright PaulJ.Geiger

42 ChapterFour:Discussion Negativecognitivebiasesandthoughtsuppressionaretwowell5established constructsrelatedtobpdsymptomseverity.thepresentstudyusedawritten laboratorytask(sst)tostudycognitivebiasandthoughtsuppressionastheyrelateto BPDsymptomseverityinastudentsample.Asexpected,correlationalanalysesshowed significantrelationshipsbetweenbpdsymptomseverityandself5reportandlabtasks designedtomeasurecognitivebiasandthoughtsuppression.additionally,thelabtask measurementsofthoughtsuppressionandcognitivebiasweresignificantlycorrelated withself5reportmeasuresofthoughtsuppressionandcognitivebias.althoughsst5load scorewassomewhatmorestronglycorrelatedwithbpdsymptomseveritycomparedto thesst5noloadscore,agroupcomparisonshowedthatthedifferencebetweenthe LoadandNo5Loadconditionswasnotsignificant,eitherforparticipantswithhighlevels ofbpdfeaturesorforthosebelowtheclinicallysignificantthreshold.however,a nonsignificanttrendwasobservedfortheparticipantswithhighbpdfeaturestocreate fewerpositivesentencesundertheloadconditionthantheno5loadcondition, suggestingthatthecognitiveload(rememberinga65digitnumber)mayhavea somewhatgreatimpactonparticipantswithhighlevelsofbpdsymptoms. Regressionanalysesshowedthataftercontrollingforstate5levelnegativeaffect, thesst5loadscoremeasuringcognitivebiassignificantlypredictedauniqueportionof varianceinbpdsymptomsunaccountedforbypbq5bpd.whilerude(2010)foundthe SSTscoretobeastrongerpredictoroffuturedepressiveepisodesthanself5report measures,thecurrentstudyfoundboththesstandself5reportmeasurestoremain 33

43 significantpredictors.tomorestringentlytestthepredictivepowerofthesst5load score,posthocanalysesenteredbothself5reportmeasuresofinterest(pbq5bpdand WBSI)followedbytheSST5score.Althoughthetwoself5reportmeasuresremained significantpredictorsinthefinalmodel,thesst5loadscorewasasignificantpredictoras wellforthepai5bortotalscoreandtheidentityproblemssubscale.acrossall regressionanalyses,thesst5loadscorewasthestrongestpredictorfortheidentity Problemssubscalescore.Thismaybearesultofthescrambledsentencestaskand cognitivedistortionsrelatedtobpd(theworldandotherpeoplearedangerousand malevolent,theselfispowerlessandvulnerable,andtheselfisunacceptable, unlovable,anddeservingofpunishment)beingmoreheavilyweightedtowards distortionsoftheselfandidentity.furtherexaminationofthesst5loadscorein mediationalanalysesfoundthatthesst5loadscorepartiallymediatedtherelationship betweenachildhoodemotionalvulnerabilityandbpdsymptomseverity,suggesting thatindividualswiththisemotionalvulnerabilityasachildwillbemorelikelytodevelop BPDsymptomswhentheyhavenegativecognitivebiasesaboutthemselves. WhiletheSSTdifferencescoreintendedtomeasurethoughtsuppressiondid significantlycorrelatewiththewbsi,itonlysignificantlycorrelatedwithonepai5bor subscaleanditdidnotaccountforsignificantvarianceoverandabovethewbsiina regressionmodel.furthermore,thesstdifferencescoredidnotmediatethe relationshipbetweenchildhoodemotionalvulnerabilityandbpdsymptomseverity. Theseinsignificantfindingsmaybeduetothefactthatthelabtaskisonlycapturingone aspectofthoughtsuppressionrelatedtobpd.forexample,nosentencesaddress 34

44 impulsivebehaviors(i.e.drugs,alcohol,sex,etc.)orself5harmbehavior twocommon BPDfeatures.SowhiletheSSTdifferencescoremaybeanaccuratemeasureofthought suppressionasitrelatestocognitivedistortionsinbpd,itmaynotbeagoodmeasure forstudyingoverallthoughtsuppressioninbpd. Takenasawhole,thisstudysuggeststhatthescrambledsentenceslabtaskisa usefulsupplementtoself5reportmeasuresinassessingcognitivedistortionsinbpd. Evenafteraccountingforself5reportedcognitivedistortionsasmeasuredbythePBQ5 BPD,theSST5loadscorestillaccountedforasignificantportionofvarianceinBPD features,albeitasmallamount.thesstdidnotappeartoworkaswellforstudying thoughtsuppressioninbpd,inthatthesstdifferencescorewasonlymodestly correlatedwithonefeatureofbpdandwithself5reportedthoughtsuppression.this maybearesultofusingsentencesforthesstthatfocusnarrowlyonbpd5related distortionsalone,asopposedtoincludingthoughtsaboutothersymptomsassociated withbpd.ontheotherhand,therewasadifferencebetweentheloadv.no5load conditionsforparticipantswithhighbpdsymptomatology,albeitnotstatistically significant.insteadofameasureofthoughtsuppression,thisdifferencemayexposean alreadydepletedattentionalsysteminbpd.thereforeitisplausiblethatthesst differencescoreisnotagoodmeasureofthoughtsuppressioninbpd. Althoughfindingsfromthecurrentstudyaresupportiveoftheuseoflaboratory tasksinconjunctionwithself5reportmeasures,otherlabtasksusingdifferentmethods shouldbedeveloped.becausethesststudiedthoughtsuppressioninthespecific contextofbpd5relatedcognitivedistortions,itmayhaveonlycapturedaportionofthe 35

45 constructasitrelatestobpd,asindividualswithbpdmaysuppressthoughtsunrelated tothespecificcognitivedistortionstargetedinthepresentstudy.also,becausethesst wasoriginallydesignedforthestudyofdepression,itmaybemoreeffectivein capturingthoughtsuppressioninadepressedpopulation.anotherlimitationofthe presentstudyistheuseofastudentsamplewithoutanyformaldiagnosesordiagnostic interviews.thecurrentstudywasunabletoreplicatemediationalmodelsusingan invalidatingchildhoodenvironmentastheindependentvariable,possiblybecausethe collegesampleuseddidnotexperienceahigherrangeofchildhoodinvalidation.this study sfindingswouldbestrengthenedifreplicatedwithaclinicalpopulation. Otherfutureresearchdirectionscouldincludeexaminingthoughtsuppressionin BPDasitcomparestootherclinicalpopulations.Inaddition,becauseBPDissucha heterogeneousdisorder,sotoowouldbethetypesofthoughtsthatmaybesuppressed. Therefore,athoughtsuppressiontaskmayshowhighlevelsofthoughtsuppressionin onebpdsubjectbutnottheotherbecauseofthevastconstellationofsymptoms associatedwiththisdisorder.futureresearchmayalsowanttoseekoutwaysfor laboratorytaskstoremainstandardized,yetsomehowincorporatepersonallyrelevant stimuliintothedesigns.overall,thisstudyshowsthestrongpresenceofcognitive biasesinbpdasevidencedinbothself5reportandlaboratorymeasures,whichcanbe targetedforfutureresearchtobetterunderstandthisdisorder Copyright PaulJ.Geiger