Unpleasant and pleasant referential thinking: Relations with self-processing, paranoia, and other schizotypal traits

Journal of Research in Personality xxx (2011) xxxndashxxx

Contents lists available at ScienceDirect

Journal of Research in Personality

journal homepage wwwelsevier com locate j rp

Unpleasant and pleasant referential thinking Relations with self-processingparanoia and other schizotypal traits

David C Cicero John G Kerns uArrDepartment of Psychological Sciences University of Missouri United States

a r t i c l e i n f o

Article historyAvailable online xxxx

KeywordsReferential thinkingParanoiaSelf-esteemSchizotypal personalityBig-five personalityConfirmatory factor analysisSchizotypySelf-consciousness

0092-6566$ - see front matter 2011 Elsevier Inc Adoi101016jjrp201102002

uArr Corresponding author Address University of Mislogical Sciences 214 McAlester Hall Columbia MO 6

E-mail address kernsjmissouriedu (JG Kerns)

Please cite this article in press as Cicero D Cother schizotypal traits Journal of Research in P

a b s t r a c t

Referential thinking is the tendency to view innocuous stimuli as having a specific meaning for the selfand is associated with personality traits and disorders In three studies this research examined the rela-tions among referential thinking self-processing and paranoia In Study 1 follow-up questions on theReferential Thinking Scale (Lenzenweger Bennett amp Lilenfeld 1997) revealed that referential thoughtsare experienced as unpleasant and pleasant In Study 2 unpleasant referential thinking was morestrongly associated with paranoia and maladaptive self-processing and personality CFAs in Studies 1and 2 found that unpleasant and pleasant referential thinking loaded on different factors In Study 3 agroup of participants with elevated schizotypal personality reported more unpleasant and pleasant ref-erential thoughts than a control group

2011 Elsevier Inc All rights reserved

1 Introduction ing (Lenzenweger et al 1997) One self-processing variable that

Referential thinking is a common feature of schizophrenia-spec-trum conditions (Lenzenweger et al 1997) which includes schizo-phrenia Cluster A personality disorders which represent odd oreccentric behavior and include schizotypal schizoid and paranoidpersonality disorders (The American Psychiatric Association 2000)and schizophrenia-like symptoms thought to reflect liability forschizophrenia (ie schizotypy Chapman Chapman Raulin ampEdell 1978 Meehl 1962 Raine 2006) Research has found thatover two-thirds of people with schizophrenia experience delusionsof reference (Frith 1992) and ideas of reference are frequently re-ported in schizotypal personality disorder (Raine 1991) Research-ers interested in the development of schizophrenia have alsosuggested that irregularities in self-concept are one of the mostimportant features of the onset of the disorder (eg Moller amp Hus-by 2000 Raballo Saebye amp Parnas 2009) At the same time refer-ential thinking might be related to other personality traits such asreliance on intuition (King amp Hicks 2009) Despite the potentialimportance of referential thinking in basic personality and schizo-phrenia-spectrum personality disorders relatively little researchhas focused on what psychological mechanisms might contributeto referential thinking (Lenzenweger et al 1997)

Given that referential thoughts reflect viewing information asself-relevant researchers have suggested that referential thinkingmight be related to problems in self-relevant information process-

ll rights reserved

souri Department of Psycho-5211 United States

amp Kerns J G Unpleasant and personality (2011) doi101016j

may be associated with referential thinking is self-esteem Self-es-teem is broadly defined as how people feel about themselves(Kernis 2003) It is possible that referential thoughts might reflecteither low or high self-esteem For instance people might haveunpleasant referential thoughts such as thinking they are beingblamed by others because of a low sense of self-worth In contrastsome referential thoughts might reflect increased self-esteem Forinstance people might have pleasant referential thoughts such aspeople playing songs on the radio just for them because of a highsense of self-worth However to our knowledge no previous re-search has examined the relation between referential thinkingand self-esteem

In addition based on previous research it is unclear whetherreferential thinking can be discriminated from paranoia As men-tioned the central feature of referential thinking is the over-inter-pretation of stimuli as having a special meaning for the self(Lenzenweger et al 1997) Similarly paranoia is the tendency tobe inappropriately suspicious of other peoplersquos motives and behav-iors directed towards oneself (Fenigstein amp Vanable 1992)Although the term lsquolsquoparanoiarsquorsquo has been used to denote any typeof delusional thought (see Lewis (1970) for an historical review)the current research uses the term to reflect the more narrow def-inition related to suspiciousness distrust and persecutory ideation(Combs amp Penn 2004 Fenigstein amp Vanable 1992) Researchershave hypothesized that both referential thinking and paranoiaare personality traits that are associated with cluster A personalitydisorders and are related to self-relevant information processing(Lenzenweger et al 1997) Previous research has found that refer-ential thinking is strongly correlated with paranoia (eg Meyer amp

leasant referential thinking Relations with self-processing paranoia andjrp201102002

2 DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx

Lenzenweger 2009 Stefanis et al 2004) Furthermore research onthe factor structure of schizotypy has found at least three factorsparanoid cognitive-perceptual and negative (eg Compton Goul-ding Bakeman amp McClure-Tone 2009 Stefanis et al 2004)Among these three factors referential thinking has been found tomost frequently load on the paranoid factor However in partdue to limited measurement of referential thinking none of thesestudies actually examined whether referential thinking could forma factor separate from paranoia A recent study that could examinethis found that referential thinking formed a factor separate fromparanoia (Cicero amp Kerns 2010) Hence this suggests that referen-tial thinking and paranoia might be distinct constructs

One issue in examining the relation between referential think-ing with both self-processing variables and paranoia is that assuggested by a number of psychopathologists referential thinkingcould be multidimensional (Startup amp Startup 2005 Wing Coo-per amp Sartorious 1974) In particular referential thoughts mightvary in terms of their experienced emotional valence For exam-ple the most comprehensive measure of referential thinkingThe Referential Thinking Scale was designed to include both pos-itively and negatively valenced referential thoughts (Lenzenwegeret al 1997 Study 1) In contrast paranoia might involve exclu-sively negatively valenced thoughts This is because paranoia in-volves a threat to self Hence paranoid thoughts always involvesome unpleasant emotional content On the other hand referen-tial thoughts do not necessarily involve a threat to the self andcould be either unpleasant or pleasant (Lenzenweger et al1997) For example referential thinking may include unpleasantthoughts such as lsquolsquowhen I see something broken I often wonderif people blame me for itrsquorsquo However it may also include pleasantthoughts such as lsquolsquowhen I hear a favorite song I often wonder if itwas written with me in mindrsquorsquo Thus although paranoia seems toalways involve negatively valenced thoughts referential thinkingcan refer to negatively or positively valenced thoughts The cur-rent research builds on the work of Lenzenweger and colleagues(1997) by empirically testing whether referential thoughts canbe experienced as positively valenced as opposed to exclusivelyunpleasant

The first goal of the current research was to empirically testwhether referential thoughts are experienced as both unpleasantand pleasant The second goal was to examine whether unpleasantand pleasant referential thoughts could be discriminated from eachother and whether they could be discriminated from paranoia Fi-nally the third goal of the current research was to examinewhether unpleasant and pleasant referential thinking had differen-tial relations with facets of self-relevant information processingparanoia other schizotypal personality traits and Big-five person-ality traits

In the current research we hypothesized that unpleasant andpleasant referential thoughts although correlated could be dis-criminated from each other and could be discriminated from para-noia In addition we expected to find that unpleasant referentialthinking would be more strongly correlated with paranoia thanpleasant referential thinking would be We hypothesized thatunpleasant referential thoughts would be associated with unpleas-ant self-relevant information processing including lower explicitand implicit self-esteem higher self-consciousness and lower fac-ets of narcissism In contrast we expected to find that pleasant ref-erential thoughts would be associated with higher implicit andexplicit self-esteem lower self-consciousness and higher facetsof narcissism Finally we expected to find that unpleasant referen-tial thinking would be associated with maladaptive Big-five per-sonality traits while pleasant referential thinking would be moreassociated with adaptive personality traits In general we expectedunpleasant referential thinking and paranoia to display similarrelations with self-processing and big-five personality

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

The current research examined the relations among referentialthinking self-processing paranoia and other schizotypal charac-teristics in three studies In Study 1 follow-up questions wereadded to the Referential Thinking Scale (Lenzenweger et al1997) to determine whether items were sometimes experiencedas pleasant as well as unpleasant In Study 2 ratings of the itemsfrom Study 1 were used to create unpleasant and pleasant sub-scales of the Referential Thinking Scale and these subscales wereused to examine the relations between unpleasant and pleasantreferential thinking with self-processing paranoia other schizo-typal characteristics and Big-five personality In addition we alsotested a series of confirmatory factor analyses that examinedwhether unpleasant referential thinking pleasant referentialthinking and paranoia could be discriminated from each other Fi-nally in Study 3 we tested whether participants with elevatedschizotypal personality had a higher level of both unpleasant andpleasant referential thoughts than a control group

2 Study 1

21 Method

211 ParticipantsParticipants (n = 348) were native English-speaking undergrad-

uate college students at the University of Missouri who completedthe study as partial completion of a course requirement Twenty-six participants were excluded for having Chapman Infrequencyscores of 3 or higher (see below) which resulted in a final sampleof 322 participants Participants ranged from 18 to 37 years oldwith an average age of 1916 (SD = 155) Participants were 47 fe-male 879 White 90 AfricanndashAmerican 06 AsianndashAmericanand 27 other

22 Measures

221 Referential ThinkingThe Referential Thinking Scale (REF Lenzenweger et al 1997)

is a 34-item truendashfalse questionnaire that measures referentialthinking For Study 1 the administration of the REF was modifiedto further assess the experience of referential thoughts None ofthe items in the REF were modified Instead participants wereasked two follow-up questions for each lsquolsquotruersquorsquo response First theywere asked lsquolsquoto what extent was this experience positiversquorsquo on ascale from 0 (not at all positive) to 6 (extremely positive) Secondthey were asked lsquolsquoto what extent was this experience negativersquorsquo ona scale from 0 (not at all negative) to 6 (extremely negative) Thisallowed for the calculation of unpleasant and pleasant referentialthinking scores by summing the 0ndash6 scores for the follow-upunpleasant and pleasant questions Additionally this modificationmade it possible to empirically examine the valence associatedwith specific referential thoughts

222 ParanoiaParanoia was measured with the Paranoia and Suspiciousness

Questionnaire (Rawlings amp Freeman 1996) a 47 item yesndashno ques-tionnaire designed to measure paranoia in a non-psychiatric sam-ple (eg Would you have been more successful if others aroundyou had not put difficulties in your way) The scale contains fivesubscales including interpersonal suspiciousnesshostility nega-tive moodwithdrawal angerimpulsiveness mistrustwarinessand perceived hardshipresentment The PSQ was developed fromseveral existing paranoia scales the PEN Psychoticism scale (Ey-senck amp Eysenck 1975) the Paranoia scale of the MMPI (Hathawayamp McKinley 1989) the Buss Hostility scale (Buss amp Perry 1992)

leasant referential thinking Relations with self-processing paranoia andjrp201102002

1 As can be seen in Table 3 several of the scales in Table 1 violate the assumption ofmultivariate normality of maximum likelihood estimation In addition to using a chi-square difference test that is robust to multivariate normality a BoxndashCox transfor-mation (Box amp Cox 1964) was used to transform the data to a normal distribution andthe same five factor models were fit to the data The pattern of results was nearlyidentical when the transformed data were used instead of the raw data (ie Model 1still fit significantly better than the four other models)

DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx 3

the 16PF Suspiciousness scale (Cattell Eber amp Tatsuoka 1970) andthe STA Paranoid Ideation subscale (Hewitt amp Claridge 1989)

223 InfrequencyParticipants also completed the Chapman Infrequency scale

which measures careless or invalid responding (eg I cannotremember a time when I talked to a person wearing eyeglasses)The Chapman Infrequency scale is composed of questions thatshould rarely truthfully be answer in the affirmative Based on pre-vious research 26 participants endorsing three or more items wereexcluded from the analysis (Chmielewski Fernandes Yee amp Miller1995)

23 Data analysis

To test whether unpleasant referential thinking pleasant refer-ential thinking and paranoia are distinct from each other we com-pared the statistical fit of five confirmatory factor measurementmodels using the sum of the valence scores for unpleasant andpleasant referential thinking First we tested a three-factor modelin which unpleasant referential thinking pleasant referentialthinking and paranoia all loaded on separate factors (Model 1)Second we tested three two-factor models including unpleasantreferential thinkingparanoia pleasant referential thinking (Model2) unpleasant referential thinkingpleasant referential thinkingparanoia (Model 3) and unpleasant referential thinking pleasantreferential thinkingparanoia (Model 4) Finally we tested a one-factor model in which unpleasant referential thinking pleasant ref-erential thinking and paranoia all loaded on a single factor (Model5) We examined whether models with more factors exhibited sig-nificantly better fit than models with fewer factors

All models were fit using Mplus3 software (Muthen amp Muthen2004) Models were fit using maximum likelihood parameter esti-mates and with standard errors and a mean adjusted chi-squarestatistic that is robust to non-normality (the SatorrandashBentler v2Satorra amp Bentler 1994) v2 difference tests of model comparisonswere done using a scaled-difference test statistic (Satorra amp Ben-tler 2001) Models were also compared with the Incremental FitIndex (Tucker amp Lewis 1973 which is also referred to as the Tuck-erndashLewis Index) which compares the fit of models while adjustingfor degrees of freedom IFI values greater than 090 indicate sub-stantial increases in model fit In all models the latent factors wereallowed to correlate freely with each other The latent factors wereallowed to correlate freely as opposed to being constrained toequal zero because we expected to find that unpleasant referentialthinking pleasant referential thinking and paranoia would bemoderately to strongly correlated with each other In most studiesattempting to examine distinct schizotypy factors the factors arespecified to correlate freely (eg Chmielewski amp Watson 2008Kwapil Barrantes-Vidal amp Silvia 2008 Raine et al 1994 Stefaniset al 2004) Thus latent factors may be correlated and still consid-ered to be distinct constructs Four test statistics were used to as-sess whether models provide a good fit to the data (Hu amp Bentler1998) (a) v2df ratio lt25 (b) CFI (comparative fit index) gt95 (c)RMSEA (root mean squared error of approximations) lt08 and (d)SRMR (standardized root mean square residual) lt05

In order to more accurately measure unpleasant referentialthinking pleasant referential thinking and paranoia each of thescales were randomly divided into three facets For example items1 4 7 etc were summed to create the first unpleasant referentialthinking facet items 2 5 8 etc comprised the second facet anditems 3 6 9 etc comprised the third facet Previous research hasused similar techniques to examine the factor structure of similarconstructs including schizotypy (Kwapil et al 2008) and self-con-sciousness (Lischetzke amp Eid 2003) Monte Carlo studies havefound that this method for measuring constructs is more valid than

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

using manifest variables (Alhija amp Wisenbaker 2006) Additionallyin model testing the errors of the manifest variables for the pleas-ant and unpleasant ratings of the referential thoughts were speci-fied to be freely correlated with each other This was done becauseunpleasant and pleasant referential thinking items shared impor-tant method variance (eg scores for unpleasant referential think-ing item 1 and for pleasant referential thinking item 1 were basedon initially endorsing having experienced the same referentialthinking item)

24 Results

241 Unpleasant and pleasant referential thoughtsAs can be seen in Table 1 14 items were experienced as more

unpleasant than pleasant and 20 items were experienced as morepleasant than unpleasant The most unpleasant experience waslsquolsquotraffic lights usually turn red because I am driving in a hurryrsquorsquoOther relatively unpleasant experiences included participants feel-ing like they were being blamed for things feeling like people sayunpleasant things about the participant while in private conversa-tions (eg laughing as the participant walks by two people criti-cizing the participant) and noticing things about the participantthat the participant tried to hide The most pleasant referentialexperiences included strangers waving at the participant radioDJs playing songs specifically for the participant favorite songswritten with the participant in mind and others imitating the par-ticipantrsquos style of dressing

242 Discriminability of unpleasant referential thinking pleasantreferential thinking and paranoia

As can be seen in Table 2 the three-factor model (Model 1) withseparate factors fit the data well and fit the data significantly betterthan all of the other models according to the chi-square differencetest and the Incremental Fit Index1 None of the other models fit thedata even moderately well Thus it appears that unpleasant referen-tial thinking pleasant referential thinking and paranoia may be dis-tinct constructs

In model 1 unpleasant and pleasant referential thinking werepositively correlated (r = 71) unpleasant referential thinking andparanoia were positively correlated (r = 57) and pleasant referen-tial thinking and paranoia were positively correlated (r = 36) InModel 2 pleasant referential thinking was positively correlatedwith the unpleasant referential thinkingparanoia factor (r = 53)In Model 3 the unpleasantpleasant referential thinking factorwas positively correlated with the paranoia factor (r = 55) In Mod-el 4 the pleasant referential thinkingparanoia factor was positivelycorrelated with the unpleasant referential thinking factor (r = 71)

25 Discussion

The first goal of Study 1 was to examine whether referentialthoughts were experienced as pleasant in addition to unpleasantIndeed Study 1 found that there was a great deal of variability inthe valence associated with the referential thoughts This is consis-tent with the original conceptualization of the REF which was de-signed to include both positively valenced and negatively valenceditems (Lenzenweger et al 1997) Study 1 also provided some evi-dence that unpleasant referential thinking pleasant referential

leasant referential thinking Relations with self-processing paranoia andjrp201102002

Table 1Mean difference scores of pleasant minus unpleasant ratings for Referential ThinkingScale items

Referential Thinking Scale Item Mean Difference Score

23 34613 34210 31825 3168 27111 26532 23920 23822 23626 23412 20515 2024 19919 17924 15334 159 09821 0732 02317 0025 03418 05427 0631 0933 1077 12514 18328 19233 19529 2056 21631 28430 29716 336

See Lenzenweger et al (1997) for the corresponding items to match the itemsnumbers

Table 2Fit statistics for confirmatory factor analysis measurement models of referentialthinking and paranoia in Study 1

Model v2 df CFI RMSEA SRMR v2 diff (vsModel 1)

IFI (vsModel 1)

Model1

2647 21 099 003 003 ndash

Model2

12114 23 082 012 014 1494fraslfrasl 094

Model3

17670 23 073 015 011 4832 096

Model4

20614 23 068 017 013 18322 097

Model5

33470 24 045 021 015 8104fraslfrasl 098

Model 1 3-factor model (unpleasant referential thinking pleasant referentialthinking paranoia) Model 2 2-factor model (unpleasant referential thinkingpar-anoia pleasant referential thinking)Model 3 2-factor model (unpleasant referential thinkingpleasant referentialthinking paranoia)Model 4 2-factor model (unpleasant referential thinking pleasant referentialthinkingparanoia)Model 5 1-factor model (unpleasant referential thinkingpleasant referentialthinkingparanoia) v2 diff = SatorrandashBentler chi-square difference test Significantdifference represents worse model fit IFI = Incremental Fit Index p lt 01

4 DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx

thinking and paranoia are all correlated but distinct from one an-other The best fitting CFA model included separate unpleasant ref-erential thinking pleasant referential thinking and paranoiafactors In addition to the results of confirmatory factor analysisif unpleasant referential thinking and pleasant referential thinkingare distinct constructs they should display differential associationswith other theoretically meaningful variables

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

3 Study 2

In Study 1 we found that referential thoughts could be experi-enced as both unpleasant and pleasant and that these thoughtscould be discriminated from each other The goal of Study 2 wasto use the information about the valence of referential thoughtsfrom Study 1 to examine the relations between unpleasant andpleasant referential thoughts with paranoia self-processingschizotypal personality and Big-five personality in a separate sam-ple Based on the valence ratings from Study 1 pleasant andunpleasant subscales of the Referential Thinking Scale were cre-ated and the correlations between scores on these subscales andother variables were examined

31 Method

311 ParticipantsParticipants (n = 347) were native English-speaking undergrad-

uate college students at the University of Missouri who completedthe study as partial completion of a course requirement Followingprevious research participants (n = 35) were excluded due toChapman infrequency scores of 3 or greater (Chapman amp Chapman1983) In addition 17 participants were excluded due to failing tocomplete all the questionnaires resulting in 295 useable partici-pants Participants ranged from 18ndash42 years old with an averageage of 1887 (SD = 185) Participants were 59 female 901White 58 AfricanndashAmerican 20 AsianndashAmerican and 17other One participant declined to specify ethnicity

312 Measures referential thinkingParticipants completed the Referential Thinking Scale (Len-

zenweger et al 1997) and participants were not asked about thevalence of their experiences in this study

313 ParanoiaFour measures of paranoia were administered in Study 2 One

measure was the Paranoia and Suspiciousness Questionnaire(Rawlings amp Freeman 1996) as in Study 1 A second paranoia mea-sure was the 8-item Suspiciousness subscale from the SchizotypalPersonality Questionnaire (SPQ-S Raine 1991 eg Do you some-times get concerned that friends or coworkers are not really loyalor trustworthy) Overall the full Schizotypal Personality Ques-tionnaire (SPQ Raine 1991) is a 74-item yesndashno questionnairedesigned to measure DSM-III-R schizotypal personality disorderThe SPQ has been the most frequently used scale in studies exam-ining the factor structure of schizotypy traits (eg Stefanis et al2004)

A third paranoia measure was the Suspiciousness subscale ofthe Dimensional Assessment of Personality Pathology-Basic Ques-tionnaire (DAPP-BQ Livesley amp Jackson 2002) which includes 14items (eg When people do something nice for me I wonder whattheir real motives are) Participants answer on a scale from 1 = veryunlike me to 5 = very like me The DAPP-BQ Suspiciousness subscalehas been shown to be highly correlated with a count of DSM-IVparanoid personality disorder symptoms (r = 67 Bagge amp Trull2003)

The fourth paranoia measure was the Paranoid Personality Dis-order Features Questionnaire (PPDFQ Useda amp Trull 2002) a 23-item questionnaire (eg I am careful about the way I act aroundother people because they may take advantage of me) Participantsrate statements on a scale from 0 = strongly disagree to 4 = stronglyagree Two items are reverse coded with higher scores reflectinghigher paranoid personality disorder characteristics The scale con-tains six subscales measuring suspiciousnessmistrust antago-nism autonomy hypersensitivity hypervigilence and rigidity

leasant referential thinking Relations with self-processing paranoia andjrp201102002

DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx 5

Useda and Trull (2002) found that the PPDFQ is highly correlated(r = 78) with the DAPP-BQ Suspiciousness Subscale Since the par-anoia scales were highly correlated with each other (rs rangedfrom 061 to 076) a composite paranoia score was calculated bytaking the mean of the standardized z-score for all four measures

314 Explicit self-esteemExplicit self-esteem was measured with the Rosenberg Self-

Esteem Scale (RSES Rosenberg 1965) a 10-item Likert scale rang-ing from 1 = strongly agree to 4 = strongly disagree (eg I feel that Iam a person of worth at least on an equal plane with others)Several items are reverse scored The RSES has been shown to havehigh internal consistency and testndashretest reliability (Rosenberg1965) is highly associated with other measures of explicit self-esteem (eg Demo 1985 Tafarodi amp Swann 1995) and may bethe most commonly used measure of trait self-esteem (LearyTambor Terdal amp Downs 1995)

315 Implicit self-esteemImplicit self-esteem was measured with the self-esteem Impli-

cit Association Test (IAT Greenwald amp Farnham 2000) The self-es-teem IAT has been found to have the highest testndashretest reliabilityof all existing measures of implicit self-esteem (Bosson Swann ampPennebaker 2000) Moreover implicit self-esteem as measuredwith the IAT has been found to predict different outcomes thanself-esteem assessed with explicit measures (Bosson Brown Zeig-ler-Hill amp Swann 2003 de Jong 2002 Schimmack amp Diener2003)

Self-esteem was measured both explicitly and implicitly be-cause previous research suggests that explicit and implicit self-es-teem may be differentially related to facets of schizotypyparticularly paranoia For example some research suggests thatparanoia is associated with a discrepancy between high explicitself-esteem and low implicit self-esteem (eg Bentall Kaney ampDewey 1991 Bentall Kinderman amp Kaney 1994) while otherresearch suggests that paranoia is associated with both decreasedexplicit and implicit self-esteem (eg Freeman 2007) No previousresearch has examined whether referential thinking is associatedwith implicit self-esteem The current research did not measureother variables on an implicit level because previous research hasnot suggested that they are associated with paranoia referentialthinking or other facets of schizotypy on an implicit level

316 Self-consciousnessSelf-consciousness was measured using the 23-item Self-

Consciousness Scale (SCS Fenigstein Scheier amp Buss 1975) Thescale was administered as a truendashfalse questionnaire It contains

Table 3Correlations among unpleasant and pleasant referential thinking and other variables in St

1 2 3 4

1 Unpleasant referential thinking 752 Pleasant referential thinking 61 753 Paranoia 62 49 -4 Rosenberg self-esteem 36 08 47 885 Implicit self-esteem 01 12 02 046 Self-consciousness 31 18 29 27 NPI-leadershipauthority 10 16 01 40

8 NPI-entitlementexhibition 19 46 24 17

9 Magical ideation scale 43 44 46 110 Perceptual aberration scale 45 29 48 3Mean 230 305 0 321Standard deviation 242 293 1 475Range 0ndash13 0ndash14 166ndash289 13ndashSkewness 159 122 062 0Kurtosis 278 126 008 017

p lt 05 The numbers on the diagonal are Cronbachrsquos Alpha

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

subscales for public self-consciousness (eg Irsquom concerned aboutwhat other people think of me) and private self-consciousness(eg Irsquom always trying to figure myself out) This self-consciousness scale has been used in previous research examiningassociations between self-consciousness and paranoia (eg Combsamp Penn 2004 Lenzenweger et al 1997)

317 NarcissismNarcissism represents relatively normal but disordered self-

processing characterized by a pattern of grandiosity and entitle-ment and is strongly associated with self-esteem (RodebaughWoods amp Heimberg 2007 Sedikides Rudich Gregg Kumashiroamp Rusbult 2004) The Narcissistic Personality Inventory (NPIRaskin amp Terry 1988) was used to measure narcissism The NPIis a 40-item questionnaire (eg If I ruled the world it would be amuch better place) that is commonly used to measure narcissismIt has been found to be correlated with staff and clinician ratings ofnarcissism in clients ratings of narcissistic behavior in an experi-mental discussion task and with dominance and sociability scores(two personality characteristics thought to be strongly related tonarcissism Raskin amp Terry 1988) Previous research suggests thatthe NPI may be multidimensional and composed of at least twofactors (Corry Merritt Mrug amp Pamp 2008 Rodebaugh et al2007) A recent item-level confirmatory factor analysis of the NPIcompared the fit of competing 2- 3- 4- and 7-factor models andconcluded that a 2-factor model was the most parsimonious andprovided the best fit to the data (Corry et al 2008) Additionallysubscale scores based on the two-factor model have high internalconsistency and are recommended for use by Corry et al Thesetwo factors have been termed leadershipauthority and exhibition-ismentitlement Previous research has found that leadershipauthority may be a more covert facet of narcissism and is stronglyrelated to extraversion dominance social boldness and high self-esteem Conversely exhibitionismentitlement may representmore overt narcissism and may be more strongly related toachievement tension anxiety and suspiciousness (Corry et al2008) If unpleasant referential thinking is associated with lowself-esteem and maladaptive personality then we would expectto find that it would not be associated with leadershipauthoritybut would be associated with exhibitionismentitlement In con-trast if pleasant referential thinking is associated with high self-esteem then we would expect to find that it would be associatedwith both leadershipauthority and exhibitionismentitlement Ascan be seen in Table 3 these two subscales of the NPI were highlycorrelated with each other and had high internal reliability

udy 2

5 6 7 8 9 10

937 04 60

01 13 78

02 17 47 767 03 04 15 30 833 01 06 01 19 67 834 012 496 595 563 545 373

007 177 233 321 463 40740 010ndash031 0ndash8 0ndash9 0ndash13 0ndash25 0ndash2736 001 053 069 032 134 249

004 059 041 065 225 882

leasant referential thinking Relations with self-processing paranoia andjrp201102002

6 DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx

318 Other schizotypal personality characteristicsThere were two scales used to measure other schizotypal per-

sonality characteristics One scale was the Magical Ideation Scale(Eckbald amp Chapman 1983) a 30-item truendashfalse questionnaire de-signed to measure lsquolsquobeliefs in forms of causation that by conven-tional standards are invalidrsquorsquo (Eckbald amp Chapman 1983 p215)For example lsquolsquoI have worried that people on other planets maybe influencing what happens on Earthrsquorsquo The Magical Ideation scalehas considerable support for its reliability and validity (for a re-view see Edell 1995) A second schizotypal personality scale wasthe Perceptual Aberration Scale (Chapman Chapman amp Raulin1978) a 35-item true false scale that includes 28 items designedto measure schizophrenic-like distortions in perception of onersquosown body and seven items for other perceptual distortions (egmy hearing is sometimes so sensitive that ordinary sounds becomeuncomfortable) The Perceptual Aberration Scale also has consider-able support for its reliability and validity (for a review see Edell1995) The authors of these scales also referred to them as mea-sures of lsquolsquopsychosis-pronenessrsquorsquo and both measures have beenfound to predict future onset of psychosis (eg Chapman Chap-man Kwapil Eckblad amp Zinser 1994)

319 Big-five personality characteristicsIf unpleasant referential thinking is associated with decreased

self-esteem and with increased paranoia then we would expectto find that unpleasant referential thinking would be associatedwith maladaptive personality characteristics Conversely if pleas-ant referential thinking is associated with increased self-esteemand less strongly associated with paranoia then we would expectthat pleasant referential thinking would be associated with adap-tive personality characteristics Big-five personality characteristicswere measured with the 100-item International Personality ItemPool (IPIP Goldberg 1999) with five 20-item subscales for eachof the five factors of personality neuroticism (eg I get stressedout easily) extroversion (eg I am the life of the party) opennessto experience (eg I have a vivid imagination) agreeableness (egI sympathize with other peoplersquos feelings) and conscientiousness(eg I am always prepared) Participants rate their agreement withitems on a 5 item Likert scale from 1 = very inaccurate to 5 veryaccurate

32 Procedure

Participants first completed the self-esteem Implicit AssociationTest Then they completed the Referential Thinking Scale the Pub-lic Self-Consciousness Subscale of the Self-Consciousness Scaleand the Paranoia and Suspiciousness Questionnaire randomlymixed together Then participants completed the Paranoid Person-ality Disorders Features Questionnaire Survey of Attitudes andExperiences (Composed of the Magical Ideation Scale PerceptualAberration Scale Revision Social Anhedonia Scale and InfrequencyScale) DAPP-BQ Suspiciousness subscale and the Rosenberg Self-Esteem Scale Participants completed the study in one occasionin an isolated room with the entire study taking approximately90 min

33 Results

331 Discriminability of unpleasant and pleasant referential thoughtsTo further test whether unpleasant referential thinking could be

discriminated from pleasant referential thinking we conducteditem-level confirmatory factor analyses on the REF items Itemsthat were rated as more unpleasant than pleasant in Study 1 werespecified to load on the first factor (ie the unpleasant factor) anditems that were rated as more pleasant than unpleasant were spec-ified to load on the second factor (ie the pleasant factor) Then

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

the fit of this model was compared to the fit of a single factor mod-el in which all of the items loaded on a single lsquolsquoreferential thinkingrsquorsquofactor Using Mplusrsquo categorical variable option and weighted leastsquare mean and variance (WLSMV) estimation the fit of thesetwo models were compared We used WLSMV as opposed to MLas in Study 1 because ML estimation cannot be used with categor-ical variables The latent variables were allowed to correlate freelybecause we expected unpleasant and pleasant referential thinkingto be correlated with each other Both the two-factor model (v2df = 202 CFI = 086 RMSEA = 006 SRMR = 012) and the one-fac-tor model fit the data reasonably well (v2df = 205 CFI = 086RMSEA = 006 SRMR = 016) The unpleasant and pleasant referen-tial thinking factors were strongly correlated with each other(r = 93) A standard chi-square difference test cannot be used withWLSMV estimation because the difference between chi-squarevalues for two models is not distributed as chi-square using thisestimation method Thus the difftest command in Mplus whichuses derivatives to correct for this distribution (Asparouhov2006) was used to compare the fit of the more restrictive model(ie the one-factor model) to the fit of the less restrictive model(ie the two-factor model) The resulting value can be interpretedlike a standard chi-square difference test The resulting v2 was sig-nificant (v2 diff (1) = 1148 p lt 001) which suggests that restrict-ing all the items to load on a single factor as opposed to twofactors worsened the fit of the model In turn this suggests thatreferential thinking may be composed of both an unpleasant-valence factor and a pleasant-valence factor that are distinct buthighly correlated

One explanation for the finding that a factor model with sepa-rate pleasant and unpleasant factors fit the data better than a sin-gle factor model could be that instead of tapping different latentconstructs our factors represent groups of items with different lev-els of item difficulty If this were the case we would expect there tobe a significant difference in the percentage of the populationendorsing the unpleasant items than the percentage of participantsendorsing the pleasant items There was not a significant differ-ence in the percentage of participants endorsing the unpleasantitems compared to percentage of participants endorsing the pleas-ant items (M = 262 SD = 016 vs M = 252 SD = 016 t (32) = 20p = 85)

332 ParanoiaAs can be seen in Table 3 unpleasant referential thinking was

more strongly correlated with paranoia than was pleasant referen-tial thinking To test whether the difference between the correla-tions was significant we computed a Z-score for the differencebetween correlated correlation coefficients as suggested by MengRosenthal and Rubin (1992) Unpleasant referential thinking wasmore strongly correlated with paranoia than was pleasant referen-tial thinking (Z = 319 p = 001) To further test whether unpleasantreferential thinking was more strongly correlated with paranoiathan was pleasant referential thinking unpleasant referentialthinking and pleasant referential thinking were simultaneously en-tered into a multiple regression equation predicting paranoiaThese results can be interpreted as the relation between unpleas-ant referential thinking and paranoia after removing shared vari-ance with pleasant referential thinking and the relation betweenpleasant referential thinking and paranoia after removing varianceshared with unpleasant referential thinking In this regressionanalysis unpleasant referential thinking seemed even morestrongly predictive of paranoia than was pleasant referential think-ing (b = 55 vs 17)

333 Explicit self-esteemAs can also be seen in Table 3 unpleasant referential thinking

was associated with low explicit self-esteem while pleasant refer-

leasant referential thinking Relations with self-processing paranoia andjrp201102002

Table 4Zero-order correlations among referential thinking scales and Big-five personalitycharacteristics in Study 2

1 2 3 4 5 6 7

1 Unpleasant referentialthinking

75

2 Pleasant referentialthinking

61 75

DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx 7

ential thinking was not significantly associated with explicit self-esteem (Z = 841 p lt 001) When entered separately into a multi-ple regression unpleasant referential thinking was negativelyassociated with explicit self-esteem (b = 44 p lt 001) and pleas-ant referential thinking was associated with increased explicit self-esteem (b = 19 p lt 01)

334 Implicit self-esteemUnpleasant referential thinking was not significantly associated

with implicit self-esteem but pleasant referential thinking wasassociated with increased implicit self-esteem (Z = 251 p = 01)Moreover when removing shared variance with pleasant referen-tial thinking unpleasant referential thinking was still unassociatedwith implicit self-esteem (b = 09 p = 25) and pleasant referen-tial thinking was still positively associated with implicit self-es-teem (b = 16 p = 04)2

335 Self-consciousnessAs shown in Table 3 both unpleasant and pleasant referential

thinking were associated with increased public self-consciousnessbut unpleasant referential thinking was more strongly correlatedwith public self-consciousness than was pleasant referential think-ing (Z = 261 p = 008) When entered simultaneously into a multi-ple regression unpleasant referential thinking was associated withincreased public self-consciousness (b = 32 p lt 001) while pleas-ant referential thinking was not significantly associated with pub-lic self-consciousness (b = 02 p = 82)

336 Facets of narcissismAs can be seen in Table 3 unpleasant referential thinking was

significantly correlated with the exhibitionentitlement facet ofnarcissism but not the leadershipauthority facet Pleasant refer-ential thinking was more strongly associated with both the exhibi-tionismentitlement facet (Z = 551 p lt 001) and the leadershipauthority facet (Z = 499 p lt 001) than was unpleasant referentialthinking When entered simultaneously into a multiple regressionequation unpleasant referential thinking was negatively associ-ated with exhibitionismentitlement (b = 17 p = 01) whilepleasant referential thinking was positively associated with exhibi-tionismentitlement (b = 56 p lt 001) Similarly unpleasant refer-ential thinking was negatively associated with leadershipauthority (b = 34 p lt 001) and pleasant referential thinkingwas positively associated with leadershipauthority (b = 37p lt 001)

337 Schizotypal personalityUnpleasant and pleasant referential thinking were both posi-

tively correlated with magical ideation and perceptual aberrationHowever unpleasant referential thinking was more strongly corre-lated with perceptual aberration than was pleasant referentialthinking (Z = 364 p lt 001) When entered simultaneously into amultiple regression both unpleasant (b = 28 p lt 001) and pleas-ant referential thinking were still associated with magical ideation(b = 24 p lt 001) Unpleasant referential thinking was still associ-ated with perceptual aberration (b = 38 p lt 001) when removingvariance shared with pleasant referential thinking but pleasantreferential thinking was not (b = 05 p = 45)

2 To test whether a discrepancy between implicit and explicit self-esteem wasassociated with unpleasant referential thinking or pleasant referential thinking wetested a series of hierarchical linear regression models Mean centered explicit andimplicit self-esteem scores were entered in step one and the product of implicit andexplicit self-esteem scores was entered in step 2 There was not a significantinteraction between implicit and explicit self-esteem scores in predicting unpleasantreferential thinking (t (279) = 49 p = 63) or pleasant referential thinking (t(279) = 31 p = 76)

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

338 Big-Five PersonalityAs can be seen in Table 4 unpleasant referential thinking was

associated with decreased extraversion agreeableness conscien-tiousness and openness to experience but increased neuroticismIn contrast pleasant referential thinking was only associated withincreased neuroticism although not as strongly as was unpleasantreferential thinking These correlations were significantly differentfor neuroticism (Z = 293 p = 003) extraversion (Z = 515 p lt 001)agreeableness (Z = 211 p = 04) conscientiousness (Z = 211p = 04) and openness to experience (Z = 404 p lt 001) Whenremoving variance shared with unpleasant referential thinkingpleasant referential thinking was associated with increased extra-version (b = 29 p lt 001) and openness to experience (b = 26p lt 001) After removing variance with pleasant referentialthinking unpleasant referential thinking was still associatedwith decreased extraversion (b = 42 p lt 001) agreeableness(b = 22 p lt 001) conscientiousness (b = 24 p lt 001) andopenness to experience (b = 32 p lt 001) but increased neuroti-cism (b = 39 p lt 001)

34 Study 2 discussion

Study 2 provided further evidence that unpleasant and pleasantreferential thoughts could be discriminated from each other in aseparate sample from Study 1 A confirmatory factor analysis withtwo factors in which items rated as being more pleasant thanunpleasant loaded on one factor and items rated as more unpleas-ant than pleasant loaded on a second factor fit the data better thana CFA in which all the items loaded on a single factor This suggeststhat unpleasant and pleasant items may be correlated but distinct

Additionally the results of Study 2 largely conformed to ourhypotheses about the relations between unpleasant and pleasantreferential thinking with paranoia self-processing other schizo-typal personality characteristics and Big-five personality traitsAs hypothesized unpleasant referential thinking was morestrongly correlated with paranoia than was pleasant referentialthinking which was found when shared variance was and wasnot removed Moreover unpleasant referential thinking was asso-ciated with lower explicit self-esteem and higher public self-con-sciousness than was pleasant referential thinking In contrastpleasant referential thinking was associated with increased impli-cit self-esteem whereas unpleasant referential thinking was notUnpleasant referential thinking was associated with personalitytraits that are generally considered to be maladaptive while therewas some evidence that pleasant referential thinking was associ-ated with personality traits that are generally considered to beadaptive Overall these results suggest that unpleasant referentialthinking is associated with more unpleasant biases in self-relevant

3 Neuroticism 42 28 934 Extraversion 26 01 35 915 Agreeableness 20 09 31 44 866 Openness to

experience16 05 20 41 42 87

7 Conscientiousness 18 07 30 26 37 22 88Mean 230 305 322 337 379 354 316Standard deviation 242 293 066 071 050 053 037

p lt 05 The numbers on the diagonal are Cronbachrsquos Alpha

leasant referential thinking Relations with self-processing paranoia andjrp201102002

8 DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx

information processing and maladaptive personality traits whilepleasant referential thinking is more associated with pleasant orthe absence of biases in self-relevant information processing andmore adaptive personality traits In the current research onlyself-esteem was measured on an implicit level Future researchcould examine the relations among unpleasant referential think-ing pleasant referential thinking other schizotypal traits andBig-five personality measured implicitly

Finally Study 2 found that both unpleasant and pleasant refer-ential thinking were associated with measures of schizotypal per-sonality but that unpleasant referential thoughts may be morestrongly associated with some other schizotypal personality char-acteristics than are pleasant referential thoughts The finding thatboth unpleasant and pleasant referential thoughts were correlatedwith schizotypal traits suggests that people with schizotypy mayhave elevated referential thoughts regardless of the valence ofthese thoughts However one limitation of Study 2 is that it in-volved unselected college student samples Thus it is not clearhow unpleasant and pleasant referential thoughts are experiencedamong people with more clinically meaningful schizotypal symp-toms In Study 3 we administered the Referential Thinking Scaleto a sample of participants with extreme levels of schizotypy andcompared them with a control group

4 Study 3

The main goal of Study 3 was to examine whether a group ofpsychometrically identified participants who have elevated schizo-typy and are at increased risk for psychosis (Chapman et al 1994)would have more unpleasant and pleasant referential thoughtsthan a control group In addition to examining risk for psychosisdimensionally by correlations with the Magical Ideation andPerceptual Aberration Scales as in Study 2 schizotypy researchershave often used a lsquolsquohigh riskrsquorsquo approach to examining the correlatesof schizotypy (eg Chapman et al 1994 Gooding Tallent amp Matts2005 Lenzenweger 1994 Miller 1995) This approach consists ofidentifying participants with extremely high scores on the MagicalIdeation and Perceptual Aberration Scales and comparing theseparticipants to a control group of participants with relatively lowscores on both of these scales In Study 3 we used this high risk ap-proach to complement and extend the results of Study 2

If unpleasant and pleasant referential thinking are both associ-ated with other schizotypal personality characteristics then wewould expect to find that a group of participants with extreme lev-els of schizotypy would have increased unpleasant and pleasantreferential thoughts However if only unpleasant or only pleasantreferential thoughts are associated with other schizotypal person-ality characteristics then we would expect to find that onlyunpleasant or only pleasant referential thoughts would be elevatedin the schizotypal sample

41 Method

411 ParticipantsParticipants were 55 (24 Schizotypal and 31 Control) under-

graduate college students at the University of Missouri who wererecruited from a large pool of participants (n = 1901) who hadcompleted a screening battery of questionnaires in partial fulfill-ment of a course requirement The questionnaires included abbre-viated versions of the Magical Ideation Scale (Eckbald amp Chapman1983) and Perceptual Aberration Scale (Chapman Chapman ampRaulin 1978) Participants completed this battery online during a1 week period Based on the results of the screening measure werecruited people who scored 196 standard deviations above themean on the abbreviated versions of the Magical Ideation Scale

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

or Perceptual Aberration Scale or a combined 3 standard deviationsabove the mean on the Magical Ideation and Perceptual AberrationScale to participate in an individual testing session We also re-cruited control participants who scored below 05 standard devia-tions above the mean on the Magical Ideation Scale PerceptualAberration Scale and Social Anhedonia Scale (Chapman Chapmanamp Raulin 1976) to take part in the individual testing session Giventhat the Social Anhedonia Scale also predicts schizophrenia-spectrum disorders (Gooding et al 2005 Kwapil 1998) followingprevious research the Social Anhedonia Scale was also included toidentify a control group (Gooding et al 2005 Kerns 2005 Kwapil1998)

In the individual testing session recruited participants com-pleted the full versions of the Magical Ideation Scale PerceptualAberration Scale and Social Anhedonia Scale From these full ver-sions of the scale participants were assigned into two groupsSchizotypal and Control Classifications were made based on previ-ous norms from large samples of similar populations (Kerns amp Ber-enbaum 2003)

412 Schizotypal groupThere were 24 participants in the schizotypal group ranging

from 18 to 20 years old with an average age of 1827 (SD = 055)Participants were 542 female 708 White 42 AfricanndashAmeri-can and 42 AsianndashAmerican and 209 other

413 Control groupThere were 31 participants in the control group ranging from 18

to 21 years old with an average age of 1835 (SD = 066) Partici-pants were 710 female 806 White 32 AfricanndashAmerican67 AsianndashAmerican and 97 other

414 ProcedureAs part of a larger study participants first completed the

Magical Ideation Scale Perceptual Aberration Scale and SocialAnhedonia Scale mixed together and titled the Survey of Attitudesand Experiences In a separate session participants completed theReferential Thinking Scale

42 Results

Pleasant and unpleasant referential thinking scores were calcu-lated in Study 3 as they were in Study 2 Participants in the schizo-typal group had both higher unpleasant referential thoughts(M = 488 SD = 280 vs M = 170 SD = 144 t (52) = 539 p lt 001effect size d = 143) and higher pleasant referential thoughts(M = 596 SD = 276 vs M = 310 SD = 204 t (52) = 438 p lt 001d = 118) than participants in the control group Next we testedwhether in either of these groups they were more likely to experi-ence unpleasant than pleasant referential thoughts Since theunpleasant referential thinking subscale had 14 items and thepleasant item subscale had 20 items unpleasant referential think-ing scores were divided by 14 and pleasant item scores were di-vided by 20 to allow for a comparison between scales Then apaired-samples t-test was run to test whether there was a differ-ence between the number of unpleasant and pleasant referentialthoughts experienced by schizotypal or control participants Therewas not a significant difference in the number of unpleasant andpleasant referential thoughts experienced by the schizotypal (t(23) = 133 p = 20) or control groups (t (29) = 153 p = 14)

43 Discussion

Study 3 found further evidence that both unpleasant and pleas-ant referential thoughts are related to other schizotypal personal-ity characteristics The schizotypal group had both elevated

leasant referential thinking Relations with self-processing paranoia andjrp201102002

DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx 9

unpleasant and pleasant referential thinking scores compared tocontrol participants

5 General discussion

The current research extended previous work on referentialthinking in several ways (Lenzenweger et al 1997) Study 1 wasthe first study to empirically examine whether referential thoughtscan be experienced as both unpleasant and pleasant Moreover theCFA in Study 1 found that unpleasant and pleasant referentialthoughts could be discriminated from each other and from para-noia with a three-factor model with separate unpleasant referen-tial thinking pleasant referential thinking and paranoia factorsproviding the best fit to the data Study 2 provided further evi-dence that unpleasant and pleasant referential thoughts could bediscriminated from each other in two ways First an item-levelCFA with unpleasant items and pleasant items on separate factorsfit the data better than a CFA with all items loading on a single fac-tor Second unpleasant and pleasant referential thoughts displayeddifferential associations with self-processing facets of narcissismand schizotypal and normal personality traits Finally study 3found that people with elevated schizotypal characteristics hadboth elevated unpleasant and elevated pleasant referentialthoughts This suggests that both unpleasant and pleasant referen-tial thinking might be important for schizophrenia-spectrumdisorders

The current research found that unpleasant and pleasant refer-ential thinking appear to be correlated but distinct constructsThese traits exhibited very different associations with self-process-ing paranoia and personality In addition the CFAs in both Study 1and Study 2 found that unpleasant and pleasant referential think-ing loaded on different factors Hence these results suggest thatthere could be important differences between unpleasant andpleasant referential thinking At the same time unpleasant andpleasant referential thinking still exhibited moderate to strongassociations with each other and both constructs were associatedwith other schizotypal personality characteristics

Taken together these results suggest that unpleasant and pleas-ant referential thinking might share some important commonmechanisms but other variables may moderate the expression ofreferential thinking For example one mechanism that might bein common between unpleasant and pleasant referential thinkingis aberrant salience Aberrant salience is the over-attribution of sal-ience to personally irrelevant objects or events and has been con-jectured to be a critical psychological mechanism in thedevelopment of psychosis (Kapur 2003) Increased aberrant sal-ience might foster the occurrence of either unpleasant or pleasantreferential thinking This might explain why both unpleasant andpleasant referential thinking are associated with other schizotypalcharacteristics associated with psychosis However whether refer-ential thoughts are experienced as unpleasant or pleasant mightdepend on other moderating factors The current research suggeststhat one moderating factor might be self-esteem Potentially thecombination of high aberrant salience and low self-esteem resultsin the occurrence of unpleasant referential thoughts In contrastthe combination of high aberrant salience and high self-esteemmight result in the occurrence of pleasant referential thoughtsHence unpleasant and pleasant referential thinking might bothshare a common mechanism such as aberrant salience but the va-lence of referential thinking might be moderated by self-esteem

The current research may also have implications for the assess-ment and conceptualization of personality disorders particularlycluster A or odd and eccentric personality disorders (AmericanPsychiatric Association 2000) Some research has suggested thatthe Big Five do not adequately account for personality characteris-

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

tics associated with schizotypal PD and that measures of lsquolsquooddityrsquorsquoor lsquolsquopeculiarityrsquorsquo may do a better job of representing schizotypal PD(Tackett Silberschmidt Krueger amp Sponheim 2008) Researchershave recently called for more work investigating these constructs(eg Watson Clark amp Chmielewski 2008) These aspects of per-sonality may be separate from Big-five personality characteristicsbut may be strongly related to Cluster A personality disorders suchas Schizotypal PD Unpleasant and pleasant referential thoughtsmay be facets of oddity and may be useful in identifying traitsand dimensions underlying personality disorders For examplethe current research found that an elevated schizotypal grouphad more unpleasant and pleasant referential thinking than thecontrol group Since the schizotypal group in Study 3 would bethought to be at least somewhat similar to a group of participantswith schizotypal personality disorder this suggests that bothunpleasant and pleasant referential thinking could be related topersonality disorders Future research could continue to examinethe relations among unpleasant referential thinking pleasant ref-erential thinking and other facets of oddity (eg odd or disorga-nized speech) which could lead to a better understanding of oddand eccentric personality disorders

The finding that there may be different types of referentialthoughts is consistent with previous theories of referential think-ing For example some previous research has suggested that refer-ential thinking may be multifaceted with differences betweenlsquolsquoguiltyrsquorsquo and lsquolsquosimplersquorsquo ideas of reference (Wing et al 1974) Fromthis perspective guilty ideas of reference involve a feeling that oth-ers are holding an individual accountable for a unpleasant outcomewhile simple ideas of reference represent referential thoughtswithout an obvious unpleasant or pleasant affective component(eg thinking people are taking special notice of you could beunpleasant or pleasant) Guilty ideas of reference may be sub-sumed within the broader construct of unpleasant referentialthinking and simple ideas of reference could fall into either cate-gory depending on the valence of the thought Thus the currentresearch is consistent with previous research that suggests thatthere may be different types of referential thoughts related tothe valence of these thoughts

The current research also provides evidence suggesting that ref-erential thinking is distinct from paranoia Previous research hasfound that referential thinking and paranoia load on the sameschizotypy factor (eg Compton et al 2009 Stefanis et al2004) However none of this research directly examined whetherreferential thinking might load on a factor separate from paranoiaIn a study that was able to directly examine this we found that ref-erential thinking and paranoia load on distinct schizotypy factors(Cicero amp Kerns 2010) However this research did not examineunpleasant versus pleasant referential thinking The current re-search examined whether unpleasant referential thinking and par-anoia might load on the same factor In current Study 1 and Study2 unpleasant referential thinking and paranoia loaded on distinctfactors Furthermore there was some evidence of differential asso-ciations between unpleasant referential thinking and paranoia asunpleasant referential thinking was more strongly associated withpleasant referential thinking and less strongly associated with neu-roticism than was paranoia Hence the current research suggeststhat even specifically unpleasant referential thinking appears tobe at least somewhat distinct from paranoia This suggests that at-tempts to measure odd and eccentric personality disorders shouldinclude distinct referential thinking and paranoia symptom dimen-sions One issue for future research would be to examine whether aCFA using additional unpleasant referential thinking scales alsofinds that unpleasant referential thinking and paranoia load on dis-tinct factors In addition another issue for future research wouldbe to further examine psychological mechanisms that might dis-tinguish referential thinking and paranoia For example it is possi-

leasant referential thinking Relations with self-processing paranoia andjrp201102002

10 DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx

ble that referential thinking might exhibit stronger associationswith aberrant salience than paranoia does

The results of this research may also have implications for thetreatment and prevention of schizophrenia Previous research sug-gests that cognitive behavioral therapy may be an effective treat-ment for schizophrenia (see Rathod Phiri and Kingdon (2010)for a review) In the current studies we found that some psy-chotic-like experiences are experienced as pleasant while otherswere experienced as unpleasant This suggests that cliniciansmay be able to focus on certain beliefs (ie the unpleasant ones)in cognitive therapy Additionally recent research has suggestedthat the identification and treatment of individuals in prodromalor early pre-psychotic stages of schizophrenia may lessen theseverity of the disorder and potentially prevent its onset altogether(Compton McGlashan amp McGorry 2007) Future research couldexamine whether unpleasant and pleasant referential thoughtscould be used to better identify people at risk for the developmentof the disorder in order to provide treatment for those individuals

Acknowledgments

Work on this article was supported by National Institute ofMental Health Grants MH072706 and MH086190 National Insti-tute on Drug Abuse Grant DA022405 National Institute on AlcoholAbuse and Alcoholism Grant AA019492 and a MU Research Boardgrant

References

Alhija F N amp Wisenbaker J (2006) A monte carlo study investigating the impactof item parceling strategies on parameter estimates and their standard errors inCFA Structural Equation Modeling 13 204ndash228 doi101207s15328007sem1302_3

American Psychiatric Association (2000) DSM-IV-TR Diagnostic and StatisticalManual of Mental Disorders (4th Text Revision ed) Washington DC AmericanPsychiatric

Asparouhov T (2006) Mean and variance adjusted test statistics Mplus WebNotes No 10

Bagge C L amp Trull T J (2003) DAPP-BQ Factor structure and relations topersonality disorder symptoms in a non-clinical sample Journal of PersonalityDisorders 17 19ndash32 doi101521pedi1711924055

Bentall R P Kaney S amp Dewey M E (1991) Paranoia and social reasoning Anattribution theory analysis British Journal of Clinical Psychology 30(Pt 1) 13ndash23

Bentall R P Kinderman P amp Kaney S (1994) The self attributional processes andabnormal beliefs Towards a model of persecutory delusions Behavior Researchand Therapy 32(3) 331ndash341 doi 0005-7967(94)90131-7 [pii]

Bosson J K Brown R P Zeigler-Hill V amp Swann W B (2003) Self-enhancementtendencies among people with high explicit self-esteem The moderating role ofimplicit self-esteem Self amp Identity 2 267ndash287 doi10108015298860390208801

Bosson J K Swann W B Jr amp Pennebaker J W (2000) Stalking the perfectmeasure of implicit self-esteem The blind men and the elephant revisitedJournal of Personality and Social Psychology 79 631ndash643 doi1010370022-3514794631

Box G E P amp Cox D R (1964) An analysis of transformations Journal of the RoyalStatistical Society 26 211ndash234

Buss A H amp Perry M (1992) The aggression questionnaire Journal of Personalityand Social Psychology 63 452ndash459 doi1010370022-3514633452

Cattell R B Eber H W amp Tatsuoka M M (1970) Handbook for the sixteenpersonality factor questionnaire Champaign IL Institute for Personality andAbility Testing

Chapman L J amp Chapman J P (1983) Infrequency Scale Unpublished TestChapman L J Chapman J P Kwapil T R Eckblad M amp Zinser M C (1994)

Putatively psychosis-prone subjects 10 years later Journal of AbnormalPsychology 103 171ndash183 doi1010370021-843X1032171

Chapman L J Chapman J P amp Raulin M L (1976) Scales for physical and socialanhedonia Journal of Abnormal Psychology 85 374ndash382 doi1010370021-843X854374

Chapman L J Chapman J P amp Raulin M L (1978a) Body-image aberration inSchizophrenia Journal of Abnormal Psychology 87 399ndash407 doi1010370021-843X874399

Chapman L J Chapman J P Raulin M L amp Edell W S (1978b) Schizotypy andthought disorder as a high risk approach to schizophrenia In G Serban (Ed)Cognitive defects in the development of mental illness (pp 351ndash360) New YorkBrunnerMazel

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

Chmielewski M Fernandes L O Yee C M amp Miller G A (1995) Ethnicity andgender in scales of psychosis proneness and mood disorders Journal ofAbnormal Psychology 104 464ndash470 doi1010370021-843X1043464

Chmielewski P M amp Watson D (2008) The heterogeneous structure ofschizotypal personality disorder Item-level factors of the SchizotypalPersonality Questionnaire and their associations with obsessive-compulsivedisorder symptoms dissociative tendencies and normal personality Journal ofAbnormal Psychology 117 364ndash376 doi1010370021-843X1172364

Cicero D C amp Kerns J G (2010) Multidimensional factor structure of positiveschizotypy Journal of Personality Disorders 24(3) 327ndash343 doi101521pedi2010243327

Combs D R amp Penn D L (2004) The role of subclinical paranoia on socialperception and behavior Schizophrenia Research 69 93ndash104 doi101016S0920-9964(03)00051-3 S0920996403000513 [pii]

Compton M T Goulding S M Bakeman R amp McClure-Tone E B (2009)Confirmation of a four-factor structure of the Schizotypal PersonalityQuestionnaire among undergraduate students Schizophrenia Research 111(1ndash3) 46ndash52 doi101016jschres200902012

Compton M T McGlashan T H amp McGorry P D (2007) Toward preventionapproaches for schizophrenia An overview of prodromal states the duration ofuntreated psychosis and early intervention paradigms Psychiatric Annals 37340ndash348

Corry N Merritt R D Mrug S amp Pamp B (2008) The factor structure of thenarcissistic personality inventory Journal of Personality Assessment 90(6)593ndash600 doi10108000223890802388590

de Jong P J (2002) Implicit self-esteem and social anxiety Differential self-favoring effects in high and low anxious individuals Behavior Research Therapy40 501ndash508 doi101016S0005-7967(01)00022-5

Demo D H (1985) The measurement of self-esteem Refining our methods Journalof Personality and Social Psychology 48 1490ndash1502 doi1010370022-35144861490

Eckbald M amp Chapman L J (1983) Magical ideation as an indicator of schizotypyJournal of Consulting and Clinical Psychology 51 215ndash225 doi1010370022-006X512215

Edell W S (1995) The psychometric measurement of schizotypy using theWisconsin Scales of Psychosis-Proneness In G Miller (Ed) The behavioral high-risk paradigm in psychopathology (pp 1ndash46) New-York Pringer-Verlag

Eysenck H J amp Eysenck S B G (1975) Manual of the eysenck personalityquestionnaire London Hodder amp Stoughton

Fenigstein A Scheier M F amp Buss A H (1975) Public and private self-consciousness Assessment and theory Journal of Consulting and ClinicalPsychology 43 522ndash527 doi101037h0076760

Fenigstein A amp Vanable P A (1992) Paranoia and self-consciousness Journal ofPersonality and Social Psychology 62 129ndash138 doi1010370022-3514621129

Freeman D (2007) Suspicious minds The psychology of persecutory delusionsClinical Psychology Review 27 425ndash457 doi101016jcpr200610004

Frith C D (1992) The cognitive neuropsychology of schizophrenia Hove HoveGoldberg L R (1999) A broad-bandwidth public health domain personality

inventory measuring the lower-level facets of several five-factor models In IMervielde I Deary F De Fruyt amp F Ostendorf (Eds) Personality psychology inEurope (Vol 77 pp 7ndash28)

Gooding D C Tallent K A amp Matts C W (2005) Clinical status of at-riskindividuals 5 years later Further validation of the psychometric high-riskstrategy Journal of Abnormal Psychology 114 170ndash175 doi1010370021-843X1141170

Greenwald A G amp Farnham S D (2000) Using the implicit association test tomeasure self-esteem and self-concept Journal of Personality and SocialPsychology 79 1022ndash1038 doi1010370022-35147961022

Hathaway S R amp McKinley J C (1989) Manual of the Minnesota multiphasicpersonality inventory Minneapolis MN University of Minnesota Press

Hewitt J K amp Claridge G S (1989) The factor structure of schizotypy in a normalpopulation Personality and Individual Differences 10 323ndash329 doi1010160191-886928892990105-0

Hu L T amp Bentler P M (1998) Fit indices in covariance structure modelingSensitivity to underparameterized model misspecification PsychologicalMethods 3 424ndash453 doi1010371082-989X34424

Kapur S (2003) Psychosis as a state of aberrant salience A framework linkingbiology phenomenology and pharmacology in schizophrenia American Journalof Psychiatry 160 13ndash23 doi101176appiajp 160113

Kernis M H (2003) Toward a conceptualization of optimal self-esteemPsychological Inquiry 14 1ndash26 doi101207S15327965PLI1401_01

Kerns J G (2005) Positive schizotypy and emotion processing Journal of AbnormalPsychology 114 392ndash401 doi1010370021-843X1143392

Kerns J G amp Berenbaum H (2003) The relationship between formal thoughtdisorder and executive functioning component processes Journal of AbnormalPsychology 112 339ndash352 doi1010370021-843X1123339

King L A amp Hicks J A (2009) Positive affect intuition and referential thinkingPersonality and Individual Differences 46 719ndash724

Kwapil T R (1998) Social anhedonia as a predictor of the development ofschizophrenia-spectrum disorders Journal of Abnormal Psychology 107558ndash565 doi1010370021-843X1074558

Kwapil T R Barrantes-Vidal N amp Silvia P J (2008) The dimensional structure ofthe Wisconsin Schizotypy Scales Factor identification and construct validitySchizophrenia Bulletin 34 444ndash457 doi101093schbulsbm098

leasant referential thinking Relations with self-processing paranoia andjrp201102002

DC Cicero JG Kerns Journal of Research in Personality xxx (2011) xxxndashxxx 11

Leary M R Tambor E S Terdal S K amp Downs D L (1995) Self-esteem as aninterpersonal monitor The sociometer hypothesis Journal of Personality andSocial Psychology 68 518ndash530 doi1010370022-3514683518

Lenzenweger M F (1994) Psychometric high-risk paradigm perceptualaberrations and schizotypy An update Schizophrenia Bulletin 20 121ndash135

Lenzenweger M F Bennett M E amp Lilenfeld L R (1997) The Referential ThinkingScale as a measure of schizotypy Scale development and initial constructvalidation Psychological Assessment 9 452ndash463 doi1010371040-359094452

Lewis A (1970) Paranoia and paranoid A historical perspective PsychologicalMedicine 1(1) 2ndash12

Lischetzke T amp Eid M (2003) Is attention to feelings beneficial or detrimental toaffective well-being Mood regulation as a moderator variable Emotion 3(4)361ndash377 doi1010371528-354234361

Livesley W J amp Jackson D N (2002) Dimensional assessment of personalitypathology-basic questionnaire (DAPP-BQ) Port Huron MI Sigma Press

Meehl P E (1962) Schizotaxia schizotypy and schizophrenia AmericanPsychologist 17 827ndash838 doi101037h0041029

Meng X Rosenthal R amp Rubin D B (1992) Comparing correlated correlationcoefficients Psychological Bulletin 111 172ndash175 doi1010370033-29091111172

Meyer E C amp Lenzenweger M F (2009) The specificity of referential thinking Acomparison of schizotypy and social anxiety Psychiatry Research 165 78ndash87doi101016jpsychres200710015

Miller G A (Ed) (1995) The behavioral high-risk paradigm in psychopathology NewYork City Springer

Moller P amp Husby R (2000) The initial prodrome in schizophrenia Searching fornaturalistic core dimensions of experience and behavior Schizophrenia Bulletin26(1) 217ndash232

Muthen L K amp Muthen B O (2004) Mplus userrsquos guide (3rd ed) Los Angeles CAMuthen amp Muthen

Raballo A Saebye D amp Parnas J (2009) Looking at the Schizophrenia SpectrumThrough the Prism of Self-disorders An Empirical Study Schizophrenia Bulletindoi101093schbulsbp056

Raine A (1991) The SPQ A scale for the assessment of schizotypal personalitybased on DSM-III-R criteria Schizophrenia Bulletin 17 555ndash564

Raine A (2006) Schizotypal personality Neurodevelopmental and psychosocialtrajectories Annual Review of Clinical Psychology 2 291ndash326 doi101146annurevclinpsy2022305095318

Raine A Reynolds C Lencz T Scerbo A Triphon N amp Kim D (1994) Cognitive-perceptual interpersonal and disorganized features of schizotypal personalitySchizophrenia Bulletin 20 191ndash201

Raskin R amp Terry H (1988) A principal-components analysis of the NarcissisticPersonality Inventory and further evidence of its construct validity Journal ofPersonality and Social Psychology 54(5) 890ndash902 doi1010370022-3514545890

Rathod S Phiri P amp Kingdon D (2010) Cognitive behavioral therapy forschizophrenia The Psychiatric Clinics of North America 33(3) 527ndash536doi101016jpsc201004009

Please cite this article in press as Cicero D C amp Kerns J G Unpleasant and pother schizotypal traits Journal of Research in Personality (2011) doi101016j

Rawlings D amp Freeman J L (1996) A questionnaire for the measurement ofparanoiasuspiciousness British Journal of Clinical Psychology 35(Pt 3) 451ndash461

Rodebaugh T L Woods C M amp Heimberg R G (2007) The reverse of socialanxiety is not always the opposite The reverse-scored items of the socialinteraction anxiety scale do not belong Behavior Therapy 38(2) 192ndash206doi101016jbeth200608001

Rosenberg M (1965) Society and the adolescent self-image Princeton New JerseyPrinceton University Press

Satorra A amp Bentler P A (1994) Corrections to test statistics and standard errorsin covariance structure analysis In A von Eye amp C C Clogg (Eds) Latentvariables analysis Applications for developmental research Thousand Oaks SageCA

Satorra A amp Bentler P A (2001) A scale difference chi-square test statistic formoment structure analysis Psychometrika 66 507ndash514 doi101007BF02296192

Schimmack U amp Diener E (2003) Predictive validity of explicit and implicit self-esteem for subjective well-being Journal of Research in Personality 37 100ndash106doi101016S0092-6566(02)00532-9

Sedikides C Rudich E A Gregg A P Kumashiro M amp Rusbult C (2004) Arenormal narcissists psychologically healthy Self-esteem matters Journal ofPersonality and Social Psychology 87(3) 400ndash416 doi1010370022-3514873400

Startup M amp Startup S (2005) On two kinds of delusions of reference PsychiatryResearch 137 87ndash92

Stefanis N C Smyrnis N Avramopoulos D Evdokimidis I Ntzoufras I ampStefanis C N (2004) Factorial composition of self-rated schizotypal traitsamong young males undergoing military training Schizophrenia Bulletin 30335ndash350

Tackett J L Silberschmidt A L Krueger R F amp Sponheim S R (2008) Adimensional model of personality disorder Incorporating DSM Cluster Acharacteristics Journal of Abnormal Psychology 117 454ndash459 doi1010370021-843X1172454

Tafarodi R W amp Swann W B Jr (1995) Self-liking and self-competence asdimensions of global self-esteem Initial validation of a measure Journal ofPersonality Assessment 65 322ndash342 doi101207s15327752jpa6502_8

Tucker L R amp Lewis C (1973) A reliability coefficient for maximum likelihoodfactor analysis Psychometrika 38 1ndash10

Useda D amp Trull T J (2002) The construct validity of the paranoid personalitydisorder features questionnaire (PPDFQ) A dimensional assessment of paranoidpersonality disorder Unpublished Dissertation

Watson D Clark L A amp Chmielewski M (2008) Structures of personality andtheir relevance to psychopathology II Further articulation of a comprehensiveunified trait structure Journal of Personality 76 1545ndash1585 doi101111j1467-6494200800531x

Wing J K Cooper J E amp Sartorious N (1974) Measurement and classification ofpsychiatric symptoms An instruction manual for the PSE and Catego ProgramCambridge England Cambridge University Press

leasant referential thinking Relations with self-processing paranoia andjrp201102002