UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Extending interpersonal problems to include the "big five" personality dimensions Pincus, Aaron Lee 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1992_fall_pincus_aaron_lee.pdf [ 2.76MB ]
JSON: 831-1.0086607.json
JSON-LD: 831-1.0086607-ld.json
RDF/XML (Pretty): 831-1.0086607-rdf.xml
RDF/JSON: 831-1.0086607-rdf.json
Turtle: 831-1.0086607-turtle.txt
N-Triples: 831-1.0086607-rdf-ntriples.txt
Original Record: 831-1.0086607-source.json
Full Text

Full Text

EXTENDING INTERPERSONAL PROBLEMS TO INCLUDE THE“BIG FIVE” PERSONALITY DIMENSIONSbyAARON LEE PINCUSB.S. (High Honours), University of California, Davis, 1985M.A. University of California, Berkeley, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF PSYCHOLOGYWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1992© Aaron Lee Pincus, 1992Signature(s) removed to protect privacyIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives, It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________________________Department of PsychologyThe University of British ColumbiaVancouver, CanadaDate September 24. 1992DE-6 (2/88)Signature(s) removed to protect privacyiiABSTRACTCurrent advances in the science of personality structurewere discussed with reference to the utility of assessingdimensions of normal personality for clinical research andpsychological treatment. The current debate in the field ofclinical personality assessment suggests that the extremebehaviors, attitudes, symptoms, and actions seen in patientsin clinical settings may be widely separated from behaviorsin the normal range, and hence, current conceptions of normalpersonality structure may be insufficient to provide usefulinformation. The basic dimensions of normal personality havenot been directly related to maladaptive, rigid, and abnormalbehavior seen in psychopathology.Based on the construct of interpersonal problems(Horowitz, 1979), it was proposed that two ways personalitytraits may be expressed rigidly and maladaptively are viachronic behavioral excesses (behaviors a person does toomuch) and chronic behavioral inhibitions (behaviors a personfinds hard to do). Three investigations were conducted todetermine if this operationalization of maladaptive behaviorwould lead to the identification of a taxonomy of personalitytraits of particular relevance to clinical assessment andtreatment. Additionally, specific methodological techniqueswere used to impose a taxonomic structure on the traitdomain, conforming to the Dyadic—Interactional Five-factorModel of personality structure (Pincus & Wiggins, in press;Trapnell & Wiggins, 1990; Wiggins & Pincus, 1992, in press).iiiIn the current research, a valid and reliable self—report instrument was derived in a large normal sample,cross—validated on an independent normal sample, and cross—validated on a small psychiatric sample. This instrumentextended a recent circumplex modification of Horowitz’ (1979)Inventory of Interpersonal Problems (lIP-C; Alden, Wiggins, &Pincus, 1990) to include the three additional personalitydimensions of neuroticism, conscientiousness, and openness toexperience. The final inventory, the IIP-B5, is a 140 itemquestionnaire that assesses maladaptive trait expression(problems) related to the five basic personality dimensionsby assessing a number of lower—order problems facets withineach superordinate trait domain.The advantages of the IIP-B5 compared to currentlyavailable five—factor model inventories for clinicalassessment and research was discussed. The instrument wasused to operationalize the five—factor model of personalityin a fourth study comparing a competing model of adjustmentto the five-factor model of personality. The clinicalutility of the IIP-B5 was demonstrated in a brief casepresentation of a patient who was seen for intensive grouppsychotherapy with the author.Results of all studies suggest that a taxonomy ofmaladaptive personality traits can be subsumed by the fivefactor model of personality and that a dimensionalperspective on abnormal behavior may be a viable alternativeto the categorical classification system of the DSM-III.ivSome structural weaknesses of the IIP—B5 were identified andfurther improvements and investigations are required.VTABLE OF CONTENTSABSTRACT.iiTABLE OF CONTENTS vLIST OF TABLES xLIST OF FIGURES XVACKNOWLEDGEMENT xviCHAPTER 1: Introduction 1Introduction 1The Circumplex Model of Interpersonal Behavior 4The Five-factor Model of Personality 5History and Current Perspectives 6Integrating the Circumplex and Five-factor Model 14Recent Applications in Clinical Psychology 26Interpersonal Problems & Maladaptive Trait Expression.33Maladaptive Trait Expression 41Brief Review of Research on Interpersonal Probleius.42Combining Personality Structure and Clinical Research.43Openness Items .66Principal Components of Selected Items 70Superordinate Structure of Derived Problems Facets.75Reliability 85Structural Validity 87Convergent Validity 89Discussion 93General Factor 93Scale Derivation 94Structural Fidelity 96General Conclusion 97viCHAPTER 2: Scale DerivationMethodsOverviewSampleInstrumentsProceduresAnalysesIpsatizationData ReductionReliability and ValidityExpectationsResultsIpsatizationDerivation of Problems Scales for N,Neuroticism ItemsConscientiousness Items454545484851515152545455550, and C 585959viiChapter 3: Scale Validation 99Methods 99Overview 99Sample 99Instruments 99Procedures 101Analyses 101Data Transformation 101Cross Validation 101Structural Validation 101Reliability and Validity 102Expectations 102Results 103Item Modifications 103Facet Validation 104Structural Validation 110Reliability 115Convergent Validity 117IIP—B5 Scale Norms 127Discussion 127vi IiChapter 4: Psychiatric Validation. 140Methods 140Overview 140Sample 140Instruments 141Procedures 141Analyses 142Data Transformation 142Structural Validation 142Reliability and Validity 142Expectations 143Results 143Structural Validation 143Reliability 150Convergent Validity 152Discussion 155Neuroticism 156Unconsientiousness 157Openness 158ixChapter 5: The Weinberger Adjustment Typology and theFive—factor Model of Personality 161Introduction 161The Weinberger Adjustment Typology 161Adjustment and the “Big Five” 165Methods 166Sample 166Instruments 166Procedures 167Analyses 167Hypotheses 168Results 168Correlations 168Structural Analyses 170Assessment of the WAI Adjustment Typology 173Discussion 177Chapter 6: A Case Presentation 183Introduction 183Discussion 189References 191xLIST OF TABLESTable 1. Costa & McCrae’s Enduring-DispositionalPerspective 10Table 2. Hogan’s Social-Competency Perspective 12Table 3. Goldberg’s Lexical Perspective 13Table 4. Wiggins’ Dyadic-Interactional Perspective 15Table 5. Inventory of Interpersonal Problems CircumplexScales 39Table 6. Intercorrelations between the general factor ofthe Inventory of Interpersonal Problems—Big Fiveversion and the domain scales of the NEO Five—Factor Inventory, the Extended InterpersonalAdjective Scales, and mean endorsement acrossthe item pool 57Table 7. Principal components of neurotic problems items.60Table 8. Derived neuroticism facets 61Table 9. Neuroticisin component intercorrelations 62Table 10. Principal components of conscientiousnessproblems items 63Table 11. Derived conscientiousness facets 64Table 12. Conscientiousness component intercorrelations. .66Table 13. Principal components of openness problems items.67Table 14 Derived openness facets 68Table 15. Openness component intercorrelations 69Table 16. Principal components of selected neuroticismitems 71Table 17. Neuroticism component intercorrelations II 72xiTable 18. Principal components of selectedconscientiousness items 73Table 19. Conscientiousness componentintercorrelations II 74Table 20. Principal components of selected openness items.76Table 21. Openness component intercorrelations II 77Table 22. Three component solution of conjoint principalcomponents analysis of derived problem facetsand domain scales of the IASR-B5 78Table 23. Correlations of Oversensitivity with domainscales of the IASR—B5 79Table 24. Correlations of Nonconformity with domainscales of the IASR-B5 80Table 25. Principal components analysis of Opennessproblems facets with Nonconformity ipsatized. . .81Table 26. Principal components analysis of Opennessproblems facets with Nonconformity unipsatized .82Table 27. Three component solution of conjoint principalcomponents analysis of remaining derivedproblem facets and domain scales of theIASR—B5 84Table 28. Alpha coefficients for derived facet and domainscales 86Table 29. Conjoint principal components analysis ofIASR—B5 domain scales and IIP—B5 domain scales .87Table 30. Full facet and circumplex octant principalcomponents analysis of the IIP-B5 88xiiTable 31. Correlations between IIP—B5 domain scales anddomain scales of the FF1 and IASR-B5 90Table 32. Convergent validity correlations between IIP-B5facets and the STAI Trait Anxiety scale, theInterpersonal Dependency Inventory, and theShyness scale 91Table 33. Summary of item modifications and additions toderived IIP—B5 facets 103Table 34. Principal components of selected neuroticismitems 105Table 35. Principal components of selected opennessitems 106Table 36. Principal components of selectedunconscientious items 108Table 37. Three component solution of conjoint principalcomponents analysis of IIP—B5 facets anddomain scales of the IASR-B5 111Table 38. Full facet and circumplex octant principalcomponents analysis of the IIP-B5 113Table 39. Conjoint principal components analysis ofIASR-B5 domain scales and IIP-B5 domain scales.ll4Table 40. Alpha coefficients for derived facet anddomain scales 116Table 41. Correlations between IIP—B5 domain scales anddomain scales of the FF1 and IASR-B5 117xiiiTable 42. External validity correlations between IIP-B5neuroticism facets and the Brief SymptomInventory 119Table 43. External validity correlations between IIP-B5facets and the Fenigstein Public and PrivateSelf—Consciousness scales 121Table 44. External validity correlations between IIP-B5facets and the Weinberger Adjustment Inventory.l22Table 45. Means and standard deviations of IIP-B5 facetand domain scales 124Table 46. IIP-B5 facet scales for neuroticism, openness,and unconscientiousness 131Table 47. Diagnostic breakdown of psychiatric sample 140Table 48. Three component solution of conjoint principalcomponents analysis of IIP—B5 facets and domainscales of the IASR—B5 144Table 49. Full facet and circumplex octant principalcomponents analysis of the IIP-B5 146Table 50. Conjoint principal components analysis ofNEO-PI domain scales and IIP-B5 domain scales .148Table 51. Intercorrelations among IIP—B5 openness facetsand domain scales and domain scales of theNEO-PI in a psychiatric sample 149Table 52. Alpha coefficients for IIP—B5 facet and domainscales in a psychiatric sample 151Table 53. Correlations between IIP—B5 domain scales anddomain scales of the NEO-PI 152xivTable 54. Correlations between IIP-B5 neuroticism facetsand NEO-PI neuroticism facets 153Table 55. Correlations between IIP—B5 openness facets andNEO—PI openness facets 153Table 56. Weinberger and Schwartz’ (1990) hypothesizedrelations between WAI scales and the Big Five .163Table 57. The Weinberger Adjustment Typology 165Table 58. Correlations between the WAI and the IIP-B5. . .169Table 59. Conjoint principal components analysis of theWAI subscales and the IIP—B5 domain scales 171Table 60. Conjoint principal components analysis of theWAI Restraint and Distress scales and theIIP-B5 domain scales 172Table 61. Tukey test comparisons of adjustment groupdifferences on dimensions of the Big Five 175Table 62. Tukey test comparisons of adjustment groupdifferences on facets of unconscientiousness. . .176xvFigure 1.Figure 2.Figure 3.Figure 4.Figure 5.Figure 6.Figure 7.Figure 8.LIST OF FIGURESThe circumplex model of interpersonal behaviorand the five-factor model of personalitystructure 3Projections of the interpersonal dimensions ofthe Hogan Personality Inventory and the NEOPersonality Inventory onto the RevisedInterpersonal Adjective Scales circumplex 18Interpersonal profiles of analogue cohortsrepresenting schizoid and borderlinepersonality disorders 25The interpersonal problems circumplex 38IIP—B5 domain scale profile 125IIP—B5 facet profile 126IIP-B5 domain scale profile for S. C 184IIP—B5 facet profile for S. C 186xviACKNOWLEDGEMENTI wish to thank Jerry S. Wiggins--mentor, teacher,colleague, and friend. There is no relationship that livesup to the cliche “I wouldn’t be where I am today without him”more appropriately than my five years with you. Jerry, Iwill always be thankful for your wisdom, your respect for me,your support and encouragement, your generosity, and thefreedom you allowed me in my academic endeavors.I would also like to thank my first mentor Alan Elms andmy second Gerald Mendelsohn, both of whom cultivated myinterests and enthusiasm for psychology. Thanks to OliverJohn who was the first to suggest I make the trip up to UBC.Since my arrival, a number of people have become veryspecial to me. I fondly thank Darrin Lehman for everythingacademic and not so academic, Anita DeLongis for her supportand ability to contain my neurotic fears, Lynn Alden foralways treating me as a colleague and great clinicalsupervision, and Dimitri Papageorgis for acknowledging theabsurdities of academia. Finally I thank my wife, Kim, whosaw me through it.What a long strange trip it’s been.1CHAPTER 1IntroductionDuring the decade of the 1980’s, a renewed (andimproved) wave of interest emerged in the application ofpersonality traits and normal personality assessment forresearch purposes in clinical psychology. This wave ofinterest coincided with considerable advances in the scienceof personality structure; and in many ways, the two domainshave complemented each other. By the mid—1980’s, a number oftheoretical perspectives on trait structure and ontology, andtheir associated personality assessment methodologies wereproposed as useful additions to the domain of clinicalassessment. Buss and Craik (1986) applied their ActFrequency Approach to personality assessment (e.g., Buss &Craik, 1983, 1984) to the categories of personality disorderfound in the Diagnostic and Statistical Manual of MentalDisorders (DSM-III; American Psychiatric Association, 1980).Wiggins (1982), Kiesler (1983, 1986), and Horowitz and Vitkus(1986) reviewed and extended the interpersonal theory ofpersonality and the application of the circumplex model ofinterpersonal behavior to a number of domains in clinicalpsychology. Costa and McCrae (1986a) proposed that“recognition of the stability of normal personalitytraits. . .points to important similarities between normalpersonality and personality disorders, facilitates researchon the psychological processes that maintain both adaptiveand maladaptive traits, reminds the clinician that current2problems in functioning may be the expression of enduringpersonality patterns, and promotes more realisticexpectations about how much therapeutic change is possible”(pp. 415—416).These proposals both advanced areas of inquiry andspurred ongoing debates regarding the merits of normalpersonality assessment in clinical psychology (Wailer & BenPorath, 1987). Adoption of a particular model of normalpersonality organization was, in fact, equally rich withdiverse opinions. For example, Costa and McCrae (l986b)proposed that the most useful model of personality forresearch applications in clinical psychology is the Five-factor Model (FFM) of personality trait organization, whichhas a long developmental history to be reviewed later (seeWiggins & Trapnell, in press). Of the numerous models ofpersonality trait structure proposed over the years, two haverecently emerged as the most enduring, most widely applied,and most complementary——the circumplex model of interpersonalbehavior and the FFM of personality (Wiggins & Pincus, 1992;see Figure 1).This emerging consensus in the field of personalitypsychology has generally refrained basic questions. Thedebate regarding the clinical utility of normal personalityassessment is now generally couched in terms of the utilityof either the FFM (Ben—Porath & Wailer, 1992a, 1992b; Costa &McCrae, i992a, 1992b; Widiger, in press), the interpersonalcircumplex (Alden, Wiggins, & Pincus, 1990; Kiesier, 1991;tIij‘HFIveFactorModelCircumplexModelAssured-‘-3H-3Dominant<CD(D IC)H)GregariousExtravertedSURGENCY/EXTRAVERSIONCDoxIIAGREEABLENESS(D ‘-‘0IIICONSCIENTIOUSNESSoCDCold-Warm-heartedAgreeable0IVNEUROTICISMCD ti (flHVOPENNESSTOEXPERIENCEPJCDH-‘5UnassumingIngenuous0rtC) .<H 0 p.)Arrogant-CalculatingUnassured-Submissive4Van Denburg, Schmidt, & Kiesler, in press), or a combinationof both models (Carson, in press, Pincus & Wiggins, in press;Trapnell & Wiggins, 1990). In order to fully evaluate theutility of normal personality assessment in clinicalsettings, and to identify possible shortcomings to beimproved upon, a brief review of both models and their recentclinical applications is necessary.The Circumplex Model of Interoersonal BehaviorThe circumplex model f interpersonal behavior has itsroots in Sullivan’s (1953a, 1953b) interpersonal theory ofpsychiatry and the psychotherapy research begun in the 1950’sby Timothy Leary and the Kaiser Foundation group (Freedman,Leary, Ossorio, & Cof fey, 1951; Leary, 1957). Interpersonaltheory proposes that personality is best viewed in ternis ofrecurrent interpersonal dispositions or tendencies to displaycertain characteristic patterns of interpersonal behavior(e.g., Carson, 1969; Leary, 1957; Sullivan, l953b; Wiggins,Phillips, & Trapnell, 1989). There is general agreement thatthe appropriate structural model for representinginterpersonal traits is a two—dimensional circumplex(Guttinan, 1954) in which variables are ordered in a circulararrangement around the basic orthogonal dimensions ofdominance (vs. submission) and nurturance (vs. coldness)(e.g., Benjamin, 1974; Kiesler, 1983; Leary, 1957, Lorr &McNair, 1963; Wiggins, 1979).5By basic, interpersonal theorists mean that ininterpersonal transactions and ongoing relationships, wecontinually define and negotiate our behavior with respect tothe relative status of interactants and the quality of theengagement, from friendly-engaged to hostile-detached(Benjamin, 1974; Foa & Foa, 1974; Kiesler, 1983; Wiggins,1979, 1980). Over the years, this model has beenoperationalized by a variety of instruments such asBenjamin’s Structural Analysis of Social Behavior (SASB;Benjamin, 1974, 1984, 1988), Wiggins’ Revised InterpersonalAdjective Scales (IAS-R; Wiggins, Trapnell, & Phillips,1988), Lorr & McNair’s Interpersonal Behavior Inventory (IBI;Lorr & McNair, 1965), LaForge and Suczek’s InterpersonalCheck List (ICL; Laforge & Suczek, 1955), Alden, Wiggins, &Pincus’s Inventory of Interpersonal Problems CircumplexScales (lIP—C; Alden et al, 1990), and Kiesler’s 1982Interpersonal Circle (IPC; Kiesler, 1983) and Impact MessageInventory Form hA (IMI-IIA; Kiesler & Schmidt, 1991). Whileall of the inventories mentioned purportedly assess acircumplex model, the empirical structure of many remainunevaluated (e.g., IPC, SASB) or manifest structuralshortcomings (e.g., ICL--see Paddock & Nowicki, 1986).The Five-Factor Model of PersonalityA review of the FF14 is a difficult task, as the currentZeitgeist of the field of personality structure includesnumerous and voluminous efforts to articulate the long and6interrupted history of the model. These efforts includeAnnual Review chapters (e.g., Digman, 1990), book chapters(e.g., Wiggins & Trapnell, in press), special issues ofjournals (Costa, 1991; McCrae & John, in press), and anupcoming edited Handbook of Personality Psychology withchapters devoted to each dimension of the model (Briggs,Hogan, & Jones, in press). The FF14 has been the subject ofconsiderable research over the last 10 years and that timespan covers only one fifth of the model’s history. Thus acomprehensive review would be well beyond the scope of thepresent introduction. The following introduction to themodel will thus emphasize important features of the FFM’shistory and current applications that are of significance forthe present research. Readers interested in a comprehensivereview are directed to the primary literature just cited.History and Current Perspectives. The FF14 consisting ofthe superordinate trait dimensions of extraversion,agreeableness, neuroticism, conscientiousness, and opennessto experience, has developmental roots tracing back to themultivariate factor analytic traditions of Cattell (1943,1945, 1957), Guilford (1948, 1959), and Eysenck (1947). Themodel’s genesis might well be traced back to Allport andOdbert’s (1936) initial attempts to assemble a comprehensivelexicon of trait-descriptive terms in the English language.Cattell was the first investigator to apply factor analysisto trait-descriptive terms, although he has generally opposedreducing factor analytic solutions to superordinate higher7order dimensions (Cattell, 1973). It was Donald Fiske (1949)who first identified a superordinate five—factor structurewithin the Cattell system. The following decade of the1950’s was rich with both the expansion of factor analyticinvestigations and the development of construct validityintroduced by Cronbach and Meehi (1955) and Jane Loevinger(1957). However, little research focussing on the FFM wasconducted. Despite the interrupted development of the FFM,the fundamentals of assessment advanced during the decadecreated the foundations for further articulation anddevelopment of the model (Wiggins & Trapnell, in press).The first true five factor advocates were Tupes andChristal (1958, 1961) and Warren Norman (1963), who providedboth convincing empirical data that personality could becaptured by five orthogonal dimensions, and a clear rationaleand scientific direction to validate and generalize theirfindings. Through the 1960’s and 1970’s methodological workcontinued to provide comprehensive evidence of the FFM.However, influenced by the critique on personality traitontology started by Walter Mischel (1968) and themethodological, rather than substantive emphasis ofpersonality structure over the years, many came to viewmultivariate trait investigations as mere number crunchingwith little relation to people whose personalities werepresumably being described.The FFM advocates and investigators of the 1980’s and1990’s and trait psychology in general have clearly turned8that perspective around. In the most recent Annual Reviewchapter, Wiggins and Pincus (1992) note a strong resurgencein the fundamental assertion that traits are real.Compelling arguments by individuals such as Auke Tellegen(1991), Arnold Buss (1989), David Buss (1991), David Funder(1991), and McCrae and Costa (1990), as well as theincreasing evidence of genetic heritability of traits (e.g.,Henderson, 1982; Plomin & Daniels, 1987; Rushton, Fulker,Neale, Nias, & Eysenck, 1986; Tellegen, Lykken, Bouchard,Wilcox, Segal, & Rich, 1988) are converging to provide a newtrait ontology that in many ways returns to and improves uponGordon Allport’s (1937) original assertions. And, thesearguments are congruent with assertions made within theinterpersonal tradition of personality. From Sullivan(l953b) onward, the assumption has been that the traitsinfluencing interpersonal behavior have real biological andpsychological substrates that are transactionally influencedby, and concurrently influence, the development of theindividual.Wiggins and Pincus (1992, in press) note at least fourmajor ontological perspectives on the FFM, one of which willbe emphasized in the following section. These perspectivesdiffer in their foci of convenience, theoreticalorientations, universes of content, assessment instruments,and representative applications. Collectively, theseperspectives, and the empirical work conducted within them,demonstrate the comprehensiveness of the FFM (Wiggins &9Pincus, 1992, p. 479). The picture gleaned from review ofthese perspectives is that the FFM endorsed today has a truesubstantive basis in addition to methodological and empiricalsupport (see Tables 1 to 4).Costa & McCrae’s Edurinc—Dispositional view asserts thatit is the enduring nature of individual’s behaviors andexperiences, which is what we apply trait names to, thatmakes them so central to, and predictive of, the ways inwhich emerging lives develop (Costa & McCrae, 1980; McCrae &Costa, 1990). Their demonstrations of longitudinal stability(Costa & McCrae, l988b) and convergent validity (McCrae &Costa, 1987) provide the empirical basis for theirassertions. Using the NEO Personality Inventory (NEO-PI;Costa & McCrae, 1985), these investigators have demonstratedthat virtually all systems of personality trait organizationcan be interpreted within the FFM framework (Costa, Busch,Zonderman, & McCrae, 1986; Costa & McCrae, 1988a, 1990;Costa, McCrae, & Holland, 1984; McCrae & Costa, l985a, l989a,1989b; McCrae, Costa, & Busch, 1986; NcCrae, Costa, &Piedmont, in press; Piedmont, NcCrae, & Costa, 1991, 1992;Watson & Clark, in press).Robert Hogan’s Social—Competency perspective assertsthat the basic needs of humans as social beings are approval,status, and predictability (Hogan, 1983, 1986; Hogan, Cheek,10Table 1. Costa & McCrae’s Enduring-Dispositional PerspectiveFocus of Convenience longitudinal studies of personalityand agingTheoretical Orientation traditional multivariate traittheory of individual differences isa legitimate alternative to othertheories of personalityUniverse of Content literature review of earlier scaleswith reference to study of agingAssessment Instruments NEO—PI: domains of extraversion,agreeableness, conscientiousness,neuroticism, and openness; eachmeasured by six facetsRepresentative relations between NEO—PI andApplication instruments from major researchtraditions in personality assessment11& Jones, 1985; Hogan & Johnson, 1981). Gratification ofthese needs are sought through ritualized interactions withothers. Trait attributions in the natural language evolvedas a means of social control necessitated by thecircumstances of group living. Thus, for Hogan, the Big Five,are exclusively dimensions employed by observers in theevaluation of actors’ contributions to social living groupsor work organizations (Hogan, Carpenter, Briggs, & Hansson,1984).Lewis Goldberg is the most prominent figure in the longstanding Lexical approach to the study of personality (John,Angleitner, & Ostendorf, 1988). Lexical approaches to thestudy of personality structure have the most resemblance tothe original multivariate approaches to trait structure.However, Goldberg’s work is not merely number crunching. Thelexical approach asserts that those individual differencesthat are most important in the daily transactions of personswill eventually become encoded in their natural language.And, the more important an individual difference is for humantransactions, the more languages that will have a term for it(Goldberg, 1981, 1982). The work of Goldberg and hiscolleagues (Goldberg, 1990, 1992; John, l990a, 1990b; Peabody& Goldberg, 1989) continues to provide evidence for the FFMin the structure of natural language trait descriptors. Hecontinues to engage in and motivate others to pursue crosscultural linguistic investigations of trait structure withgood success (Angleitner & Ostendorf, 1989; Angleitner,12Table 2. Hogan’s Social—Competency PerspectiveFocus of Convenience prediction of effective performancein work and social settingsTheoretical Orientation actors have needs for socialapproval, status, andpredictability; observers use traitterms to evaluate social usefulnessof actorsUniverse of Content review of earlier five—factorstudies from a social—competencyperspectiveAssessment Instruments j: primary scales of ambition,likability, sociability, adjustment,prudence, and intellectance; eachmeasured by subsets of homogeneousitem clusters (HICs)ReDresentative prediction of organizational andApplication occupational performance13Table 3. Goldberg’s Lexical PerspectiveFocus of Convenience development of compelling taxonomyof personality-descriptive terms inthe natural languageTheoretical Orientation those individual differences thatare of the most significance in thedaily transactions of persons willeventually become encoded in theirnatural languageUniverse of Content semantic relations among traitterms selected from dictionarysearchesAssessment Instruments Standard Markers: domain scores forsurgency, agreeableness,conscientiousness, emotionalstability, and intellect; eachmarked by 20 adjectivesRepresentative investigations of the generali—application zability of English taxonomy toDutch and German languages14Ostendorf, & John, 1990; Borkenau & Osteridorf, 1989; Hofstee,1990; Hofstee & Van Heck, 1990: Ostendorf, 1990; Ostendorf &Angleitner, 1990).Integrating the Circumplex and FFM.Over the last five years it has become increasinglyapparent to those who are familiar with both systems, thatthe circumplex model of interpersonal behavior and the FFMare complementary rather than competing models of traitstructure (e.g., McCrae & Costa, l989b). Wiggins and hiscolleagues (Pincus & Wiggins, in press; Trapnell & Wiggins,1990; Wiggins & Pincus, in press; Wiggins & Trapnell, inpress) have proposed a Dyadic—Interactional perspective onthe FFM that supplements circumplex classifications ofinterpersonal behaviors with reference to the additionaldimensions of conscientiousness, neuroticism, and openness.Considerable empirical, theoretical, and practicalevidence has accumulated that argues for the integration ofthe interpersonal circumplex and the FFM perspectives ontrait structure. The circumplex model provides analternative to the simple—structure model of factor analysisin which all variables are expected to have their principalloadings on one or the other of two orthogonal factors. TheDyadic—Interactional perspective argues that the circumplexbe used to represent the first two dimensions of the FFM. Inthe circumplex, there is no optimal rotation of the principalaxes, because on only empirical grounds, no15Table 4 wiggins’ Dyadic-Interactional PerspectiveFocus of Convenience dyadic interactions in psychotherapeutic settingsTheoretical Orientation agency and communion arepropaedutic to the study ofcharacterological, emotional, andcognitive dispositionsUniverse of Content theoretically—based taxonomy oftrait terms derived from Goldberg’searlier taxonomyAssessment Instruments IASR—B5: domains of dominance andnurturance form eight circumplexscales; domain scores for openness,conscientiousness, and neuroticismRepresentative relations between IASR-B5 andapplication conceptions of personality disorders16rotation is superior. Therefore arguments which emphasizethat placing the axes through dominance and nurturanceprovides considerable advantages over simple—structureassessment of extraversion and agreeableness for the study ofinterpersonal behavior must be presented.There are a number of empirical advantages of thecircumplex model. The geometric properties of precisecircumplex measurement allow investigators to plot outsidevariables onto two dimensional interpersonal space andevaluate their relations to interpersonal traits, thusoperationalizing a superordinate framework to organize aninterpersonal trait taxonomy (Gurtman, 1991; Wiggins &Broughton, 1985, 1991). To locate any outside variable ontocircumplex space, one obtains the correlations between thevariable and the circumplex factor scores for dominance andnurturance. The angular location of that variable, theta, isderived by taking the arctangent of the variable’s dominancevalue over its nurturance value. Angular location providessubstantive information regarding which interpersonal traitsare most related to the variable. The distance from thecenter of the circle, labelled r, can be obtained using thePythagorean Theorem. This vector length is an indication ofthe strength of relation between an outside variable and theinterpersonal dimensions of personality. Individual personscan be located similarly by replacing the correlations withself or observer ratings on dominance and nurturance (Wigginset al, 1989). With regard to an individual person, vector17length informs us of the level of rigidity or strength of hisor her central interpersonal tendencies identified by theindividual’s angular location (Kiesler, 1983; Leary, 1957;Wiggins et al, 1989). One potential clinical application ofthe ability to place individuals at a single point in twodimensional interpersonal space is to determine discrepanciesin angular location and vector length between self—ratingsand significant other-ratings of a target. This would beparticularly helpful in understanding some components ofmarital distress.This ability to plot outside variables onto thecircuxnplex can be demonstrated by examining the relationsbetween the circumplex and the first two dimensions of theFFM as assessed by Costa & McCrae’s NEO-PI and the HoganPersonality Inventory (HPI; Hogan, 1986). The NEO-PI, theHPI, and the IAS-R were administered to 581 universitystudents and their results were plotted as explained (Wiggins& Pincus, in press). Examination of Figure 2 clearlydemonstrates that the issue is one of rotation, and notsubstantive differences in the trait domain itself. All thescales plot relatively equidistant from the center of thecircle. If these scales were unrelated to dominance andnurturance, they would not fall in the common space. Thus,Extraversion and Agreeableness (NEO—PI) andAmbition/Sociability and Likability (HPI) are theinterpersonal dimensions of the simple structure form of theFFM. Empirical advantages of the circumplex include the18Figure 2. Projections of the interpersonal dimensions of theHogan Personality Inventory and the NECPersonality Inventory onto the Revised Interpersonal Adjective Scales circumplex.IDOMEO-E19possibility of more detailed analyses of the relationsbetween additional clinical and personality measures and theinterpersonal dimensions of the circumplex. If we hadconjointly factored the scale sets we would have extractedtwo dimensions with a number of split factor loadings. Thisway, we see, for example that the NEO—E scale assesses a“warm dominance” and the NEO—A scale assesses a “submissivefriendliness.” Additionally, precise circumplex measurementallows one to accurately assess an individual’s or diagnosticcohort of patients’ central interpersonal tendencies bylocating the person(s) in a similar way.Additional empirical advantages of the circumplex modelhave been articulated by the work of Gurtman (in press—a; inpress—b). Gurtman has provided additional circuinpiex—baseditem statistics and scale construction methods which cangenerate assessment scales with optimal interpersonalconstruct validity. He has additionally proposedtrigonometric curve—fitting analysis procedures that can beapplied to enhance our understanding of the interpersonalaspects of clinically relevant personality constructs (inpress—c).Theoretical evidence arguing for the fundamental natureof dominance and nurturance has also accumulated across anumber of domains in the social sciences and humanities.Wiggins (1991) has reviewed this evidence in a chapterrecently published in the volumes honoring Paul Meehl. Thesuperordinate concept of agency (including individual20differentiation and strivings for mastery and power), and thesuperordinate concept of communion (including strivings forintimacy, union, and solidarity with a larger entity) arerepresented among the basic tenets of a number ofphilosophies and personality theories, the study of gender,and the study of language. For example, Agency and Communionresemble Alfred Adler’s (1912, 1964) concepts of “strivingfor superiority” and “social interest”; and Karen Homey’s(1937) concepts of “moving against others” and “moving towardothers.” One of the most fascinating findings is that ofBenjafield and Carson (1985), who found that words classifiedas falling at the nodal points of the circumplex (e.g.,dominant, submissive) have significantly earlier dates ofentry into the language than words falling in the four offquadrants (e.g., extraverted, aloof).For practical reasons, clinicians interested inpsychotherapy and psychopathology clearly require accuratedescriptions of interpersonal behavior both for diagnosticpurposes and for effective psychotherapeutic interventions.Interpersonal constructs have been increasingly applied topsychotherapy practice and research (e.g., Anchin & Kiesler,1982; Safran & Segal, 1990). The predictive concepts ofinterpersonal complementarity are often used to evaluatetreatment effectiveness (e.g., Henry, Schacht, & Strupp,1990; Kiesler & Watkins, 1989; Talley, Strupp, & Morey,1990), and to systematize therapeutic interventions(Benjamin, in press; Kiesler, 1988).21Interpersonal complementarity is assessed with referenceto dominance and nurturance. Sullivan (1953a, 1953b)observed that our interpersonal behavior is enacted toconfirm our own self-definitions. Leary (1957) proposed theprinciple of reciprocal interpersonal relations as anexplanation of how interpersonal behavior could fulfill aself—definitional function: “Interpersonal reflexes tend(with probability significantly greater than chance) toinitiate and invite reciprocal interpersonal responses fromthe other person in the interactions that lead to arepetition of the original reflex” (p. 123). An interpersonaltransaction is. complementary if the behaviors of the twoparticipants endorse and confirm each others’ self—definitions relative to the circumplex axes. On thecircumplex, complementarity occurs on the basis ofreciprocity in regard to dominance (dominance pullssubmission; submission pulls dominance) and correspondence inregard to nurturance (hostility pulls hostility; friendlinesspulls friendliness) (Carson, 1969). Since any interpersonalbehavior can be located at a point around the circle, we canidentify a complementary behavior based on reciprocity ofdominance and correspondence on nurturance (Kiesler, 1983).What happens when noncomplementary responses occur?Typically, conflict, anxiety, termination of the ongoingtransaction, or negotiation with respect to dominance andnurturance is likely to ensue (Carson, 1982; Horowitz, Locke,22Morse, Waiker, & Dryer, 1991; Kiesler, 1983, 1988; Sullivan,1953a, l953b; Tally et al, 1990).An additional practical use of the circumplex isdiagnostic description. McLemore and Benjamin (1979)endorsed the rigorous, systematic description ofinterpersonal behavior as uniquely critical for effectivedefinition and treatment of psychopathology. Theinterpersonal nature of psychopathology can be wellarticulated with the circumplex. A number of recentapplications of the circuxaplex model of interpersonalbehavior to diagnostic assessment have demonstrated theutility of the model. Alden and Phillips (1990) used theinterpersonal problems circumplex to differentiate sociallyanxious depressives from pure depressives, suggesting thatsubtypes of depression may involve the presence or absence ofinterpersonal problems. Tunis, Fridhandler, and Horowitz(1990) used the SASB to assess maladaptive schemas in phobicpatients. Johnson, Popp, Schacht, Mellon, and Strupp (1989)also use the SASB as a component of their “CyclicalMaladaptive Pattern” assessment of patients. Humphrey (1989)used the SASB to differentiate interaction patterns amongfamilies of anorexic, bulimic, and patients with mixed eatingdisorders.A number of recent studies have investigated the utilityof the Interpersonal Circle in discriminating and describingDSM-III, Axis II personality disorders. These studies arenoteworthy for the range of methodologies and variety of23assessment instruments that have been brought to bear on thistopic. Strack, Lorr, and Campbell (1990) conducted principalcomponents analyses of the Millon Clinical MultiaxialInventory—Il (MCMI—II, Millon, 1987) and the PersonalityAdjective Check List (PACL; Strack, 1987, 1990) personalitydisorder scales and concluded that the interpersonaldimensions of personality can be identified in instrumentsoperationalizing Millon’s personality theory. Wiggins andPincus (1989) used principal components analyses todemonstrate the relations between the IAS—R and a number ofself—report personality disorder scales. Romney and Bynner(1989) used structural equations to reanalyze a number ofpreviously published sets of correlational data involvingpersonality disorder diagnoses. Kiesler, Van Denburg, Sikes—Nova, Larus, and Goldston (1990) provided interpersonalbehavior profiles for eight personality-disordered patientsbased on ratings of multiple judges who viewed videotapedpsychotherapy sessions. Pincus and Wiggins (l990a, 1990b)demonstrated that a subset of the DSM-III personalitydisorders are systematically related to the interpersonalproblems circumplex. Sim and Romney (1990) usedmultidimensional scaling techniques to investigate therelations between the Interpersonal Circle and personalitydisorders. Their results, and the results of DeJong, van denBrink, Jansen, and Schippers (1989) are less clear, and maybe limited by the structural shortcomings of theInterpersonal Check List (see Paddock & Nowicki, 1986).24Given the general consistency of the other results, however,it appears that interpersonal dysfunction seems central tohistrionic, narcissistic, dependent, avoidant, schizoid, andantisocial personality disorders.Figure 3 presents interpersonal profiles for twoanalogue groups of PDs (Wiggins & Pincus, in press). Theschizoid individuals consistently describe themselves and arerated by others as aloof, introverted, and socially avoidant,thus the schizoid profile has a systematic shape referred toas an interpersonal space ship (Wiggins et al, 1989). Thetarget octant is highly elevated and adjacent octantsconsistently decrease in elevation, ending in a highlytruncated opposing octant of the circle. Given such aprofile, the likelihood of schizoid patients enactingstrongly dominant, gregarious, and nurturing behaviors isquite low.Thus, I argue that for empirical, theoretical, andpractical reasons, the circumplex is a more useful andpromising way of conceptualizing interpersonal behavior thanthe simple structure assessment of extraversion andagreeableness. However, Figure 3 also shows a weakness inthe circumplex model——that is its comprehensiveness. Thesecond profile in this figure is an analogue cohort ofindividuals manifesting a borderline personality disorder.This disorder is characterized by intense oscillations ininterpersonal behavior and marked ambivalence with regard to25Figure 3. Interpersonal profiles of analogue cohortsrepresenting schizoid and borderline personalitydisorders.SCHIZOID PERSONALITY DISORDERAssured-DominantPA90•BORDERLiNE PERSONALITY DiSORDERArro9ant-CaIculangBCGregariousExtravettedNO45.Cold.heartedWarm.Agreeab’e4jf. 270Introverted HIUnassured.SubmissiveUnassuming.IngenuousAssured.DominantPA90ColdheartedWarm-AgreeableAioof- 270Introverted HIUnassured.Submissive31 5JKUnassumingIngenuous26relationships. Hence, the marked differences among thecohort seem to cancel each other out and one cannot identifya single central interpersonal tendency. A rather circularprofile is generated. It is perhaps another personalitydimension that is central to the dysfunction of theborderline personality. Additionally, even if we couldidentify an individual as extremely dominant, hischaracteristic interactions and the impressions he leaveswith others will also be colored by his emotional labilityand distress (or Neuroticism), his impulse control,motivation, and responsibility (or Conscientiousness), andhis conformity, tolerance, and imagination (or Openness).Hence, the three additional dimensions of the FFM arerequired to provide a truly comprehensive description of anindividual personality. Thus, the Dyadic-Interactionalperspective on the FFM combines the advantages of circumplexassessment of interpersonal behavior and traits with thecomprehensiveness of the FFM. Currently, the only assessmentinstrument used to assess the Dyadic—Interactional FFM isTrapnell and Wiggins’ (1990) Extended Interpersonal AdjectiveScales (IASR—B5). This is an adjective-based ratinginstrument which has been little used for clinical researchpurposes.Recent Applications in Clinical Psychology.Returning to the debate introduced at the beginning ofthis review, a small number of studies have applied the FFM27to research domains in clinical psychology. The majority ofinvestigations have used the NEO-PI as the measure of theFFM. Wiggins and Pincus (1989) demonstrated that the FFMcaptures the range of psychopathology found in the DSM-IIIpersonality disorders through the use of canonical analysesand by conjointly factoring the NEO-PI, IASR-B5, PACL, andthe MNPI Personality Disorder scales (Morey, Waugh, &Blashfield, 1985). These results were replicated by Costaand McCrae (1990) and discussed at length by Widiger andTrull (in press). Additional work has been focussed on thearea of psychopathy (Harpur, Hart, & Hare, in press). For analternative approach, see Schroeder, Wormworth, and Livesley(1992).In a special section of the Journal of PersonalityAssessment (Costa, 1991), a number of research effortsapplying the FFM to clinical domains were presented. Usingstandard correlational methods, McCrae (1991) demonstratedsignificant correlational relations between the NEO-PI andthe MMPI and the MCMI—I (Millon, 1983). Miller (1991)presents his attempts at using the NEO-PI to facilitatepsychotherapy treatment in private practice. His paper issubstantially theoretical in nature and proposes: a)Neuroticism influences the intensity and duration of thepatient’s distress, b) Extraversion influences the patient’senthusiasm for treatment, c) Openness influences thepatient’s reactions to the therapist’s interventions, d)Agreeableness influences the patient’s reaction to the person28of the therapist, and e) Conscientiousness influences thepatient’s willingness to do the work of psychotherapy.Fagan, Wise, Schmidt, Ponticas, Marshall, and Costa (1991)compared personality profiles of men with sexual dysfunctionto those of age-matched men with a primary diagnosis ofparaphilia. Analysis of variance showed significantdifferences between dysfunctional and paraphillic groups onneuroticism and agreeableness. The group personality profileof the sexually dysfunctional men was comparable to thenormative sample of the NEO-PI, except for a slight elevationin neuroticism. By contrast, men with paraphilia had apersonality profile marked by high neuroticism, lowagreeableness, and low conscientiousness. The authorsdiscuss treatment implications given differential personalityprofiles for subsets of sexual dysfunction. Muten (1991)describes the use of the FFM as a component of assessmentprocedures used in an outpatient behavioral medicine program.The use of FFM personality information to assist indiagnosis, rapport building, style of treatment delivery,tailoring goals to the individuals’ intrapersonal andinterpersonal dynamics, and predicting relative success andcompliance with noninvasive self—regulation procedures andpsychotherapy is reviewed. Costa and McCrae (1992a)demonstrate significant correlational relations between theNEO-PI and two additional clinical assessment instruments-theBasic Personality Inventory (BPI; Jackson, 1989) and thePersonality Assessment Inventory (PAl; Morey, 1991).29This brief review exhausts the primary literatureapplying the FFM to research domains in clinical psychology.In a special section of Psychological Assessment (Butcher,1992), Costa and McCrae (l992a, 1992b) summarized thisresearch as initial evidence arguing in favor of assessingnormal personality characteristics in clinical settings andproposed that useful clinical information is obtained by suchassessment. Ben—Porath and Wailer (1992a, 1992b) take issuewith these claims and point out possible shortcomings of suchassessment procedures for clinical purposes.Costa and McCrae (1992a, 1992b) suggest that knowledgeof patient’s enduring emotional, interpersonal, experiential,attitudinal, and motivational styles (i.e., personality) canaid clinicians in six specific areas. Information regardingthe basic personality of patients will help the clinicians’understanding of individuals they treat. Costa and McCraevaguely define “understanding” as “a sense of both thepatient’s strengths and weaknesses” (p. 11). Additionally,they suggest that such information will help determine if thepresenting problems are reactive or symptoms of enduringmaladjustments.Knowledge of patients’ personality traits may aid indiagnosis. They suggest “extreme scores on normalpersonality traits are not necessarily an indication ofpsychopathology, but extreme scores are oftencontraindicative of certain diagnoses (e.g., very high30extraversion scores are inconsistent with the diagnosis ofschizoid personality disorder)” (p. 11).Costa and McCrae suggest that personality informationmay help clinician’s establish rapport and empathy withpatients, as well as aid in providing feedback and insight inclient—centered assessment (see McReynolds, 1989). Theysuggest that relating scale scores to concrete examples ofproblematic behavior can be of benefit to the patient andprovide further insight into his/her behavior.Finally, they suggest that personality information willaid in anticipating treatment course and matching treatmentsto patients. They provide hypotheses regarding the responseto treatment types based on patient individual differences onthe dimensions of the FF14. Miller (1991) provides a moreextensive discussion and elaboration of relations betweenindividual differences in personality and patientpresentation and treatment selection, and treatment efficacy.Ben—Porath and Waller’s (l992a, l992b) response appearsto focus on two specific assessment instruments (the NEO—PIand the MMPI-2). In comparing instruments they tend to avoidthe conceptual argument itself. They outline the basic tasksof clinical assessment as assessment of protocol validity,provision of diagnostic information, description of currentlevel of adjustment and stable personality patterns, andarticulating treatment implications. They argue that theMMPI—2 can accomplish such tasks to an acceptable degree.This may be so; however, in so arguing for the use of the31MMPI-2, they inherently endorse the potential benefit of moredirect assessment of personality traits. The strengths oftheir argument are not directed toward the benefit of normalpersonality assessment in clinical psychology, but at the useof the NEO—PI as a stand alone clinical measure (a claimnever made by Costa and McCrae). The strength of theirargument lies in the fact that the MMPI-2 has extensiveextratest correlates of significant clinical utility.Additionally, they correctly point out the fact that Costaand McCrae’s (1992a, l992b), and Miller’s (1991) suggestionsare yet to be tested empirically.Like many debates in psychology, further research is thebest way to advance this field of inquiry. However, if wewere to limit ourselves to a debate between the utility ofthe NEO-PI and the MMPI-2, our frame of inquiry would belimited indeed. It is my contention that a fruitful way toadvance the applications of personality traits to domains ofclinical psychology is to integrate the fields of personalitystructure and psychopathology and psychotherapy. Butcher(1992) asserts, “Clinical personality assessment deals withextremes of behavior, symptoms, attitudes, and actions thatmay seem to many to be a class in themselves, widelyseparated from behaviors in the normal range. In the currentconceptualization of the Big Five dimensions, practitionersmight find it difficult to typify the deviant behavior andpersonality traits that are found in many who are referredfor clinical assessment. Whether the Big Five and the NEO32can be found to provide sufficiently useful information inthe ‘extremes’ of behavior is yet to be determined” (p. 3).Costa and McCrae (l992a) also refer to the need to relateNEO—PI scale scores to concrete examples of dysfunctionalbehavior.The potential for the FFM to provide clinically usefulinformation is partially hampered by it’s long historyembedded in the study of normal personality. Butcher’s(1992) contention that the FFM has not ventured into therealm of problematic trait—behaviors seems true. The DSM—III, for example, defines personality disorders in terms ofpersonality traits which when “inflexible and maladaptive cancause either significant impairment in social or occupationalfunctioning or subjective distress.. .constitute PersonalityDisorders” (APA, 1980, p. 305). One immediate question ishow to operationalize and define inflexible and maladaptivetrait expression.What seems to be a shortcoming of the current FFMassessment instruments is their clinical applicability. Mostclinical assessment instruments focus on either symptoms ordysfunctional behavior; whereas normal personalityinventories have historically avoided particularly evaluativeitems because of the long debated issues of socialdesirability (e.g., Edwards, 1953, 1966, 1967; Jackson &Messick, 1958; Wiggins, 1968, 1973). A clinical instrumentscaled to the FFM is lacking in the repertoire of clinicalresearchers. Assuming that the basic dimensions of33personality are related to abnormal, deviant, rigid, ormaladaptive behaviors, it would be a productive researchprogram that would articulate the maladaptive orproblematical trait domain. I believe such a taxonomy ofmaladaptive traits could be captured by the FFM. Theinterpersonal tradition in personality theory has alwaysconsidered abnormal behavior to be a matter of extreme orrigid manifestation of traits, rather than a qualitativelydifferent domain (Carson, 1969, Leary, 1957, Millon, 1981).Therefore, it may be that combining the interpersonaltradition and the FFM as achieved in the Dyadic-Interactionalperspective will be of some service in the attempt to improveFFM assessment for clinical purposes.Interpersonal Problems and Maladaptive Trait ExpressionThe problems which bring individuals into psychotherapyvary from symptoms (e.g., insomnia), to disturbing views ofthe self (e.g., low self-esteem), to difficulties interactingwith other people (e.g., social avoidance). Horowitz (1979)noted that in therapy outcome research, there was often amismatch between the outcome measure and the focus ofpsychotherapy. Specifically, he noted that psychotherapyoutcome was often evaluated by assessment of symptomaticchange, yet the interventions of psychotherapy often focusedon specific behavioral disabilities of an interpersonal type(e.g., I can’t get along with other people).34In order to improve the accuracy of outcome assessment,Horowitz (1979) proposed that an inventory of interpersonalproblems would provide the investigator with a method toevaluate changes in dysfunctional behavior. Theinterpersonal tradition of personality and psychotherapy haslong recognized the clinical importance of interpersonalproblems and their relations to personality. However, asystematic and comprehensive measure of such concerns waslacking.Through programmatic research, Horowitz and hiscolleagues (Horowitz, 1979; Horowitz & Vitkus, 1986;Horowitz, Rosenberg, Baer, Ureno, & Villasenor, 1988)articulated the construct of interpersonal problems anddeveloped an assessment instrument——the Inventory ofInterpersonal Problems (lIP). To identify problems, intakeinterviews of patients seeking outpatient psychotherapy werevideotaped. Two observers recorded statements made by theseindividuals which began with the phrases “I can’t,” “I haveto,” or acceptable synonyms. A total of 192 problems thatwere agreed upon by both observers as having been expressedin the interviews were generated, but not all of theseproblems were interpersonal (e.g., “I can’t seem to fallasleep at night”). Fourteen judges then rated each problemas interpersonal or not interpersonal. Items identified asinterpersonal by 13 or more judges were retained. Whenredundant statements were removed, a total of 127 statementsreflecting a wide range of interpersonal difficulties35remained. The final items were divided into two types ofinterpersonal problems: behaviors that are hard for theperson to do (behavioral inhibitions), and behaviors theperson does too much (behavioral excesses).Multidimensional scaling procedures identified threemajor dimensions of the lIP domain which Horowitz (1979)labeled: a) “degree of psychological involvement,” b)“nature of involvement” (friendly to hostile), and c)“intention to influence, change, or control other” (dominanceto submissiveness). More recently, factor analyticallyderived scales were developed in a sample of individualsbeginning outpatient psychotherapy who rated lIP items on twooccasions separated by two months (Horowitz et al, 1988). Aprincipal components analysis of the correlations among theitems from the first assessment yielded a first unrotatedfactor with an eigenvalue of 28.8, accounting for 23% of thevariance. Because every item had a positive loading on thisfactor, a general complaint factor was postulated (Horowitzet al, 1988). Principal components analyses conductedindependently on data from each assessment period identifiedsix replicable factors following varimax rotation: Hard tobe Assertive, Hard to be Social, Hard to be Intimate, Hard tobe Submissive, Too Controlling, and Too Responsible.The substantial correlations found among the subscaleswere ascribed to a complaint factor, patient’s generaltendency to report distress. This general tendency is seenas varying across patients, systematically raising or36lowering a person’s score on all subscales. Ipsatizedscores, expressing an individual’s response as a deviationfrom his or her mean response across all items, significantlyreduced intercorrelations among the subscales. Thus, when ageneral complaint factor was removed by ipsatizing scores,the various subscales seemed relatively independent of oneanother (Horowitz et al, 1988, p. 888).The intercorrelations among ipsatized subscales weresubjected to a principal components analysis, which yieldedtwo factors accounting for 73% of the variance. The firstfactor was identified as a dimension ranging from hostilityto friendliness. The second factor was identified as adimension ranging from submissiveness to dominance (Horowitzet al, 1988). Earlier, Horowitz (1979) noted that themultidimensional scaling dimensions of nature of involvementand intention to influence, change, or control other weresimilar to the dimensions of Leary’s (1957) circumplex modelof interpersonal dispositions. Horowitz et al (1988)concluded that the two higher order factors of the lIP“matched the two interpersonal dimensions postulated byinterpersonal theorists like Wiggins (1979) and Kiesler(1983)” (p. 888).It appeared possible to construct scales assessinginterpersonal problems that would be scaled to the structureof the interpersonal circuiuplex in order to fully capitalizeon the conceptual and structural advantages of theinterpersonal circumplex and its associated assessment37methodologies described earlier. To accomplish this, it isnecessary to derive scales in accord with a strong geometricmodel. Alden et al (1990) plotted the 127 items of lIP onthe two interpersonal factors. Three findings were clear: a)The distribution of items was circular, b) The universe ofcontent of interpersonal problems was well represented (therewere no major gaps in item distribution, and c) Some itemswere more strongly interpersonal than others, that is, theyhad longer vector lengths.Alden et al (1990) divided this space into eight sectorsand selected eight items with high vector length values thatwere close to the midpoint of each sector. This resulted ineight 8-item liP circumplex scales (TIP-C). Principalcomponents analysis revealed a robust circumplex structurethat can be seen in Figure 4. A more detailed description ofthe final lIP-C scales can be found in Table 5.Alden et al (1990) investigated the structural relationsbetween interpersonal problems and interpersonal traits asassessed by the IAS-R (Wiggins et al, 1988). These twoinstruments were independently derived, one assessinginterpersonal problems derived from videotaped intakeinterviews and one assessing interpersonal dispositions fromlexical investigations of trait-descriptive adjectives. Aconjoint factor analysis of the two scale sets in a sample of974 subjects demonstrated the domain of interpersonalproblems manifested homologous structure with the domain ofinterpersonal traits.38Figure 4. The interpersonal problems circumplex.Domineering90°PA1.0 I.Vindictive Intrusive135 45.BC .5 NOCold Overly.180 i Nurturant-.5 .5 • 1.0 o°DE -1.0 LMFG..225° JKSocially I 315°Avoidant-1.0 - ExploitableHI2700Nonassertive39Circumplex Scale.Domineering. (PA) High scorersreport problems related tocontrolling, manipulating,expressing aggression towardand trying to change others.Sample Items“I am too aggressivetoward other people”“I try to control otherpeople too much”Vindictive. (BC) High scorersreport problems related todistrust and suspicion ofothers, and an inability tocare about other’s needs andhappiness.“It is hard for me totrust other people”“I want to get revengeagainst other peopletoo much”Cold. (DE) High scorers reportan inability to expressaffection toward and feel lovefor another person, difficultymaking long—term commitmentsto others, and an inability tobe generous to, get along with,and forgive others.“It is hard for me tofeel close to otherpeople”“It is hard for me toget along with otherpeople”Table 5. Inventory of Interpersonal Problems CircumplexScales40Table 5 continuedSocially Avoidant. (EG) Highscorers feel anxious andembarrassed in the presenceof others, and have difficultyinitiating social interactions,expressing feelings, and socializing with others.“It is hard for me toask other people toget togethersocially with me”“I feel embarrassed infront of others toomuch”Nonassertive. (HI). High scorersreport difficulty making theirneeds known to others, discomfort in authoritative roles,and an inability to be firmand assertive toward others.“It is hard for me totell a person tostop bothering me”“It is hard for me to beassertive withanother person”Exploitable. (JK) High scorersfind it difficult to feelanger and to express angerfor fear of offending others.They describe themselves asgullible and readily takenadvantage of by others.“I am too easily persuaded by otherpeople”“I let other people takeadvantage of me toomuch”41Table 5 continuedOverly-Nurturant. (LM) High scorers “I try to please otherreport that they try too hard to people too much”please others and are too “I put other people’sgenerous, trusting, caring and needs before my ownpermissive in dealing with too much”others.Intrusive. (NO) High scorers are “I want to be noticedinappropriately seif—disciosive, too much”attention-seeking, and find it “It is hard for me todifficult to spend time alone, stay out of otherpeople’s business”Maladaptive Trait Expression. Butcher (1992) expressedconcern regarding the ability of the FFM to incorporateextreme behaviors. Costa and McCrae (1992b) suggested thatNEO—PI scale scores needed to be related to specificdysfunctional behaviors in patients. In a studyinvestigating interpersonal problems associated withpersonality disorders (defined in the DSM—III as inflexibleand maladaptive expressions of personality traits), Pincusand Wiggins (1990) proposed that two ways in which a traitcould be manifested in an inflexible and maladaptive mannerare: in behaviors one “does too much” and in behaviors oneconsistently finds “hard to do.” Pincus and Wiggins (1990)42also noted that such behavioral excesses and behavioralinhibitions (problems) can currently be assessed forinterpersonal traits through the use of the lIP-C; and, theapplication of the problems construct to additionalpersonality traits was also possible. Thus, one way toaddress the concerns of Butcher and Ben—Porath and Wailer isto operationalize inflexible and maladaptive trait expressionthrough the problems construct of chronic behavioral excessesand chronic behavioral inhibitions.A Brief Review of Research on Interpersonal Problems.If the application of normal personality assessment toclinical domains can be improved by the adoption of theproblems construct, some empirical evidence of itsapplicability and advantages is required. The liP has provenuseful in predicting problems that would become the focus ofpsychotherapy, even when judge’s liP ratings were based onindependently formulated case summaries (Horowitz, Rosenberg,Ureno, Kahlehzan, & O’Halloran, 1989). This report alsoprovided important evidence indicating that patients withprimarily interpersonal (as opposed to symptomatic) problemsare better candidates for brief dynamic therapy. A number ofstudies have used the lIP and the lIP-C to predict positiveand negative treatment response and demonstrated theinstrument’s sensitivity to patient change (Horowitz et al,1988, 1989; Horowitz, 1991; Mohr, Beutler, Engle, ShohanSolomon, Bergan, Kaszniak, & Yost, 1990). Mohr et al (1990)concluded that interpersonal problems and distress activate43the individual to seek change in psychotherapy. Alden andPhillips (1990) used the liP—C to differentiate sociallyanxious depressives from pure depressives, suggesting thatsubtypes of depression may involve the presence or absence ofinterpersonal problems. Alden (1991) and Alden and Capreol(1992) demonstrated that the lip—C was useful indiscriminating subtypes of avoidant personality disorder thatresponded differentially to two treatment approaches. Pincusand Wiggins (l990a, 1990b) have shown that a subset ofpersonality disorders are systematically related to theinterpersonal problems circumplex. Finally, Gurtman,Fernandez, and Phillips (1991) have investigatedinterpersonal problems related to narcissism; and, Gurtman(1992) has done the same for trust and distrust.Combining Personality Structure and Clinical ResearchThe interpersonal problems circumplex is a model ofmaladaptive interpersonal trait-behaviors operationalized bythe assessment of chronic behavioral inhibitions and chronicbehavioral excesses. It conforms to a model of personalitytrait structure which is now recognized as a component of theFFM of personality trait organization. When theinterpersonal circumplex is used to operationalize the firsttwo dimensions of the FFM, a number of empirical,theoretical, and practical advantages accrue. This isreferred to as the Dyadic—Interactional perspective on theFFN (Pincus & Wiggins, in press; Trapnell & Wiggins, 1990;44Wiggins & Pincus, 1992, in press). The present researchprogram builds on the development of the lIP-C (Alden et al,1990) by expanding the problems construct to operationalizemaladaptive and inflexible expression of the remaining threedimensions of the FFM: neuroticism, conscientiousness andopenness to experience. The goal is to derive an assessmentinstrument of significant clinical utility that is scaled tothe FFM of personality trait structure. A FFM of maladaptivetrait—behaviors (problems) may succeed in mapping a taxonomyof personality traits of significance to the understanding ofpsychopathology, diagnosis, and psychotherapy.45CHAPTER 2Study 1: Scale DerivationMethodsOverview.The major purpose of Study 1 was to derive initialproblems scales for the domains of neuroticism,conscientiousness, and openness. A combinedrational/empirical scale construction strategy was adopted.Such an approach combines the “internal” and the “intuitive”approaches to scale construction outlined by Hase andGoldberg (1967). These two approaches have also beenreferred to as “inductive” and “deductive” strategies(Burisch, 1986). A rational (intuitive/deductive) approachinvolves choosing the concepts to be measured first, and thenwriting items to fit implicit or explicit definitions of thechosen concepts. An empirical (internal/inductive) approachinvolves amassing a large heterogeneous item pool, often fromexisting inventories, and applying multivariate datareduction analyses to identify groups of correlated items.The approach to constructing additional scales for the liPcombines these methods by rationally generating items to fitthe chosen constructs of neuroticisiu, conscientiousness, andopenness at the superordinate level of the constructhierarchy (Comrey, 1988), while identifying substantivesubscales at a lower level of the construct hierarchy throughfactor analytic techniques.46Four psychologists (Dr. Jerry S. Wiggins, Dr. KimBartholomew, Aaron L. Pincus, and Paul D. Trapnell) familiarwith both the FFM and interpersonal problems generated aninitial pooi of 236 items which, a priori, spanned thedomains of neuroticism, conscientiousness, and openness. Theitem content was based on familiarity with a number of FFMassessment questionnaires, such as the NEO—PI (Costa &McCrae, 1985) and the IASR—B5 (Trapnell & Wiggins, 1990).Items were worded in the same form as the original lIP; thatis, behaviors “you do too much” and behaviors “that are hardfor you to do.”It was also considered that problems ofconscientiousness and openness could be bipolar in nature,hence problems were generated that indicated both high andlow ranges of these two dimensions (e.g., overly—conscientious and unconscientious problems; and problems ofbeing too open and too closed). The problems related toneuroticism were considered unipolar in nature and items weregenerated to capture high levels of neuroticism. Allproblems items were assembled in a test booklet and theformat was identical to that of the original lIP. Subjectswere asked to rate how much of a problem each item has beenfor them on a Likert scale ranging from “not at all” (0) to“extremely” (5).An additional empirical method based on Loevinger’s(1957) concept of structural validity was used to clarify thestructure of the derived scales. Structural validity refers47to the extent to which structural relations between testitems parallel the structural relations of othermanifestations of the trait being measured. In the presentcase the convergence of a number of theoretical perspectiveson trait ontology and the results of numerous investigationsof the validity, stability, and predictive utility of thefive—factor and interpersonal circumplex models cited in theintroduction, provide information regarding both test andnontest structural relations of neuroticism,conscientiousness, and openness.Loevinger’s (1957) concept of “structural fidelity”proposes that items be selected from a large pool on thebasis of empirical properties that indicate which items bestconform to an appropriate structural model. Wiggins (1973)proposes: “One should not seek to discover structuralrelationships among test items. The choice of measurementmodel should be determined by the structural relationshipsposited by a theory of nontest behaviors. Once again, weemphasize the importance of developing theories ofpersonality trait organization that are sufficientlyarticulated to permit the selection of appropriatemeasurement models” (p. 405).Therefore, a structure was imposed on the item pool byincluding each of the twelve items marking the dimensions ofneuroticism, conscientiousness, and openness from the Five—Factor Inventory, a short form of the NEO-PI (FF1; Costa &McCrae, 1989). These items were modified to problem item48format and included in the initial item pool as structuralmarkers. These modified items were removed from final scaleitem selection. The importance of including appropriatestructural markers when investigating the superordinatestructure of questionnaire scales is increasingly recognized(Goldberg, 1990; Ostendorf, 1990). The use of structuralmarkers for scale construction is one way of empiricallyoperationalizing Loevinger’s and Wiggins’ recommendations(see also Gurtman, 1991).Sample. The entire 272 item pool, along with a varietyof FFM inventories and additional clinical, adjustment, andpersonality questionnaires was administered to a sample of706 university students. This sample consisted of 393 women(56%) with an average age of 19.7 years, and 313 men (44%)with an average age of 20.2 years. The sample was used toderive the initial version of the Inventory of InterpersonalProblems—Big Five version (IIP-B5).Instruments. In addition to the experimental IIP-B5item pool, a number of self—report questionnaires wereadministered in a large assessment battery. Some of thequestionnaires included in the battery were used forindependent research projects. Only those pertaining to thederivation of the IIP-B5 will be discussed.The Inventory of Interpersonal Problems CircumplexScales (lIP—C; Alden et al, 1990), consist of eight 8—itemscales that form a circumplex of interpersonal problemsaround the dimensions of dominance and nurturance. These49scales were initially developed independently from any othercircumplex instrument. The lIP—C scales have shown strongconvergence between self— and peer—rating profiles(Bartholomew & Horowitz, 1991); ability to discriminatesubgroups of depressed patients (Alden & Phillips, 1990);homologous structure to the IAS-R (Alden et al, 1990), and anability to discriminate important features of diverse self-report scales measuring personality disorders (Pincus &Wiggins, 1990), trust and distrust (Gurtman, in press-b) andnarcissism (Gurtman, Fernandez, & Phillips, 1991). Alphacoefficients range from .72 for the Intrusive scale to .85for the Nonassertive and Socially Avoidant scales.The Extended Interpersonal Adjective Scales (IASR-B5;Trapnell & Wiggins, 1990) consists of a) eight 8—itemadjectival scales that form a circumplex around thedimensions of dominance and nurturance (IAS—R; Wiggins et al,1988), and b) three 20-item adjectival scales measuring thedimensions of neuroticism, conscientiousness, and openness.The instrument has excellent structural properties at theitem level, and promising convergent and discriminantproperties when compared with the NEO-PI and the HPI. Alphacoefficients range from .87 for the Openness scale to .94 forthe composite Dominance and Nurturance Scales.The NEO-Five Factor Inventory (FF1; Costa & McCrae,1989) consists of five 12—item scales assessing extraversion,agreeableness, conscientiousness, neuroticism, and openness.For each domain, the 12 NEO-PI items with the highest factor50loadings were selected. Ten item substitutions were made todiversify item content. FF1 scales show excellent convergentvalidity with the full NEO-PI. Alpha coefficients range from.74 for the Agreeableness scale to .89 for the Neuroticismscale.The revised Shyness scale (Cheek & Melchior, 1985) is a20—item scale which assesses individual’s discomfort andinhibition in the presence of others. This scale has beenused widely for research purposes (e.g., Arnold & Cheek,1986; Briggs, 1988; Cheek & Stahl, 1986). The alphacoefficient for this scale is .94.The Trait Anxiety scale from the State-Trait AnxietyInventory (STAI; Spielberger, 1983) is a 20—item scaleassessing individuals general level of anxiousness. Itemsare face valid and assess a variety of physical, cognitive,and behavioral manifestations of anxiety. The reported alphacoefficient for this scale has ranged from .86 to .92 invarious versions of the STAI manual.The Interpersonal Dependency Inventory (IDI; Hirschfeld,Klerman, Gough, Barret, Korchin, & Chodoff, 1977) is a 48-item questionnaire that assesses three components ofinterpersonal dependency: Emotional Reliance on Others, Lackof Social Self-Confidence, and Defensive Autonomy. The threesubscales demonstrate adequate validity and the split-halfreliabilities range from .72 to .87. Including DefensiveAutonomy in scoring subjects’ dependency ratings is a matterof some debate, and some investigators (e.g., Birtchnell,511991) argue against including the third subscale whenassessing dependency.Procedures. Subjects were recruited from the Departmentof Psychology subject pool at the University of BritishColumbia and were given course credit for theirparticipation. Subjects completed the assessment battery intwo 1—hour sessions. In all testing sessions, subjectscompleted the IASR-B5 first. The order of all otherinventories including the IIP-B5 item pooi was varied.AnalysesIpsatization. The derivation of IIP-B5 scales forneuroticism, conscientiousness, and openness proceeded in astep-wise fashion, beginning with data transformation. Intheir principal components analyses, Horowitz et al (1988)found a large general factor in the original lIP item pool.A general factor accounts for a large proportion of itemvariance and all items have a positive loading on the factor.This general factor led them to conclude that subjects havedifferential tendencies to endorse complaints. Whether ornot this factor is viewed as a “complaint factor,” or an“acquiescence factor,” or an “intensity factor,” such acomponent is thought to reflect individual differences in theuse of response format, rather than (substantive) differencesin the perception of self or others (Wiggins, Steiger, andGaelick, 1981, p. 283). Therefore, it must be treatedseparately from the substantive components in interpreting agiven solution.52One way to control for this individual difference is toipsatize the individual item scores by expressing each scoreas a deviation from the subject’s mean score across all items(Alden et al, 1990; Cronbach, 1949; Horowitz et al, 1988;Strack, 1987; Wiggins & Pincus, 1989). One particularlyambiguous issue in this procedure is determining thesubstantive effects of ipsatization on the content of thedata. Interpretation of the effects would be facilitated bya determination of the content domain of the general factorand the effects of ipsatization on derived scales. In Study1, the first unrotated general factor was interpreted byexamining its external correlates.Data Reduction. Ipsatization produces anintercorrelation matrix which is not of full rank. Althoughthere are a number of statistical procedures which allow oneto utilize common factor analysis on such matrices, the mostparsimonious dimension reduction technique is principalcomponents analyses, as such analyses do not require thecalculation of a matrix inverse. Therefore, scales werederived through the use of principal components analysis. Inthe current study, the likelihood of significantly differentresults occurring from the use of different factor analytictechniques is small. In their comparison of factor andcomponent analyses, Bentler and Kano (1990) concluded that ifthe number of factors (m) stays small as the number ofvariables (p) gets large, there is no need to choose betweencomponents and factors because the component and maximum53likelihood factor patterns, and corresponding scores, becomeequivalent. When one is concerned solely with datasummarization, especially dimension reduction, principalcomponents analysis may be fruitfully employed (p. 73). Inthe present case the number of items (p) is large, but thenumber of derived factors (m) was relatively small.To develop superordinate higher order problems factorsfor neuroticism, conscientiousness, and openness, severallower order scales should be developed in each major factorarea (Comrey, 1988). To obtain the broader, more generalpersonality constructs, the intercorrelations among severallower order factor scales must be subjected to furtherprincipal components analyses. Each of the item setsrepresenting the three superordinate domains were firstanalyzed separately to develop subscales, or facets ofhomogeneous items. Costa and McCrae (1985) refer to suchlower order constructs as facets, whereas Hogan (1986) refersto “homogeneous item clusters” or HIC’s, and Comrey (1988)refers to “factored homogeneous item dimensions” or FHID’s.After removal of modified FF1 items, these derived facetswere reanalyzed to ensure structural fidelity.Following their derivation, the entire set of problemfacets was subjected to a conjoint principal componentsanalysis with markers of neuroticism, conscientiousness, andopenness from the IASR—B5. The use of adjectival scalesallows for appropriate structural markers while reducing thecommon method variance in the analyses (i.e., adjective vs.54questionnaire scales). The derived facet scales should be ata level in the construct hierarchy that is similar to thelIP—C octant scales. The superordinate problems factors forthe three domains should be at the equivalent level as thecircumplex dimensions of dominance and nurturance. Conjointprincipal components of all derived problems facets and liP-Coctant scales was performed. Conjoint principal componentsanalyses of the superordinate dimensions of the derivedproblems domains and the interpersonal problems circumplexaxes followed.Reliability and Validity. Alpha coefficients and item—total correlations for derived facets and domain scales weredetermined to assess internal consistency. Itemmodifications may be introduced based on such results.Convergent validity correlations for the three new problemsdimensions were obtained by examining the relations betweenderived problems scales and the neuroticism,conscientiousness, and openness scales of the IASR—B5 and theFF1 concurrently assessed in the sample. Additional validitycorrelations were obtained by examining the relations betweenthe derived scales and the IDI, STAI Trait Anxiety scale, andthe Shyness scale.Expectations1. Chronic behavioral excesses and chronic behavioralinhibitions (problems) can by systematically related to theremaining three dimensions of the FFM.552. The general factor running through the original liPwill replicate across the enlarged item pool.3. A number of studies have shown that traitneuroticism is associated with both recall of, and distressedreactivity to physical symptoms and daily stressors (Affleck,Tennen, Urrows, & Higgins, 1992; Bolger & Schilling, 1991;Costa & McCrae, 1987; Larsen & Kasimatis, 1991; Larsen, inpress; Watson & Pennebaker, 1989). Therefore, it ishypothesized that the general factor found in problemsendorsement will be related to trait neuroticism.4. Additional problems facets can be derived to expandthe problems construct from the interpersonal domain to thebroadened domain of the FFM of personality and a valid andreliable inventory to assess such problems can beconstructed.5. The newly constructed inventory will demonstratestructural fidelity with the Dyadic-Interactional perspectiveon the FFM (Trapnell & Wiggins, 1990; Wiggins & Pincus, 1992,in press).ResultsIpsatization.The entire 272 experimental IIP-B5 item pool wassubjected to a principal components analysis prior toipsatization procedures to determine if an unrotated generalfactor could be extracted; and if so, correlationalrelationships between this general factor and majorpersonality dimensions could be examined. The results of56this principal components analysis indicated that a largeunrotated general factor, with an eigenvalue of 47.6,accounted for 20.0% of the item variance. All but six items(97.5%) had positive loadings on this first unrotated factor.The second factor extracted had an eigenvalue of 11.5 andaccounted for 4.8% of the variance. Notably, the pattern ofitem loadings on this factor were of mixed valence.The general factor was then correlated with the FFMscales of the IASR-B5, the NEO-FFI, and mean item endorsement(elevation) across the item pool. These correlations arepresented in Table 6. The statistical power of the largesample size allows for correlations as low as .09 to besignificant at p < .01. Therefore, the relative magnitudesof the correlations are of most importance. Mean endorsementacross the item pool correlates .98 with the first unrotatedfactor. Thus this factor clearly represents the elevation ofthe problems items. The general factor correlates .65 withthe IASR-B5 Neuroticism scale and .76 with the NEO-FFINeuroticism scale. The average correlation between thegeneral factor and additional dimensions of the FFM is .28.When the item pool is ipsatized, the general factor isremoved. If this general factor is interpreted assubstantially trait neuroticism, rotation of the factorthrough derived component scales would create increased scaleinterdependencies similar to those seen in Horowitz et al(1988). While Horowitz et al (1988) and Alden et al (1990)57Table 6. Intercorrelations between the general factor of theInventory of Interpersonal Problems—Big Five version and thedomain scales of the NEC-Five Factor Inventory, the ExtendedInterpersonal Adjective Scales, and mean endorsement acrossthe item pool.General FactorMean Endorsement .98IASR-B5 ScalesDominance —.28Nurturance —.27Neuroticism .65Conscientiousness —. 29Openness — .14NEO—FFI ScalesExtraversion —. 30Agreeableness -.38Neuroticism .76Conscientiousness — .41Openness — .18Note. N = 706. All correlations p < .01.58were concerned with a two—factor interpersonal model, thecurrent study involves a five—factor model. One of theadditional three factors in the model (neuroticism) is, infact, assessed within the item pool. Since ipsatizationsubstantively removes neuroticism from item content, theprocedure was only applied to the circumplex, openness, andconscientiousness domains during scale derivation procedures.Derivation of Problems Scales for Neuroticism,Conscientiousness, and Openness.The problems items generated for neuroticism,conscientiousness, and openness were first subjected tointradomain principal components analyses with obliquerotations to develop facet scales. Oblique rotations wereextracted because facets for a superordinate domain areassumed to be substantively related, but not redundant. AsBen-Porath and Waller (l992a) point out: “The greater thepsychometric independence among facets within domains (i.e.,the greater the discriminant validity of the facets), theless likely it is that the factorial integrity of each domainwill be maintained” (p. 17). By performing oblique rotationsand reviewing component intercorrelations, it is possible toidentify nonredundant, but interrelated facets that will aidin maintaining domain integrity. The number of factorsextracted was determined by two criteria: a) components witheigenvalues greater than one, and b) examination of the screeplot. Items were selected based on relative loadings acrosscomponents and substantive homogeneity. Items with component59loadings greater than .40 on the target component and lessthan .30 on additional components were examined. Itemsmeeting these criteria were then examined for substantivehomogeneity and final selection was made with a goal ofretaining a minimum of five items per facet to a maximum often items per facet. Modified FF1 items were discarded atthis point.Neuroticism Items. All 52 unipsatized a priorineuroticism items were subjected to a principal componentsanalysis with an oblique rotation. Extraction criteriasuggested either a five or six component solution was best.Examination of the sixth component determined that too fewitems met selection criteria and a homogeneous theme couldnot be determined. Therefore, a five component solutionaccounting for 31.6% of the item variance was extracted. Theitem loadings can be examined in Table 7. The derivedneurotic problems facets can be seen in Table 8.The five facets of neurotic problems retained wereidentified as Depression (sad affect), Anger, Anxiety, LowSelf—esteem, and Urge Control. The average correlationbetween oblique components is .16 (see Table 9).Conscientiousness Items. All 67 ipsatized a prioriconscientiousness items were subjected to a principalcomponents analysis with an oblique rotation. Extractioncriteria suggested a six component solution. Examination ofthe sixth component determined it was a doublet, hence toofew items met selection criteria and a five component60Table 7. Principal components of neurotic problems items.FactorsA 2ItemX2l5 .80X107 .77X229 .66X224 .64X157 .40X1l6 .80X54 .76X154 .59Xl53 .47Xl43 .40X69 .72X88 .65X62 .63X78 .62X84 .54X41 .71X49 .65X186 .57X52 .40X45 .40X194 .66X148 .62X167 .4861Table 7 ContinuedX35 .46Xll0 —.40Note. N = 706. Loadings < .30 deleted.Table 8. Derived Neuroticism FacetsFactor 1: Depression215. I feel sad too much.107. I feel depressed too much.229. I feel empty inside too much.224. I feel discouraged too much.157. Hard to feel cheerful.Factor 2: Anger116. I lose my temper too much.54. Hard to control my temper.154. Hard to keep from exploding at someone when they’vereally angered me.153. Hard to keep calm when things don’t go my way.143. Hard to keep from smashing something when I’m in a rage.Factor 3: Anxiety69. I feel panicky too much.88. I worry about what bad things might happen to me toomuch.62. I worry about things going wrong too much.78. I feel tense and jittery too much.62Table 8 continued84. I am too fearful.Factor 4: Low Self-Esteem41. Hard for me to like myself.49. Hard for me to have faith in myself.186. I feel I am inferior too much.52. Hard for me to feel ambitious.45. Hard for me to feel I deserve affection from others.Factor 5: Urge Control194. I do things I know aren’t good for me too much.148. Hard for me to stop doing things I know are hurting me.167. Hard for me to control my urges.35. Hard for me to control my bad habits.110. I think about my health too much.Table 9. Neuroticism Component Intercorrelations.1. 2. 3. 4. 5.1. Depression2. Anger .293. Anxiety .29 .144. Low Self—Esteem .36 .05 .185. Urge Control .10 .04 .11 .03Note. N = 706.63Table 10. Principal Components of conscientiousness problemsitems.FactorsaItemXl65 .70X85 .69Xl05 .67X8 .62X140 .61X225 .70X203 .69X226 .66X66 .60X104 .57X74 .57X217 .55X109 .49X112 .44X208 .43Xl59 .38X121 .33X199 .31X65 .61X36 .58X99 .55X61 .50X118 .4664Table 10 continuedX191 .75X234 .66X222 .58X190 .56X201 .55Note. N = 706. Loadings < .30 deleted.Table 11. Derived Conscientiousness Facets.Factor 1: Laziness165. Hard for me to concentrate on work for really longperiods of time.85. I waste too much time before settling down to work.105. I feel lazy towards my work too much.8. Hard to work systematically toward goals.140. Hard for me to work hard at something that I’m notparticularly interested in.Factor 2: Recklessness225. I crave excitement too much.203. I’m attracted to dangerous or adventurous activities toomuch.226. I get “out of control” too much.66. I do impulsive, “crazy” things too much.104. I act recklessly too much.65Table 11 continuedFactor 3: Compulsivity74. I constantly strive for perfection too much.217. I fret over keeping things well-organized too much.109. I make lists too much.112. I’m concerned about keeping my appearance neat too much.208. I schedule my time too much.159. Hard to be relaxed in the middle of disorder oruncleanliness.121. I’m concerned about others being on time too much.199. I watch the clock too much.Factor 4: Unreliability65. I’m late for appointments too much.36. Hard for me to meet my obligations.99. I break commitments too much.61. I forget to pay bills on time too often.118. I forget to return phone calls too much.Factor 5: Impulsivity191. I start something too quickly without thinking itthrough first.234. I act without thinking of the consequences too much.222. I overlook important details too much.190. I blurt things out that would be better left unsaid toomuch.201. I react impulsively too much.66solution accounting for 31.2% of the item variance wasextracted. The item loadings can be examined in Table 10.The derived conscientiousness problems facets can be seen inTable 11.The five facets of conscientiousness problems retainedwere identified as Laziness, Recklessness, Compulsivity,Unreliability, and Impulsivity. The average correlationbetween facets extracted obliquely is .12 (see Table 12).Table 12. Conscientiousness component intercorrelations.1. 2. 3. 4. 5.1. Laziness2. Recklessness .053. Compulsivity -.11 -.054. Unreliability .16 .18 —.155. Impulsivity .24 .11 —.01 .14Note. N = 706.Openness Items. All 63 ipsatized a priori opennessitems were subjected to a principal components analysis withan oblique rotation. Extraction criteria suggested a sixcomponent solution. Examination of the sixth componentdetermined that too few items met selection criteria and afive component solution accounting for 27.8% of the itemvariance was extracted. The item loadings can be examined in67Table 13. Principal components of openness problems itemsFactorsItemX133 .68Xl49 .66X134 .62X51 .44X124 — .50Xl31 —.43Xl64 —.43X97 —.43Xl80 .37X2l4 .57X202 .55X235 .53X238 .52X77 .49X56 .55Xll9 .54X48 —.50X59 . 49Xl20 —.47Xl63 .62X128 .54Xl76 .51X209 .4568Table 13 continuedX155 .43Note. N = 706. Loadings < .30 deleted._________Derived Openness FacetsUnimaginativeto feel captivated when looking at a work of art.to play with theories or abstract ideas.to be creative.to imagine things just by thinking about them.Nonconformityto break the law, no matter what the circumstance.to break with traditions.to question authority.too conservative.to adjust to activities that are routine andpredictable.Factor 3: Lack of FocusI am too openminded.I am interested in too many things.I am attracted to variety too much.I am too eclectic in my thinking.I am too curious.Table 14.Factor 1:133. Hard149. Hard134. Hard51. HardFactor 2:124. flard131. Hard164. Hard97. I am180. Hard214. 14 continuedFactor 4: Oversensitivity56. Hard to resist crying when other people are crying.119. I am too affected by sad or frightening movies.48. Hard to experience intense emotions.59. I get caught up in the feelings of the moment too much.120. I ignore my emotions too much.Factor 5: Egocentrism163. Hard to accept new and different ideas.128. Hard to appreciate other cultures.176. Hard to understand people who are unusual.209. I am too rigid about the way I do things.155. Hard to be tolerant of differing religious beliefs.Table 15. Openness component interrcorrelations1. 2. 3. 4. 5.1. Unimaginative2. Oversensitive -.093. Lack of Focus -.19 .024. Egocentrism -.12 .09 .065. Nonconformity -.03 .03 .08 .10Note. N 706.70Table 13. The derived openness problems facets can be seenin Table 14.The five facets of openness problems retained wereidentified as Unimaginative, Nonconformity, Lack of Focus,Oversensitivity, and Egocentrism. The average correlationbetween facets extracted obliquely is .08 (see Table 15).Principal Components of Selected Items. The selecteditems making up each of the five facets of neuroticism,conscientiousness, and openness generated a reduced itempool: 25 neurotic problems items, 28 conscientious problemsitems, and 24 openness problems items. The reduced, domain—specific problems items were then subjected to a second roundof principal components analyses to verify facet structurewith nonselected items removed. All analyses extracted fivefacets per domain using oblique rotations.Principal components of the 25 selected neuroticproblems items generated a five component solution accountingfor 47.3% of the item variance. As can be seen in Table 16,all but one selected item loaded on the expected facet. Theitem “I feel inferior too much” (X186) previously loaded onthe Low Self-esteem facet. In the second analysis, it splitits loading between Depression (.51) and Low Self—esteem(.47). A decision was made to retain item 186 in the LowSelf—esteem facet scale pending further analyses. Theaverage facet intercorrelation was .09 (see Table 17).Principal components of the 28 selected conscientiousproblems items generated a five component solution accounting71Table 16. Principal components of selected neuroticism items.Factors1ItemX215 .81X107 .74X229 .74X224 .70Xl57 .50Xl16 .78X54 .76X154 .74X153 .56X143 .49X69 .77X88 .72X78 .68X62 .65X84 .62X49 .68X41 .65X52 .56X45 .56X186 .51 .47X194 .72X148 .65X167 .5672Table 16 ContinuedX35 .51xiioa—.24Note. N = 706. aAll other loadings < .3 deleted.Table 17. Neuroticism component intercorrelations II.1. 2. 3. 4. 5.1. Depression2. Anger .063. Anxiety .21 .004. Low Self—Esteem .13 —.08 .075. Urge Control .13 .09 -.04 .10Note. N = 706.<<<<<><X><<<<><><><><><Ht3WHa.Dat’Jt.JHHHaHOHODHçtHaHH(iio.Dw‘.Doo’cauioct\)HC).HU01C0101H CD H•••••I•’-d0101HctOOD.‘D0cnC) I-,. Ha--IJC)HiOUQ 0 CD r1lUITjC11p) C)0r1 0 hICflCflCD H•••••CD0101010IC)(71—3Ct CD C) 0 U)101C) CD Ct 0 U) CD U) U)74Table 18 continuedX208 .61X74 .59X112 .52X109 .50X159 .43X199 .40Xl2la—.38 .27Note. N = 706. aAll other loadings < .3 deleted.for 41.3% of the item variance. As can be seen in Table 18,all but one selected item loaded on the expected facet. Theitem “I’m concerned about others being on time too much”(X12l) previously loaded on the Compulsivity facet. In theTable 19. Conscientiousness component intercorrelations II.1. 2. 3. 4. 5.1. Laziness2. Recklessness .013. Impulsivity .14 .144. Unreliability .07 .11 .075. Compulsivity —.19 —.02 —.07 —.19Note. N 706.75second analysis, it split its rather low loading betweenUnreliability (-.38) and Compulsivity (.27). A decision wasmade to retain item 121 in the Compulsivity facet scalepending further analyses. The average facet intercorrelationwas .10 (see Table 19).Principal components of the 24 selected opennessproblems items generated a five component solution accountingfor 47.1% of the item variance. As can be seen in Table 20,all but two selected items loaded on their expected facets.The items “I ignore my emotions too much” (X120) and “Hard toexperience intense emotions” (X48) previously loaded on theOversensitivity facet. In the second analysis, item 120split its loading between Nonconformity (.64) andOversensitivity (—.38). Item 48 also split its loadingbetween Nonconformity (.61) and Oversensivity (-.45). Adecision was made to retain item 120 and item 48 in theOversensitivity facet scale pending further analyses. Theaverage facet intercorrelation was .07 (see Table 21).Superordinate Structure of Derived Problems Facets. Inorder to evaluate the superordinate structure of the derivedproblems facets, all fifteen facet scales were conjointlyfactored with the appropriate domain scales of the IASR-B5.Oblique and simple structure extractions were nearlyidentical and only the varimax rotations will be presented.Extraction criteria indicated a three component solutionaccounting for 51% of the variance was indicated (see Table22).76Table 20. Principal components of selected openness items.Factors1ItemX97 .79X164 .78X131 .73X124 .55Xl80 .32X134 .77X149 .74Xl33 .65X51 .62X235 .73X202 .71X214 .67X77 .54X238 .53X56 .68X119 .64X59 .59X48 .64 .45X120 .61 .38X128 .71X155 .62X176 .6277Table 20 continuedX163 .60X209 .40Note. N = 706. All loadings < .3 deleted.Table 21. Openness component intercorrelations II.1. 2. 3. 4. 5.1. Nonconformity2. Unimaginative .073. Lack of Focus .03 —.154. Oversensitivity .06 —.02 .065. Egocentrism —.04 .13 —.14 -.04Note. N = 706.78Table 22. Three component solution of conjoint principalcomponents analysis of derived problem facets anddomain scales of the IASR-B5.FactorI U 111ScaleNeuroticism (IASR-B5) .81Anxiety .75Depression .71Nonconformity -.70Low Self-esteem .54Oversensitivity .35Anger .34Conscientious (IASR-B5) .81Compulsive .79Lazy —.63Unreliable —.51Impulsive -.39Urge Control -.35Openness (IASR-B5) .75unimaginativeLack of Focus .56Egocentrism .34 -.53Reckless —.36 .48Note. N = 706. Loadings < .30 deleted.79The conjoint principal components analysis of thederived problems facets demonstrated that the superordinatestructure of the facets could be captured by the dimensionsof neuroticism, openness, and conscientiousness. However, anumber of facets loaded on unexpected dimensions. BothNonconformity and Oversensitivity (derived from the a prioriopenness item pool) loaded on the neuroticism dimension,while Reckless (derived from the a priori conscientiousnessitem pool) split its loading between neuroticism andopenness. Urge Control (derived from the a priorineuroticism item pool) loaded on the conscientiousnessdimension.In order to determine the most accurate superordinatestructure, a number of secondary analyses were performed.From the item component analyses, it was apparent that theOversensitivity items were of questionable homogeneity.Table 23. Correlations of Oversensitivity with domain scalesof the IASR-B5.OversensitivityScaleNeuroticism .28Openness .25Conscientiousness .19Dominance .25Nurturance .33Note. N = 706. All correlations p < .01.80Examination of the facet’s correlations with the dimensionsof the FF14 indicated no significant differentialrelationships (see Table 23).The Oversensitivity facet failed to demonstratediscriminant validity among the dimensions of the FFM.Examination of the item content indicated that the itemsinvolved the presence or absence of strong affectiveexperiences; whereas the remainder of the initially derivedneuroticism problem facets assessed specific affectiveexperiences. A decision was made to drop the Oversensitivityfacet from further analyses.The Nonconformity facet was also derived from the apriori openness item pool. Item component analysis indicatedgreater item homogeneity than the Oversensitivity facet.Examination of the facet’s correlations with the dimensionsTable 24. Correlations of Nonconformity with domain scalesof the IASR—B5.NonconformityScaleNeuroticism -.52Openness .26Conscientiousness .06Dominance .30Nurturance .13Note. N = 706. All correlations p < .01.81of the FFM demonstrated significant differentialrelationships (see Table 24).The Nonconformity facet correlated -.52 with traitneuroticism. As this facet appeared to belong on theneuroticism problems dimension, it was rescored withoutipsatization as were all other neurotic problems facets.When the scales were refactored, the unipsatizedNonconformity facet loaded most highly on the opennessdimension. Thus, subsequent principal components analyses ofthe openness facets including either the ipsatized orunipsatized Nonconformity facet were conducted (see Tables 25and 26).Table 25. Principal components analysis of Openness problemsfacets with Nonconformity ipsatized.FactorI IIScaleEgocentrism .07Unimaginative -.10Lack of Focus -.62 -.09Nonconformity .04 .99Note. N = 706.82Table 26. Principal components analysis of Openness problemsfacets with Nonconformity unipsatized.FactorIScaleEgocentrism .75Unimaginative .73Lack of Focus -.53Nonconformity -.43Note. N = 706.The principal components analysis presented in Table 25indicated a 2 component solution accounting for 62.3% of thevariance. When the Nonconformity facet is ipsatized (as areall other facets not loading on the neuroticism dimension) itis clearly unrelated to the three additional derived opennessfacets. And, the facet loads on the neuroticism dimension.However, when the Nonconformity facet is scored in anunipsatized manner (as are all other neuroticism problemsfacets), it loads substantially on the openness dimension.This indicates that the ipsatization process may be morecomplex than was originally hypothesized. A decision wasmade to score the Nonconformity facet in an unipsatizedmanner and retain it as a facet of the openness dimensionpending further analyses.83Examination of the correlations between the Urge Controlfacet and the dimensions of the FF11 indicated significantlygreater relationship between the facet and traitconscientiousness and a decision was made to retain the UrgeControl facet on the conscientiousness problems dimension.Examination of correlations between the Reckless facet andthe dimensions of the FF11 indicated significantly greaterrelationship between the facet and trait dominance assessedby the IASR—B5 (r = .44) and trait extraversion assessed bythe FF1 (r = .43) than trait conscientiousness (r = —.13 asassessed by the IASR-B5; r = —.05 as assessed by the FF1). Adecision was made to drop the Reckless facet from furtherstructural analyses. However, the facet was retained in theitem pool, as reckless behavior is of significant importanceis clinical assessment.After removal of the Oversensitivity and Reckless facetsfrom structural analyses, the remaining 13 derived problemsfacets were subjected to a conjoint principal componentsanalysis with appropriate domain scales from the IASR-B5.Examination of the Scree Plot indicated a three factorsolution accounting for 52% of the variance which was rotatedto varimax criterion. The final superordinate structure ofthe derived problems facets can be seen in Table 27.84Table 27. Three component solution of conjoint principalcomponents analysis of remaining derived problemfacets and domain scales of the IASR-B5.FactorI IScaleAnxietya.00 .03Depressiona.15 .00Neuroticism (IASR—B5) .83 —.03 .10Low Seif_esteema .76 .24 —.14Angera 5 .05 .03Conscientious (IASR—B5) -.12 —.83 .00Compulsive .05 —.80 .10Lazy .03 -.08Unreliable -.26 .52 .22Impulsive .19 .19Urge Control .15 .03Openness (IASR—B5) —.08 —.06Unimaginative -.23 -.07 -.78Egocentrism -.20 -.34 -.56Lack of Focus -.09 -.09 .47Nonconformitya-.35 .28Note. N = 706.aUnipsatized facet scales.85Reliability. Cronbach’s alpha was calculated for boththe individual facet scales and the domain scales of the lipB5. Examination of item-total correlations indicatedsignificant improvements in alpha reliabilities due toremoval of the following items from the following facets: a)deletion of item X112 (I’m concerned with keeping myappearance neat too much) and item Xl99 (I watch the clocktoo much) from the Compulsive facet, b) deletion of item X169(Hard for me to abide by rules and regulations even if theyare good ones) from the Nonconformity facet, and c) deletionof item XllO (I think about my health too much) from the UrgeControl facet. It was decided to retain the reduced itemfacets and generate new items to improve scale reliabilityand thematic homogeneity in further investigations. Finalalpha coefficients for the derived IIP-B5 facet scales can beseen in Table 28. IIP—B5 domain scales are scored asfollows:1. Neuroticism =Depression + Anxiety + Anger + Low Self-esteem2. Unconscientiousness =Lazy + Unreliable + Impulsive + Urge Control -Compulsive3. Openness =Nonconformity + Lack of Focus — Egocentric —UnimaginativeReliability coefficients ranged from .52 for Unreliabilityand Urge Control to .93 for domain Neuroticism. Aiphas for86Table 28. Alpha coefficients for derived facet and domainscales.AlphaScaleDepression .89Anxiety .88Low Self—esteem .81Anger .82Neuroticism (IIP-B5) .93Lazy .70Unreliable .52Impulsive .71Urge Control .52Compulsive .80Unconscientiousness (IIP—B5) .67Unimaginative .74Egocentric .78Lack of Focus .65Nonconformity .62Openness (IIP—B5) .73Reckless .74Note. N = 706the former two scales may be attenuated due to moderatevariance of these scales in a college sample.87Structural Validity. The final derived IIP-B5 problemsdomain scales were subjected to a conjoint principalcomponents analysis with the domain scales of the IASR-B5.This analysis also included the circumplex dimensions ofdominance and nurturance from both the IASR-B5 and the lIP-Cto determine the structural fidelity of the derived scaleswith the Dyadic-Interactional perspective on the FFM. Theanalysis indicated a five factor solution accounting for84.8% of the variance which was rotated to varimax criterion(see Table 29). Structural fidelity was supported.Table 29. Conjoint principal components analysis of IASR-B5domain scales and IIP—B5 domain scales.FactorsI II III IV VScalesNeuroticism (IASR-B5) .94 —.03 —.05 .00 .05Neuroticism (11P-B5) .89 .13 —.15 —.17 —.04Unconscientious (IIP-B5) .02 .90 .06 .08 .07Conscientious (IASR-B5) -.07 —.90 .12 .14 —.04Dominance (IIP—B5) .01 .01 .91 —.19 .11Dominance (IASR—B5) -.27 —.07 .87 .13 .16Nurturance (IIP—B5) —.02 .09 .07 .91 .08Nurturance (IASR—B5) -.15 —.15 —.15 .86 .03Openness (IASR—B5) —.01 -.10 .19 .08Openness (IIP—B5) .02 .22 .06 .03Note. N = 706.88A principal component analysis of the domain scales ofthe IIP-B5 also demonstrated strong support for a five factorstructure of the instrument. The final structural analysesapplied to the derived IIP-B5 scales was to subject allderived facet scales and circumplex octant scales to aprincipal components analysis with markers of theneuroticism, conscientiousness and openness from the IASR—B5.Results of the analysis can be found in Table 30. A fivefactor solution accounting for 59.4% of the variance wasrotated to varimax criterion. All derived facets loaded mosthighly on the expected dimensions, further supporting thestructural fidelity of the IIP-B5 at the facet level.Table 30. Full facet and circumplex octant principalcomponents analysis of the IIP-B5.FactorsI II III IV VScalesAnxietya-.06 -.04 -.02 .03Depressiona.85 —.18 .00 —.18 .03Neuroticism (IASR—B5) .83 .04 .01 .00 .07Low Self_esteema .77 —.26 —.17 —.26 —.08Angera.11 .42 —.09 —.04liP—C (FG) .22 —.83 —.17 .02 —.14liP—C (DE) —.06 —.72 .27 .09 —.06lIP—C (NO) —.13 .71 .06 —.04 .1789Table 30 continuedlIP—C (JK) —.07 —.06 —.84 .06 —.07lIP—C (HI) .11 —.45 —.72 —.05 —.21lIP—C (PA) —.20 .34 .68 .18 .17lIP—C (BC) .03 —.29 .13 —.14lIP—C (LM) —.17 .35 —.66 .23 .17Conscientious (IASR—B5) —.17 —.01 .03 .82 —.09Compulsive .00 .05 .19 -.01Lazy .08 .05 .00 —.63 —.02Unreliable —.24 .09 .05 —.51 .28Impulsive .24 .40 .21 —.48 .12Urge Control .19 .15 .12 -.40 .01Openness (IASR—B5) —.12 .17 .13 .06 .81Unimaginative —.19 —.04 —.02 .07 —.79Egocentric —.20 -.01 .21 .33 -.60Lack of Focus -.10 .36 .03 .12 .42Nonconformitya-.35 .10 .29 -.29Note. N = 706. aunipsatized scales.Convergent Validity. The IIP-B5 domain scales werecorrelated with the domain scales of the FF1 and IASR-B5 toassess convergent validity. These correlations can be foundin Table 31. Additional validity correlations between IIP-B5facets and the Interpersonal Dependency Inventory (Hirschfeld90et al, 1977), the State-Trait Anxiety Inventory (Spielberger,1983), and the Shyness Scale (Cheek & Melchior, 1985) can befound in Table 32.Table 31. Correlations between IIP—B5 domain scales anddomain scales of the FF1 and IASR-B5.IASR-B5N QIIP—B5 ScalesNeuroticism .80 -.13 -.31 .72 -.09 -.20openness .03* —.14 .04* -.20Unconscientious .15 .10 —.72 .00* .05* —.66Note. N = 706. *r is not significant.The convergent validity correlations between the IIP-B5domain scales and the IASR—B5 domain scales support thecontention that each derived problem domain is assessingmaladaptive traits related to one of the dimensions of theFFM. Each IIP-B5 domain correlated most highly with its FF1and IASR—B5 counterpart. While these correlations areacceptable, further revisions of the IIP—B5 scales may reducethe off diagonal correlations in Table 31 and increase thecorrelation between IIP—B5 openness problems and traitopenness assessed by the FF1 and IASR-B5. It should be notedthat in the IIP—B5, the conscientiousness domain is scored inthe direction of unconscientiousness, leading to the91consistent negative factor loadings and correlations withother measures of trait conscientiousness.Table 32. convergent validity correlations between IIP-B5facets and the STAI Trait anxiety scale, theInterpersonal Dependency Inventory, and theShyness Scale.Validity ScalesSTAI—T IDI—1 IDI—2 IDI—3 IDI-Tot SHYIIP-B5 FacetsDepression .78 .48 .44 —.21 .48 .58Anger .41 .28 .21 .28Anxiety .78 .53 .47 —.24 .53 .57Low Self—esteem .76 .60 .38 —.24 .52 .59LazyCompulsiveUnreliableImpulsiveUrge ControlNonconformity -.29 -.25 -.27Lack of FocusUnimaginative -.23Egocentric -.25Note. N = 706. All r’s p < .01. r < .20 deleted.Abbreviations: STAI-T = Trait Anxiety Scale; IDI-1 = Lack ofSocial Self-confidence; IDI-2 = Emotional Reliance on Others;IDI-3 = Assertion of Autonomy; IDI-Tot Dependency; SHY =Shyness Scale.92The majority of correlations above .20 in Table 32involve the IIP—B5 neuroticism facets. The depression facetis strongly related to trait anxiety (.78) and moderatelyrelated to shyness (.58). It is moderately related todependency (.48) and negatively related to the assertion ofautonomy (-.21). The anxiety facet shows a similar patternof convergent correlations——strongly related to trait anxiety(.78), moderately related to shyness (.57) and dependency(.53), and negatively related to the assertion of autonomy(-.24). The anger facet is moderately related to traitanxiety (.41), and was less strongly related to dependency(.21) and shyness (.28). The low self—esteem facet exhibitsstrong associations with trait anxiety (.76), dependency(.52) (particularly lack of social self—confidence r = .60),and shyness (.59).None of the unconscientious problems facets exhibitedsignificant relationships to anxiety, dependency, or shyness.The nonconformity facet exhibited moderate negativerelationships with a lack of social self-confidence (—.29),dependency (-.25), and shyness (-.27). This seemsintuitively congruent, as nonconforming behaviorcontraindicates lack of confidence, shyness, or dependence onothers. The egocentric facet was negatively correlated withtrait anxiety (-.25) indicating a relationship betweenemotional stability and the ability to reject people orcustoms very different from one’s own. The unimaginativefacet was also negatively correlated with trait anxiety93(-.23). Interpretation of this relationship is difficult;however it could be that concrete thinking attenuates self—consciousness and the ability to ruminate over possiblefuture events or possible alternatives to previouslyexperienced situations. Simply put, concrete thinking mayreduce a person’s tendency to worry.DiscussionGeneral Factor. The general factor running through thedomain of self-reported problems noted by Horowitz et al(1988) and Alden et al (1990) replicated in the enlargedexperimental item pool of neurotic, conscientious, andopenness problems. When this factor was correlated withindependent measures of the FFM, it was most strongly relatedto trait neuroticism and mean item endorsement.Gurtman (in press-b, in press-c) noted that for the lIPC this general factor contributed to scale elevation (orinterdependencies among octant scales) which could beassessed by calculating the average intercorrelation amongliP—C octant scales. When these scales are unipsatized, theintercorrelation matrix includes all positive correlationsrather than a circumplex pattern of correlations (Wiggins etal, 1981).Gurtman applied multidimensional scaling techniques tothe all positive liP-C intercorrelation matrix and determinedthat a strong circulant property remained intact. This ledhim to conclude that the ipsatization procedure may excludeinformation of importance for assessment. However, Gurtman94was working within a two—dimensional interpersonal system;whereas the present study involved the FFM. One of theadditional dimensions of the FFM is, in fact, stronglyrelated to the general factor. Thus it was concluded that:a) the general factor is most strongly related to traitneuroticism, b) neurotic problem items should be excludedfrom the ipsatization procedure, and c) ipsatization would beused to remove the effects of the factor on facetintercorrelations for the remaining FFM domains. Therelationship between trait neuroticism and elevation in self—reported problems is congruent with recent findings that thistrait is significantly related to both recall of, anddistressed reactivity to physical symptoms and dailystressors (Affleck, Tennen, Urrows, & Higgins, 1992; Bolger &Schilling, 1991; Costa & McCrae, 1987; Larsen & Kasimatis,1991; Larsen, in press; Watson & Pennebaker, 1989).Scale Derivation. Initial facet and domain scales forneurotic, conscientious, and openness problems were derivedfrom the experimental item pool. The problems construct ofbehavioral excess and behavioral inhibition was expanded fromthe interpersonal domain to the additional personalitydomains of neuroticism, conscientiousness, and openness.Initial facet scales for neuroticism are unipolar andinclude: Depression (sad affect) (+), Anxiety (+), Low Self—Esteem (+), and Anger Control (+). Initial facet scales forunconscientiousness are bipolar and include: Compulsivity95(—), Lazy (+), Impulsivity (+), Urge Control (+), andUnreliable (+). Initial facets scales for Openness arebipolar and include: Egocentric (-), Unimaginative (-), Lackof Focus (+), and Nonconformity (+). The Nonconformity facetloaded strongly on the openness dimension only if it remainedunipsatized. This indicated that the substantive effects ofipsatization may differ if the target facet or scale isstrongly negatively correlated with trait neuroticism than ifthe target facet or scale is positively correlated withneuroticism. If a problems scale is negatively related totrait neuroticism, retaining the distress component appearsto reduce the strong negative correlation, leaving the facetor scale free to load on the most substantively relateddimension.An additional problems facet, Reckless, was retained asclinically significant, although it is most strongly relatedto trait extraversion. This facet appears to assessmaladaptive sensation—seeking (Zuckerman, 1971; Zuckerman,Bone, Neary, Mangelsdorff, & Brustman, 1972), a constructthat has been a thorn in the structural side of the FFM overthe years (see Zuckerman, ICuhiman, & Camac, 1988).Scale reliabilities ranged from .52 to .93. While thelowest reliabilities may have been attenuated by low variancein university sample, item changes may be required to raisealpha coefficients to acceptable levels, Convergent validitycorrelations between IIP—B5 domain scales for neurotic,unconscientious, and openness problems and independent96measures of the FFM (FF1, IASR-B5) supported the expectationthat the new scales were most strongly related to the FFMdimension for which scale construction was targeted.Validity correlations for derived neuroticism facet scalesindicated excellent convergence. The five item Anxietyproblems facet correlated .78 with the 20-item Trait Anxietyscale of the STAI. All neuroticism facets showed significantrelations with measures of dependency and shyness. Theunconscientiousness problems facets failed to demonstratesignificant relations with the scales from the IDI, STAI, orthe Cheek Shyness scale. These facets may be a more uniquecomponent of the derived scales, and such constructs may notbe well represented in existing adjustment inventories (e.g.,Costa et al, 1986). Derived openness facets demonstratedsmall and meaningful relations with dependency, lack ofsocial self—confidence, shyness, and trait anxiety. As withthe facets of unconscientiousness, openness problems mayrepresent relatively unique constructs not well representedin available self—report inventories.Structural Fidelity. Loevinger’s (1957) concept of“structural fidelity” proposes that items be selected from alarge pool on the basis of empirical properties that indicatewhich items best conform to an appropriate structural model.In the present research, items were selected to conform tothe Dyadic-Interactional perspective on the FFM (Pincus &Wiggins, in press; Trapnell & Wiggins, 1990, Wiggins &Pincus, 1992, in press). To assess the structural fidelity97of the derived problems scales, the IIP-B5 domain scales(including the interpersonal problems circumplex axes) andthe IASR-B5 were subjected to a conjoint principal componentsanalysis. A clear five factor solution was extracted thatshowed the two instruments demonstrate homologous structure(see Table 29). No off-diagonal factor loadings ofsignificance emerged. At the facet and octant level, aprincipal components analysis of the full IIP-B5 alsogenerated a clear five factor solution (see Table 30). Facetfactor loadings on nontarget dimensions are comparable tosuch loadings reported for full factor analyses of therevised NEO—PI (NEO—PIR; Costa, McCrae, & Dye, 1991) and theStandard Markers of the Big Five Factor Structure (Goldberg,1992).General Conclusion. The results of Study 1 support theexpectations that the problems construct can be expandedbeyond the interpersonal domain of personality. An initialfive-factor problems inventory (IIP-B5) was developed whichshowed structural fidelity with the Dyadic-Interactionalperspective on the FFM. Additionally, the derived facetscales appear to assess constructs which may be interpretedas lualadaptive trait representatives of the dimensions of theFFM, rather than redundant constructs found in other FF14inventories. Facet scale reliability requires improvement;and validation of the inventory at both the facet and domainlevel of the construct hierarchy in an independent sample is98required. The results of the scale validation study can befound in Chapter 3.99CHAPTER 3Study 2: Scale ValidationMethodsOverviewThe major goal of Study 2 was to validate the IIP-B5scales derived in Study 1.Sanrnle. The derived IIP-B5 (including additional newitems), along with a variety of FFM inventories, andadditional clinical, adjustment, and personalityquestionnaires was administered to a sample of 572 universitystudents. This sample consisted of 201 men (35%) with anaverage age of 19.9 years, and 371 women (65%) with anaverage age of 20.0 years. This sample was used to validatethe IIP—B5 derived in Study 1.Instruments. The initial version of the IIP-B5 derivedin Study 1 with item modifications to clarify substantiveinterpretation of derived factors and to improve internalconsistency was included in a large assessment battery. Inthis study, the lIP-C (Alden et al, 1990; described inChapter 2) was embedded within the test booklet. A varietyof adjustment and personality inventories were included inthe battery. Some of these pertained to independent researchprojects and are not be discussed.The IASR-B5 (Trapnell & Wiggins, 1990) and the FF1(Costa & McCrae, 1989) will again be included to provideindependent and method-different markers of the FFM. Theseinventories have been described in Chapter 2.100The Brief Symptom Inventory (BSI; Derogatis &Melisaratos, 1983) is a self—report symptom scale that wasdesigned to measure levels of psychopathology. It is ashortened form of the revised version of the SymptomChecklist-90 (SCL—R-90; Derogatis, 1977). The BSI consistsof 53 items describing a variety of difficulties andcomplaints. The items are rated on a 5-point scale (0 to 4),reflecting degrees of distress ranging from “not at all” to“extremely.” Scores are obtained on the following ninedimensions: Somatization, Obsessive—Compulsive,Interpersonal Sensitivity, Depression, Anxiety, Hostility,Phobic Anxiety, Paranoid Ideation, and Psychoticism. Threeglobal indexes can be calculated from raw scores.The Weinberger Adjustment Inventory (WAI; Weinberger &Schwartz, 1990) is an 84 item inventory designed to assesslong term functioning rather than short term symptoms. Thetwo primary dimensions are “Distress” (as assessed bysubscales measuring anxiety, depression, low self—esteem, andlow well—being) and “Restraint” (as assessed by subscalesmeasuring impulse control, suppression of aggression,consideration of others, and responsibility). There are twovalidity indicators consisting of a “Denial of Distress”scale, which refers to defensiveness about normativeexperiences of distress, and a “Repression” scale whichrefers to claims of nearly absolute restraint. Subjectsendorse items on a 5—point scale ranging from completelyfalse to completely true. Alpha reliabilities range from .72101for the Impulse Control scale to .92 for the compositeDistress scale.The Private and Public Self—Consciousness scales(PRSC/PUSC; Fenigstein, Scheier, & Buss, 1975) are 10 and 7-item scales which assess a) the tendency to attend to one’sinner thoughts and feelings and b) the awareness of the selfas a social object having an effect on others, respectively.Subjects respond to each item on a 5—point scale from“Extremely Uncharacteristic” to “Extremely Characteristic.”Test—retest reliability for these scales was greater than .7(Fenigstein et al, 1975).Procedures. Data collection procedures proceeded asdescribed in Study 1.AnalysesData Transformation. The ipsatization proceduredescribed in Study 1 was again used to transform the data.In this study, both ipsatized and unipsatized versions of thescales were scored and saved for further analyses.Cross Validation. Each of the problems domains ofneuroticism, conscientiousness, and openness were subjectedto independent principal components analyses at the itemlevel to demonstrate scale stability across samples.Structural Validation. The initial IIP—B5 facetsderived from Study 1 were subjected to a conjoint principalcomponents analysis with markers of neuroticism,conscientiousness, and openness from the IASR—B5. Followingthe same step-wise procedure detailed in Study 1, the facets102were then conjointly factored with the liP-C octant scales.Following confirmation of the lower—level structuralstability of the IIP-B5 facets, the domain scales wereconjointly factored with the IASR-B5.Reliability and Validity. Alpha coefficients and item-total correlations for derived facets and domain scales weredetermined to assess internal consistency. Convergentvalidity correlations for the three new problems dimensionswere obtained by examining the relations between derivedproblems scales and the neuroticism, conscientiousness, andopenness scales of the IASR—B5 and the FF1 concurrentlyassessed in the sample. Additional validity correlationswere obtained by examining the relations between the derivedIIP-B5 scales and the BSI, WAI, and Self-consciousnessscales.Gender differences in normative endorsement rates wereexamined. Normative data were used to generate standardizedscores for purposes of clinical assessment and research.Expectations1. Both the lower order and superordinate structure ofthe IIP-B5 will be validated on a second large sample.2. The reliability, convergent validity, anddiscrimininant validity of the IIP-B5 will be supported.3. Significant gender differences will emerge inendorsement rates of problems.103ResultsItem modifications. Based on a review of alphareliabilities, item loadings, and the thematic content offacets, a limited number of additional items or itemmodifications were introduced into the IIP-B5 validation itempool. Item changes and additions are presented in Table 33.Table 33. Summary of item modifications and additions toderived IIP-B5 facets.Facet Item ModificationLazy “Hard to set clear goals and work efficientlytowards them”MODIFIED TO“Hard to work systematically toward my goals”Unimaginative “Hard to feel captivated when looking at awork of art”MODIFIED TO“Hard to be moved by a work of art”Urge Control “I think about my health too much” (reversed)MODIFIED TO“Hard to act responsibly towards my health”104Table 33 continuedItem AdditionUnimaginative “Hard to be absorbed when listening to music”“Hard to reflect seriously about things”“I get bored too often when conversationsturn to deep topics”Nonconformity “I want to stay with the usual way of doingthings too much” (reversed)Egocentric “Hard to accept customs different from myown’,“I get too upset by others’ unconventionalbehavior”Facet Validation. The 20 unipsatized neurotic problemsitems were subjected to a principal components analysis.Both oblique and varimax rotations were performed. As thesewere virtually identical, only the varimax rotation will bepresented here. A four component solution accounting for49.2% of the item variance was extracted. This solution ispresented in Table 34.All items had their highest loadings on the originalderived problems facets. Facets were again identified asAnxiety, Depression (sad affect), Anger Control, and Low105Table 34. Principal components of selected neuroticism items.Factors2. 3 4.ItemX94 .73X146 .69X85 .67X137 .65Xl28 .61X104 .76X136 .73X144 .37 .72X125 .38 .44X29 .44X69 .82X87 .75X12 .63X61 .61X50 .49X19 .73X10 .70X99 .57X62 .48X71 .36 .40Note. N 572. Loadings < .33 deleted.106Self—esteem. Two depression items (“I feel depressed toomuch” and “I feel discouraged too much”) had secondaryloadings on the Anxiety facet. A decision was made to retainthese items on the Depression facet. One low self—esteemitem (“Hard for me to feel ambitious”) split its loadingbetween its target facet (.40) and the Depression facet(.36). A decision was made to retain the item on the LowSelf-esteem facet.The 25 ipsatized openness problems items were subjectedto a principal components analysis. Both oblique and varimaxrotations were performed. As these were extremely similar,only the varimax rotation is presented here. A fourcomponent solution accounting for 41.8% of the item variancewas obtained. This solution is presented in Table 35.Table 35. Principal components of selected openness items.Factors1 2. 4.ItemX143 .84X130 .82X3l .76X2 .73X27 .59X55 24aX46 .70X58 .70107Table 35 continuedX47 .62X18 .57X5l .56X138 .46X22 .38X25 .79X49 .77X7 .68X52 .62X54 .41X134 .37X141 24aX88 .72X135 .67X101 .66X89 .65X147 .40Note. N 572. aAll other loadings < .33 deleted.All items had their highest loadings on the originalderived problems facets. New items’ highest loadings alsofell on expected facets. Facets were again identified asNonconformity, Unimaginative, Egocentric, and Lack of focus.Two items had low loadings (.24) on their target facets.However, these items showed significantly lower loadings,108ranging from .02 to .09 on other facets. Thus, a decisionwas made to retain these items.The 26 ipsatized unconscientious problems items weresubjected to a principal components analysis. Both obliqueand varimax rotations were performed. Only the varimaxrotation is presented here. A five component solutionaccounting for 42.7% of the item variance was obtained. Thissolution is presented in Table 36.Table 36. Principal components of selected unconscientiousitems.Factors1 2.I ternX81 .70X84 .68X9 .68X42 .65X39 .55X102 .75Xl22 .70X93 .69X97 —.40 .44X126 .42X20 .41X124 .72Xl29 .68109Table 36 continuedX91 .53XlOO .49X82 .38 .45X53 .71X79 .68X24 .60X4 .48X43 .43X77 .76X1l3 .68X40 .53X105 .50X127 .50Note. N = 572. All loadings < .33 deleted.All items had their highest loadings on the originalderived problems facets. New items’ highest loadings alsofell on expected facets. Facets were again identified asLazy, Compulsive, Impulsive, Urge Control, and Unreliable.Item X97 (“I constantly strive for perfection too much”)split its loading on Lazy (-.40) and Compulsive (.44). Adecision was made to retain the item in the Compulsive facet.Item X82 (“I overlook important details too much”) split itsloading on Impulsive (.45) and Lazy (.38). A decision wasmade to retain the item in the Impulsive facet.110Structural Validation. In order to demonstratestructural stability of the IIP—B5 across samples, the 13derived problems facets were subjected to a conjointprincipal components analysis with domain scales from theIASR-B5. Examination of the Scree Plot indicated a threefactor solution accounting for 51.2% of the variance whichwas rotated to a varimax criterion. The final superordinatestructure of the derived problems facets can be seen in Table37. The three component solution replicates the solution ofthe same analysis conducted in Study 1.The only facet with a significantly split loading wasNonconformity (.43 on openness and .40 on neuroticism). Thisfacet demonstrated a similar pattern of factor loadings inStudy 1. In a recent study of the structure of self—reportschizotypy (psychosis proneness) scales, Kendler and Hewitt(1992) conjointly factored ten scales including Chapman’sImpulsive Nonconformity scale (Chapman, Chapman, Numbers,Edell, Carpenter, & Blackfield, 1984) with Eysenck’s revisedNeuroticism and Extraversion scales (Eysenck, Eysenck, &Barrett, 1985). Nonconformity demonstrated a similar patternof loadings to the present study. The ImpulsiveNonconformity scale loaded most highly on a factorinterpreted as “nonconformity”, and had a moderate secondaryloading on a dimension strongly marked by Eysenck’sNeuroticism scale. Thus, the Nonconformity facet wasretained in the openness problems domain.111Table 37. Three component solution of conjoint principalcomponents analysis of IIP—B5 facets and domainscales of the IASR-B5.FactorU .1.11ScaleAnxietya-.02 .07Depressiona.13 .14Neuroticism (IASR—B5) .81 -.05 .15Low Seif_esteelna .22 —.09Angera.03 -.02Conscientious (IASR-B5) -.24 -.79 —.04Compulsive .02 —.77 .06Lazy —.04 —.10Unreliable -.15 .21Urge Control .03 .50 .14Impulsive .19 —.05Openness (IASR-B5) —.10 —.04Unimaginative —.27 —.14 -.77Egocentrism -.16 -.21 - .50Lack of Focus —.04 —.20 .46Nonconformitya-.40 .27 .43Note. N = 572. aUnipsatized facet scales.112As a second test of the superordinate facet structure ofthe derived IIP-B5 scales, they were combined with the liP-Coctant scales and conjointly factored with markers ofneuroticism, conscientiousness, and openness from the IASR—B5. A principal components analysis indicated a clear fivefactor solution emerged that accounted for 60.1% of thevariance. The solution was rotated to a varimax criterionand can be seen in Table 38. The full facet/octant structureof the IIP-B5 from Study 1 was replicated in this analysis.While all facet scales load most highly on the targetdimension, a number of facets show moderate secondaryloadings. Both the Anger Control facet (neuroticism) and theEgocentric facet (openness) have moderate secondary loadingson an interpersonal dimension marked by the liP—C Domineeringand lIP—C Vindictive scales. Both the Lack of Focus facet(openness) and the Impulsive facet (unconscientious) havesecondary loadings on an interpersonal dimension marked bythe liP—C Intrusive scale. The Nonconformity facet has asecondary loading on the neuroticism dimension (as expected)and a moderate secondary loading on the unconscientiousnessdimension.To validate the structural fidelity of the IIP-B5 withthe Dyadic-Interactional perspective on the FFM, a conjointprincipal components analysis of the IIP-B5 domain scaleswith adjectival domain scales for the same dimensions fromthe IASR-B5 was conducted. A clear five factor solution113Table 38. Full facet and circumplex octant principalcomponents analysis of the IIP-B5.FactorsI II III IV VScalesAnxietya.88 -.08 -.03 -.03 .04Depressiona—.06 —.12 .10 .15Neuroticism (IASR—B5) 2 -.03 .03 —.05 .12Low Seif_esteema —.23 —.17 .18 —.05Angera.42 .11 .05 -.10liP—C (JK) .01 —.83 .03 —.06 —.02lIP—C (HI) .10 —.72 —.45 .03 —.06lIP—C (PA) —.25 .22 —.12 .16lIP—C (BC) —.02 &9. —.41 —.08 —.03liP—C (LM) —.05 —.59 .46 —.28 .13liP—C (FG) .20 —.20 —.77 —.01 —.07lIP—C (NO) —.02 —.01 7.4. — .02 — .03lIP—C (DE) .02 .23 —.71 —.06 .05Conscientious (IASR—B5) —.25 .06 .07 -.76 -.05Compulsive .04 .18 .25 —.73 .00Lazy —.03 —.12 .09 7Q. —.12Unreliable —.11 .11 —.04 .18Urge Control .05 .12 -.04 .14Impulsive .24 .10 .40 —.16114Table 38 continuedOpenness (IASR—B5) —.08 .10 .02 —.04Unimaginative —.29 —.02 —.01 —.14 —.76Egocentric -.13 .36 -.04 -.17 -.52Lack of Focus .01 .10 .34 —.16 .49Nonconformitya—.38 .27 .02 .31 4.Q.Note. N = 572. aunjpsatjzed scales.Table 39. Conjoint principal components analysis of IASR-B5domain scales and IIP—B5 domain scales.FactorsI II III IV VScalesNeuroticism (IASR—B5) .92 —.09 —.02 .07 .01Neuroticism (IIP—B5) .88 —.18 .14 —.02 —.13Dominance (IIP—B5) —.03 .93 —.05 .08 —.15Dominance (IASR—B5) —.32 .85 —.12 .16 —.11Unconscientious (IIP-B5) -.05 -.01 .03 .04Conscientious (IASR—B5) —.19 .15 -.87 -.06 .14Openness (IASR—B5) —.03 .08 —.10 .92 .00Openness (IIP—B5) .09 .11 .22 .87 .08Nurturance (IIP-B5) .05 .10 .02 .02 .92Nurturance (IASR-B5) -.18 -.19 -.12 .05 .85Note. N = 572.115accounting for 85.2% of the variance was rotated to varimaxcriterion. This solution is presented in Table 39.Examination of Table 39 indicates a clear homologousfive factor structure between the IIP-B5 and the IASR—B5.The only moderate off-diagonal loading is found for IASR-B5Dominance which has a secondary negative loading on theneuroticism dimension. This is consistent with the resultsof Trapnell and Wiggins (1990) who found that the IASR-B5dominance dimension correlated negatively with the NEO-PINeuroticism scale (r = —.35) and positively with the HPIAdjustment scale (r = .42).Reliability. Cronbach’s alpha was calculated for allIIP—B5 facet and domain scales for neuroticism,conscientiousness, and openness. Reliability coefficientsare presented in Table 40. Alpha coefficients for all scalesremained stable across derivation and validation samples, andscales which were modified specifically to improve internalconsistency demonstrated modest gains. Neuroticism problemfacets’ and domain alpha coefficients are virtually identicalacross samples. The Lazy facet included one itemmodification and the facet alpha rose from .70 to .75 acrosssamples. The remaining unconscientiousness facets’ anddomain alpha coefficients remained stable across samples.The Unimaginative and Egocentric facets of openness includedadditional items which produced modest increases in internalconsistency. The alpha for the Unimaginative facet rose from.74 in Study 1 to .81 in Study 2. The alpha for the116Table 40. Alpha coefficients for derived facet and domainscales.AlphaScaleDepression .87Anxiety .87Low Self—esteem .80Anger .80Neuroticism (IIP-B5) .92Lazy .75Unreliable .53Impulsive .70Urge Control .55Compulsive .82Unconscientiousness (IIP—B5) .68Unimaginative .81Egocentric .84Lack of Focus .65Nonconformity .60Openness (IIP—B5) .79Reckless .71Note. N = 572Egocentric facet rose from .78 in Study 1 to .84 in Study 2.The Nonconformity and Lack of Focus facets exhibited stablealphas across the samples. Increases in openness facet117aiphas contributed to a modest increase in internalconsistency for the IIP—B5 Openness domain scale, which rosefrom .73 in Study 1 to .79 in Study 2.Convergent Validity. The IIP-B5 domain scales werecorrelated with the domain scales of the FF1 and IASR-B5 toassess convergent validity. These correlations can be foundin Table 41. Additional validity correlations between IIP—B5facets and the Brief Symptom Inventory (Derogatis &Melisaratos, 1983), the Weinberger Adjustment Inventory(Weinberger & Schwartz, 1990), and the Private and PublicSelf—Consciousness scales (Fenigstein, Scheier, & Buss, 1975)can be found in Tables 42 to 44.Table 41. Correlations between IIP—B5 domain scales anddomain scales of the FF1 and IASR-B5.FF1 IASR-B5N N 2 2IIP—B5 ScalesNeuroticism .76 _.02* —.31 .70 _.08* —.31Openness .01* —.18 .09* —.22* * * *Unconscientious .09 .04 —.72 —.02 —.04 —.63Note. N = 572. *r is not significant.Examination of Table 41 indicates significant convergentvalidity between the domain scales of the IIP—B5 and those ofthe FF1 and IASR-B5. The IIP-B5 Unconscientiousness domain118scale correlates significantly with only its FF1 and IASR-B5counterparts. The IIP—B5 Neuroticism domain scale convergessignificantly with its FF1 and IASR-B5 counterparts, isunrelated to trait openness, and exhibits a modest negativesecondary correlation with trait conscientiousness. The lIP—B5 Openness domain scale also converges significantly withits FF1 and IASR-B5 counterparts, is unrelated to traitneuroticism, and exhibits a modest negative secondarycorrelation with trait conscientiousness.When examining external validity correlations for theIIP-B5 facets and the Brief Symptom Inventory (Derogatis &Melisaratos, 1983), statistical power of the large samplesize allows for correlations as low as .12 to be significantat p < .01. As scales were constructed to conform to aspecific dimensional model I felt it was important to focuson the important relations between the two instruments.Therefore, a cut—off doubling the significance value (r =.25) was chosen as an indication of correlation magnitudes ofrelative significance. When this value was selected, theonly IIP-B5 facets to demonstrate relatively significantrelations with the BSI scales were the facets of neuroticism.Only one other facet (Nonconformity) exhibited a singlecorrelation greater than .24 with a BSI scale. TheNonconformity facet correlated with the BSI Obsessivecompulsive scale (r = .28). The correlations between the BSIand the IIP-B5 neuroticism facets are presented in Table 42.119The IIP-B5 Depression facet correlated most highly withthe BSI Depression scale (r = .64) and exhibited an averagecorrelation with the remaining BSI scales of .46. This facetalso correlated .59 with the global severity index (GSI).Table 42. External validity correlations between IIP-B5neuroticism facets and the Brief SymptomInventory.IIP-B5 Neuroticism FacetsDEP ANX LSE ANGBSI ScalesSCM .36 .36 .30 .25CBS .51 .56 .45 .36INT .55 .55 .59BSI DEP .64 .53 .53 .29BSI ANX .49 .42 .34HOS .38 .36 .25PHO .43 .49 .48PAR .45 .43 .41 .29PSY .53 .47 .45 .30GSI .59 .58 .52 .37Note. N 572. All r’s p < .01. r < .25 deleted.Abbreviations: BSI=Brief Symptom Inventory;SOM=somatization; OBS=obsessive—compulsive; INT=interpersonalsensitivity; BSI DEP=depression; BSI ANX=anxiety;HOS=hositility; PHO=phobic anxiety; PAR=paranoid ideation;PSY=psychoticism; GSI=global severity index; DEP=depression(IIP—B5); ANX=anxiety (IIP—B5); LSE=low self—esteem;ANG=anger control.120The IIP-B5 Anxiety facet correlated most highly with the BSIAnxiety scale (r .58) and exhibited an average correlationwith the remaining BSI scales of .47. This facet alsocorrelated .58 with the GSI. The IIP-B5 Low Self—esteemfacet correlated most highly with the BSI InterpersonalSensitivity scale (r = .59) and exhibited an averagecorrelation with the remaining BSI scales of .41. This facetalso correlated with .52 with the GSI. The IIP-B5 AngerControl facet correlated most highly with the BSI Hostilityscale (r = .45) and exhibited an average correlation with theremaining BSI scales of .29. This facet also correlated .37with the GSI. The relationship between self-reportedsymptoms and neuroticism is supported by the consistentmoderate correlations among IIP—B5 neuroticisin facets and theBSI scales. Additionally, the moderate to high correlationsfound between IIP-B5 Neuroticism facets and all BSI scalessuggest that the general distress factor runs through theself—report BSI scales and creates scale interdependencieswhich blur substantive discriminability. The lack ofrelations among IIP—B5 unconscientious and openness facetsand BSI scales suggests that these facets assess problemsrather than psychiatric symptoms (e.g., Horowitz, 1979) andsupport the use of the ipsatization procedure to control forthe general factor.The same correlation magnitude cut—off (r = .25) wasadopted when examining the correlations between the IIP-B5facets and the Private and Public Self—Consciousness scales121(Fenigstein et al, 1975). Correlations greater than .24 arepresented in Table 43. Only three IIP-B5 facets exhibitedcorrelations greater than .24 with the self—consciousnessscales. The IIP-B5 depression facet correlated .32 with thePrivate Self-Consciousness scale (PRSC) and .25 with thePublic Self-Consciousness scale (PUSC). The IIP-B5 anxietyfacet correlated .30 with both the PRSC and PUSC scales. Theonly other IIP—B5 facet to correlate with these scales wasUnimaginative, which correlated -.39 with PRSC. ThisTable 43. External validity correlations between IIP-B5facets and the Fenigstein Public and Private Self-Consciousness scales.Self—Consciousness ScalesPrivate PublicIIP—B5 FacetsDepression .32 .25Anxiety .30 .30Unimaginative -.39Note. N 572. Correlations < .25 deleted.negative correlation suggests that a concrete thinking stylemay interfere with the ability for self—reflection.A subset of 307 of the 572 subjects in Study 2 completedthe Weinberger Adjustment Inventory (WAI). When examiningthe external validity correlations among IIP-B5 facets andWAI scales, the statistical power of the sample size allows122for correlations as low as .15 to be significant at p < .01.A cut—off doubling this value (r .30) was chosen as anindication of correlation magnitudes of relativesignificance. IIP—B5 facets correlating greater than orequal to .30 with WAI scales are presented in Table 44.Table 44. External validity correlations between IIP—B5facets and the Weinberger Adjustment Inventory.IIP—B5 FacetsDEP ANX LSE ANG IMP NONC RECKWAI ScalesWAI ANX .59 fl .53 .30WAI DEP .79 .65 .65WAI LSE .64 .61LWB .58 .48 .47 —.35DISTRESS .78 .76 .72 .31 —.33SUPRESS AGG -.57IMPL CON -.39 —.43 —.47CONSID -.31RESPONS -.38RESTRAINT —.56 -.37 —.31Note. N = 307. All r’s p < .01. r < .30 deleted.Abbreviations: DEP=depression; ANX=anxiety; LSE=low selfesteem; ANG=anger control; IMP=impulsivity;NONC=nonconformity; RECK=reckless; WAF=Weinberger AdjustmentInventory; WAI ANX=WAI anxiety; WAI DEPWAI depression; WAILSE=WAI low self-esteem; LWB=low well-being; SUPRESSAGG=suppression of aggression; IMPL CON=impulse control;CONSID=consideration of others; RESPONS=responsibility.123Examination of Table 44 indicates significant relationsamong the WAI Distress subscales and domain scale and theIIP—B5 neuroticism facets of Depression, Anxiety, and LowSelf-Esteem. IIP—B5 Depression correlatesmost highly with WAI Depression (r .79). IIP-B5 Anxietycorrelates most highly with WAI anxiety (r = .77). IIP-B5Low Self-Esteem correlates most highly with WAI Low Self-Esteem (r .75). Each of these three IIP-B5 facetscorrelates highly with the WAI domain of Distress.The IIP-B5 neuroticism facet of Anger Controldemonstrates moderate relations with the subscales of the WAIRestraint dimension. This facet correlates most highly withWAI Suppression of Aggression (r = -.57) and the WAI domainRestraint scale (r = —.56). IIP—B5 Anger Control alsodemonstrates small but relatively significant relations withWAI Anxiety (r = .30) and WAI domain Distress (r = .31).Only three other IIP-B5 facets (excluding circumplexoctant scales) exhibited relatively significant correlationswith WAI scales. IIP-B5 Impulsivity exhibited moderatenegative correlations with WAI Impulse Control (r -.43) andWAI domain Restraint (r = -.37). IIP-B5 Nonconformityexhibited a modest negative correlation with WAI domainDistress (r = -.33). IIP-B5 Reckless exhibited a moderatenegative correlation with WAI Impulse Control (r = -.47), andmodest negative correlations with WAI Low Well-Being(r = -.35) and WAI domain Restraint (r = -.31).124Abbreviations: DEP=depression; ANX=anxiety; LSE=Low Self-Esteem; ANG=anger control; NEUR=neuroticism; EGOC=egocentricUNIM=unimaginative; UNFO=lack of focus; NONC=nonconformity;OPEN=openness; COMP=compulsive; URGEurge control;IMPL=impulsive; UNRE=unreliable; UNCON=unconscientious;RECK=reckless.aunipsati zedbscored with nonconformity unipsatized* p < .05 ** p < .01 p < .001Table 45.MaleMeans and standard deviations of IIP-B5 facet anddomain scales.Female Total GenderM SD M SD M SD TratioDEpa 5.83 4.90 6.54 4.88 6.29 4.89 —1.62ANXa 6.16 4.65 7.19 4.84 6.82 4.80LSEa 5.31 3.95 6.51 4.35 6.09 4.25ANGa 6.22 4.40 6.42 3.91 6.35 4.07 —0.54NEURa 23.25 14.43 26.77 14.27 25.55 14.39UNIM —1.65 4.08 —1.91 3.79 —1.80 3.90 0.71EGOC —1.56 3.33 —2.72 2.88 —2.33 3.09 4.l2***UNFO 0.29 2.84 0.08 2.89 0.14 2.87 0.82NONCa 13.71 3.54 12.92 3.88 13.18 3.78 2.24*OPENb 17.26 7.70 17.60 8.43 17.44 8.21 —0.44COMP 0.07 3.53 0.86 3.88 0.58 3.78LAZY 2.71 3.28 2.85 3.26 2.81 3.27 —0.48URGE 0.56 2.54 0.73 2.56 0.66 2.55 —0.70IMPL 0.02 2.29 —0.16 2.46 —0.09 2.40 0.89UNRE —2.31 2.28 —2.60 2.15 —2.50 2.19 1.50UNCON 0.76 8.59 0.02 9.15 0.29 8.95 0.89RECK —0.83 2.88 —1.70 2.81 —1.40 2.86 3.41***n 201 371 572Nam.______________FiveFactorProfileDomineenngPACD90°Vindve112067°InlrtjsrveInstrument:lip-B5BCNO135045080157023°700Overty________________________IJ.Cold_____________________DEnurturant601800LM500°________0 CD2040203°_____________________030________________020225°0_____________________315°ioFGJK______________Socially247°avoant27002930Exploitable0HINEURUNCONSCOPENNonassertiveScor..PABCDEFGHI.3KLMNORawfIandardIIIIAngb:____________V.clorI.ngth:126Figure 6. IIP-B5 facet profile.LINCONSCIENTIOUSLazy Urges Unreliable Impulsive Compulsive (Reckless)70604020101 IOPENNESS TO EXPERIENCENonconformity Unfocused Unimaginative Egocentric807060cn403020101 I INEUROTICISMDepression Anger Anxiety Low self-esteem807C605040302010127IIP-B5 Scale Norms. Based on the results from Study 2,scale norms were calculated in order to provide standardizedscores for clinical assessment and research purposes. Genderdifferences in endorsement rates were assessed through the T—test. The IIP-B5 norms for this sample can be found in Table45. Only seven IIP—B5 facet and domain scales exhibitedsignificant gender differences in endorsement rates. Womenreported significantly greater rates of anxiety, low self—esteem, and compulsive problems; and they exhibited asignificantly higher mean for the domain Neuroticism scale.Males reported significantly greater rates of egocentric,nonconformity, and reckless problems. There were notsignificant gender differences for the domain scales foropenness and unconscientious problems. These normative datacan be used to generate standardized scores for clinicalassessment and research purposes. Standard profile sheetsfor the IIP—B5 are available and examples can be found inFigures 5 and 6. A standardized score with a mean of 40 anda standard deviation of 10 was selected. This selection wasbased on the readability of the circumplex graph. A mean of40 rather than 50 reduces the number of concentric circles,making the profile easier to read and use.DiscussionThe results of Study 2 support the conclusion thatoperationalization of rigid and maladaptive trait expressionthrough assessment of chronic behavioral inhibitions andchronic behavioral excesses (problems) can be extended from128the interpersonal domain of personality to a morecomprehensive five—factor model of personality traitorganization. The three additional domains of neuroticism,conscientiousness, and openness to experience appear toinclude maladaptive traits and trait—behaviors assessed viathe problems construct.A valid and reliable 140-item inventory, the IIP—B5, wasconstructed to assess rigid and maladaptive trait expressionfrom a FFM framework (64 items assess the interpersonalproblems circumplex; 71 items assess problems related toneuroticism, unconscientiousness, and openness; and 5 itemsassess problems of recklessness). The IIP—B5 demonstratedcross sample stability at the item, facet, and domain level.Additionally, the inventory was constructed using methods toensure its structural fidelity (Loevinger, 1957; Wiggins,1973) with a particular measurement model: The DyadicInteractional perspective on the FFM (Pincus & Wiggins, inpress; Trapnell & Wiggins, 1990; Wiggins & Pincus, 1992, inpress; Wiggins & Trapnell, in press). In so doing, theinventory makes use of circuniplex assessment methodology inaddition to identifying maladaptive trait expressionassociated with the five major personality dimensions.Convergent validity with existing FFM assessmentinventories supports the conclusion that the IIP—B5 assessesthe same FFM dimensions; however, use of the IIP—B5 mayprovide additional advantages for clinical assessment. Inaddition to assessing the major personality dimensions, the129utility of the problems format for both self-report andobserver ratings includes the identification of dysfunctionalbehaviors at the item level which can be targeted forpsychotherapy (Alden & Capreol, 1992; Horowitz, 1991;Horowitz et al, 1989). This directly improves upon the NEOP1 and the adjectival IASR-B5 as clinical assessmentinstruments. Assessment of the FFM with these latterinstruments requires the therapist or clinical researcher toinfer from scale scores the types of dysfunctional trait—behaivors the patient may manifest (Butcher, 1992; Costa &McCrae, l992b). If a therapist chooses to provide feedbackto the patient regarding his/her assessment (i.e., client-centered assessment; McReynolds, 1989), reviewing highlyendorsed items, as well as dimensional results, may improverapport and empathy as well as providing insight. Providingpatient feedback for the MMPI-2 is a complex, multistep task(Butcher, 1990; Lewak, Marks, & Nelson, 1990). Providingpatient feedback for the NEO-PI using the standardized “YourNEO Summary” still requires that the therapist and patientinfer from scale scores, the relations between personalitydimensions and presenting problems.One of the original motives for assessing interpersonalproblems was to provide an instrument to detect patientchange in psychotherapy research that was congruent withtreatment focus (Horowitz, 1979). The lIP and the lIP-C havebeen used to predict positive and negative treatment responseand have demonstrated sensitivity to patient change (Alden &130Capreol, 1992; Horowitz, 1991; Horowitz et al, 1988, 1989;Mohr et al, 1990). The sensitivity to patient change of theadditional facets of the IIP—B5 has yet to be demonstrated,however little evidence exists to contraindicate suchpotential. The IIP-B5 may provide a tool for treatmentevaluation that allows for the efficacy of a broader range oftreatment foci to be assessed.An additional goal in extending the problems constructto the FFM was to develop a taxonomy of maladaptive traitswhich may have significant utility for understanding andinvestigating diagnosis, psychopathology, and psychotherapy.With the exception of problems facets for neuroticism, thenew facets of the IIP—B5 unconcscientiousness, and opennessdomains appear both nonredundant with the facets of the NEOP1 and the HIC’s of the HPI, and unique with regard toconstructs assessed by typical clinical and adjustmentinventories such as the BSI and WAI. External correlationsbetween the majority of unconscientious and openness facetsand these inventories were low or nonsignificant. Newproblems constructs from these domains may be of significantutility for clinical assessment and research. A summary ofthe IIP—B5 facets for neuroticism, unconscientiousness, andopenness can be found in Table 46.The potential for the taxonomy of problems assessed bythe IIP-B5 to provide significant and useful information forclinical assessment requires further empirical investigation.131Table 46. IIP—B5 facet scales for neuroticism, openness, andunconscientiousness.Neuroticism FacetDepression. High scorers report problems of feelingchronically sad, empty, and discouraged, as well asbeing unable to experience positive affect.Anxiety. High scorers report problems of chronic worry,panic, and fearfulness, and indicate chronic symptoms ofautonomic arousal.Low Self-Esteem. High scorers report chronic difficultiesfeeling ambitious, feeling deserving of affection,believing in their abilities, and liking themselves.High scorers chronically perceive themselves as inferiorto others.Anger Control. High scorers report chronic feelings of rage,difficulty controlling their temper, and an inability totolerate disappointment without becoming angry.Openness FacetUnimaginative. High scorers report chronic inability toexperience vivid mental imagery, to think abstractly ortheoretically, to feel creative, and to cathect stronglyto aesthetic stimuli.132Table 46 continuedEgocentric. High scorers report chronic difficultiestolerating customs, philosophies, ideas, actions, andpeople seen as very different from themselves and theirown beliefs.Lack of Focus. High scorers report chronic novelty—seekingbehaviors, inability to adopt a specific belief systemor philosophy of life, and see themselves as tooopen—minded.Nonconformity. High scorers report chronic rebellion againsttraditions, authority, and standard practices. Highscorers report difficulty tolerating routines andpredictable activities.Unconscientious FacetLazy. High scorers report chronic difficulties workingefficiently and persistently, inability to engage innecessary tasks that are of little personal interest tothem, and chronic problems with procrastination.Unreliable. High scorers report chronic difficulty beingpunctual for appointments, paying bills, and returningphone calls, an inability to meet their obligations, anda chronic tendency to break previous commitments.133Table 46 continuedImpulsive. High scorers report difficulties planning ahead,thinking things through thoroughly, and considering theconsequences of their actions.Urge Control. High scorers report chronic engagement inself—defeating and self harmful activities, and aninability to act responsibly toward their personal wellbeing.Compulsive. High scorers report chronic problems withperfectionism, expending too much energy organizing,overconcern with others’ organization and punctuality,and an inability to tolerate disorder or uncleanliness.Additional Clinical FacetReckless. High scorers report chronic sensation seekingbehaviors, chronic involvement in dangerous activities,and a tendency to “get out of control.”With regard to psychotherapy, Miller (1991) proposes a numberof possible characteristics of patients’ presentation andengagement in psychotherapy with regard to their standings onthe dimensions of the FF14. It appears that neuroticismrelates to distressed presentations (e.g., Costa & McCrae,1987; Watson & Pennebaker, 1989). Miller suggests high N134patients present with a variety of painful feelings whichmotivate them to seek relief. Clearly the IIP-B5 neuroticproblem facets assess aspects psychological distress commonto many patients’ presentations.Of interest, Miller’s (1991) theoretical proposalsregarding the effects of conscientiousness and openness onpatient presentation and treatment mention a number ofcharacteristics which are assessed directly by the IIP—B5.He suggests that low openness patients may seem “unable tofantasize or symbolize, their speech seems boring, pedantic,and overly—conventional; and they do not easily understand oraccept elementary psychodynamic interpretations” (p. 425).Miller (1991) also suggests that high openness patients mayhave more vivid and emotional recollections of past events,are more capable of vivid internal imagery, use moremetaphorical and analogical speech, lead more varied andunconventional life styles, and are more willing to try newways of thinking or relating to others. He suggests thathigh openness patients will prefer imaginative approaches,but their “excessive curiosity can scatter resources” (p.418). Low openness patients will respond well to practicalapproaches like education, support, and behavior therapy.The qualities of openness Miller suggests are relevant topsychotherapy seem well represented by the IIP-B5 Opennessfacets of Unimaginative, Egocentric, Nonconformity, and Lackof Focus.135Miller (1991) proposes that conscientiousness influencespatients’ willingness to do the work of psychotherapy. He isclearly convinced that low conscientiousness(unconscientiousness) “might represent one of the absolutelimits to the power of psychotherapy”(p.430). Millersuggests that the key problems of unconscientious patientsare low achievement, impulsivity, and half—hearted problemsolving. Miller suggests there may be no treatmentopportunities with very unconscientious patients and predictsthey are “unlikely to do homework, and are likely to rejectinterventions that require hard work or toleration ofdiscomfort” (p. 419). He presents a brief description of apatient: “A woman who has hated herself for years becauseshe is overweight is encouraged to keep an eating diary andcalculate her daily caloric intake. Despite continuingencouragement, she never buys a calorie counter and neverrecords any of her meals. Her explanation is she is afraidshe will be upset if she learns how much she really eats. Weagree that it might be a good thing if she got upset abouther eating habits. She continues to claim that low self-esteem due to obesity is her main problem, and she nevercomplies with the plan” (p. 430). The qualities of thispsychotherapy patient appear quite similar to the IIP-B5unconscientious problems facets of Lazy, Impulsive, UrgeControl, and Unreliable.With regard to psychopathology, research regarding theinfluence of personality traits on Axis I disorders is in its136infancy (see Widiger & Trull, in press). The majority of theinvestigations involving the FFM have focussed on thepersonality disorders (e.g., Costa & McCrae, 1990; Widiger,in press; Wiggins & Pincus, 1989, in press). For example,the problem domain of unconscientiousness may be particularlyrelevant to a number of personality disorders including thecompulsive, passive—aggressive, sadistic, and antisocialdisorders (Hyler & Lyons, 1988; Shapiro, 1989; Widiger, inpress, Widiger & Trull, in press; Wiggins & Pincus, 1989, inpress).While the current work in the field of personalitystructure appears to attest to the comprehensiveness of theFFM for descriptions of normal personality (e.g., Wiggins &Pincus, 1992), the potential that a FFM of even maladaptivepersonality traits (problems) may not be a comprehensivetaxonomy for abnormal personality is of serious concern.Butcher (1992) suggests that clinical personality assessmentalso deals with behavioral extremities that may be a class inthemselves, widely separated from behaviors in the normalrange. This claim is not arguable. Many behaviors andcharacteristics of patients presenting with severepsychopathology may fall outside the FFM and are not relatedto normal personality traits. One can simply imagine theundifferentiated schizophrenic patient exhibiting significantconfusion and positive psychotic symptoms. What dimensionsof personality shall his/her delusions, hallucinations,catatonic postures, and cognitive disorganization be subsumed137by? The fact is that all psychopathology is not pathology ofpersonality and no claim should be made that this be thecase.However, there is a more direct challenge to thecomprehensiveness of the FFM to account for all themanifestations of abnormal personality seen in clinicalassessment. A research program which focussed on thedimensions of personality pathology (Livesley, 1986, 1987;Livesley, Jackson, & Schroeder, 1989) rather than normalpersonality could provide support or challenge to thecomprehensiveness of the FFM for clinical personalityassessment.Livesley and his colleagues developed the DimensionalAssessment of Personality Pathology—Basic Questionnaire(DAPP-BQ) based on descriptive features of the DSM-IIIpersonality disorders. The basic dimensions of personalitypathology which emerged from their research program were:Affective Lability, Anxiousness, Cognitive Distortion,Compulsivity, Conduct Problems, Diffidence, IdentityProblems, Insecure Attachment, Interpersonal Disesteem,Intimacy Problems, Narcissism, Passive Oppositionality,Rejection, Restricted Expression, Self—Harm, SocialAvoidance, Stimulus Seeking, and Suspiciousness. Each ofthese broad dimensions is assessed by multiple subscales ofthe DAPP-BQ.Schroeder et al (1992) conjointly factored 16 of the 18DAPP-BQ dimensions and the NEO-PI. Additionally, they138multiply regressed the NEO-PI domain scales onto each of theDAPP-BQ dimensions. The results of their principalcomponents analysis indicated that four dimensions of the FFM(neuroticism, conscientiousness, agreeableness, andextraversion) marked components on which DAPP—BQ dimensionsstrongly loaded. Openness was not strongly related todimensions of personality pathology. Additionally, multiple-correlation analyses showed that four DAPP—BQ dimensions(Intimacy Problems, Conduct Problems, Restricted Expression,and Insecure Attachment) had little shared variance with NEOP1 factors. The majority of results suggest relativeconvergence between the DAPP-BQ and the NEO-PI. The authorsconclude, “The results of this study largely confirm ourexpectations that these dimensions of personality disorderare closely related to the Big Five factors of normalpersonality” (p. 52).It appears that even when the dimensions of personalitypathology are derived and their relations with the FFMassessed, the comprehensiveness of the FFH is supported. Itis also clear that at a lower level of the constructhierarchy, many of the 18 DAPP-BQ dimensions and 100subscales do not have construct counterparts in any FFMassessment instrument. Two DAPP—BQ dimensions, Self—Harm andCognitive Distortion, were deleted from the Schroeder et al(1992) analyses. These two dimensions may not relatestrongly to the FFM. Additionally, the IIP-B5 facets139assessing problems may have different relations withdimensions of personality pathology than the NEO-PI.However, the IIP—B5 was constructed specifically toassess maladaptive and rigid expression of traits subsumed bythe FFM. It was not constructed to necessarily assess alldimensions of personality pathology, nor was it constructedto assess all classes of behavior found in clinicalpersonality assessment. The IIP—B5 does show convergencewith many aspects of these domains. It is an inventory whichmay be useful as a component of clinical assessment for allclasses of psychopathology and all individuals’ who seekpsychotherapy (many of whom do not meet criteria for a DSMdiagnosis). Assessment of a FFM of problems with the IIP-B5may be useful for client—centered assessment, diagnosis,treatment formulation, and clinical research focussing onboth psychopathology and psychotherapy.140CHAPTER 4Study 3: Psychiatric ValidationMethodsOverviewIn order to determine the full applicability of the lIPB5 for psychological assessment of clinical populations, itis necessary to determine the psychometric and structuralcharacteristics of the instrument in an initial psychiatricsample. The major goal of Study 3 was to assess thesecharacteristics in a diverse sample of psychiatricoutpatients.Sample. The IIP-B5 and the NEO-PI were administered toa sample of 72 consecutive psychiatric outpatients assessedat University Hospital-University of British Columbia Site aspart of a more detailed clinical assessment. The goal ofthis assessment was to determine which of the variety ofoutpatient psychiatric programs was most suitable for aparticular patient. The present sample consisted of 41 women(58.5%) with an average age of 33.69 years and 31 (42.5%) menwith an average age of 28.25 years. The diagnostic breakdownof the sample can be found in Table 47.Table 47. Diagnostic breakdown of psychiatric sample.Primary Diagnosis Percent (%)Dysthymic Disorder 34.7%Major Depression 19.4%Adjustment Disorder 9.7%141Table 46 continuedPanic Disorder 5.5%Dependent Personality Disorder 5.5%Borderline Personality Disorder 5.5%Obsessive—Compulsive Disorder 4. 2%Narcissistic Personality Disorder 2.7%Mixed Personality Disorder 2.7%Avoidant Personality Disorder 1.4%Social Phobia 1.4%Post Traumatic Stress Disorder 1.4%Generalized Anxiety Disorder 1.4%No DSM-III-R Diagnosis 4.2%Note. N = 72.Instruments. The final version of the IIP-B5 asdetailed in the previous chapters was administered as part ofa large self—report battery. Some instruments pertained toindependent research projects and are not discussed here.The NEO-PI (Costa & McCrae, 1985) is a 181-iteminventory that yields measures of the Big Five factors ofnormal personality. The instrument’s Neuroticism,Extraversion, and Openness factor items can also be summed toyield six facet measures for each factor, thus providing amore detailed, representation of the dimensions.Procedures. All patients referred to outpatientpsychiatry services are assessed with clinical interviews142carried out by psychiatric residents under supervision.Additional self-report instruments are completed and allinformation is used to determine the appropriate outpatientservice for the referred patient. The NEO—PI is a standardcomponent of the assessment battery. The IIP—B5 was includedfor a period of two months. During this period, 72outpatients completed both self—report instruments and aclinical interview. These patients were assigned to variousoutpatient services.AnalysesData Transformation. The ipsatization proceduredescribed in Study 1 was again used to transform the data.Structural Validation. The final IIP-B5 facetsvalidated in Study 2 were subjected to a conjoint principalcomponents analysis with markers of neuroticism,conscientiousness, and openness from the NEO—PI. Followingthe same step-wise procedure detailed in Studies 1 and 2, thefacets were then conjointly factored with the liP-C octantscales. Following confirmation of the lower—level structuralstability of the IIP-B5 facets, the domain scales wereconjointly factored with the NEO-PI domain scales.Reliability and Validity. Alpha coefficients and itemtotal correlations for derived facets and domain scales weredetermined to assess internal consistency. Convergentvalidity correlations for the three new problems dimensionswere obtained by examining the relations between derivedproblems scales and the neuroticism, conscientiousness, and143openness scales of the NEO—PI concurrently assessed in thesample. Additional validity correlations were obtained byexamining the relations between the IIP-B5 facets and theNEO-PI facets.Expectations1. The structure of the IIP-B5 will replicate in asample of psychiatric outpatients.2. The reliability of some problems facets will improvedue to greater variance in the psychiatric sample.3. The convergent validity of the IIP-B5 will besupported.ResultsStructural Validation. In order to demonstratestructural stability of the IIP—B5 across samples, the 13derived problems facets were subjected to a conjointprincipal components analysis with domain scales from theNEO-PI. Examination of the Scree Plot indicated a threefactor solution accounting for 53.2% of the variance whichwas rotated to varimax criterion. The final superordinatestructure of the derived problems facets can be seen in Table48. Only one IIP-B5 facet (Lack of Focus) departed from theexpected pattern of factor loadings. This facet loaded mosthighly on the Neuroticism dimension (-.39) rather than on theOpenness dimension. Examination of the Pearson correlationsbetween this facet and the NEO-PI domain scores indicatedthat in this sample of 72 psychiatric outpatients, Lack ofFocus correlated (r —.30, P < .01) with Neuroticism and144Table 48. Three component solution of conjoint principalcomponents analysis of IIP—B5 facets and domainscales of the IASR-B5.FactorI U InScaleNeuroticism (NEO-PI) .92 .13 —.01Depressiona—.08 .04Anxietya 84 -.12— .08Low Seif_esteema .04 .07Angera.60 .13 -.09Lack of Focusb -.39 -.25 .17Conscientious (NEO—PI) —.09 —.89 .05Unreliable -.04 .64 .21Lazy —.16 .05Compulsive — .13 —.59 —.09Impulsive .06 -.22Urge Control .19 .29Openness (NEO-PI) -.24 -.01 .85Unimaginative -.02 .12 -.75Egocentrism -.06 -.15 -.54Nonconformitya-.42 .25Note. N = 72. aunipsatized facet scales. bunexpected loadingon nontarget dimension.145(r = .17, ns) with Openness.In the validation sample described in Study 2, Lack ofFocus exhibited nonsignificant correlations with IASR-B5Neuroticism (r = .02) and Conscientiousness (r = .06) whilecorrelating significantly with IASR—B5 Openness (r = .22, p <.001). A similar pattern of correlations was exhibited whenneuroticism, conscientiousness, and openness were assessed bythe FF1. This unexpected factor loading suggests the Lack ofFocus facet is a less robust component of the IIP—B5structure. The facet may require further development toimprove its relation with trait openness. However, thesubstantial difference in sample size between the derivationsample (n = 706), validation sample (n 572), and thepsychiatric sample (n = 72) was considered and the Lack ofFocus facet was retained on Openness in the remaininganalyses. The remaining pattern of factor loadings seen inTable 48 support the structural stability of the IIP—B5across samples.As a second test of the superordinate facet structure ofthe IIP—B5 scales, they were combined with the lIP-C octantscales and conjointly factored with markers of neuroticism,conscientiousness, and openness from the NEO—PI. A principalcomponents analysis indicated a clear five factor solutionemerged that accounted for 61.5% of the variance. Thesolution was rotated to variinax criterion and can be seen inTable 49. The full facet/octant structure of the IIP-B5 from146Table 49. Full facet and circumplex octant principalcomponents analysis of the IIP—B5.FactorsI II III IV VScalesNeuroticism (NEO—PI) .91 .19 .14 —.02 -.03Depressiona—.02 —.04 —.16 —.01Anxietya.11 -.08 -.05 -.11Low Seif_esteema .81 —.10 .09 —.27 .03Angera.51 .09 .10 - .07liP—C (JK) —.09 —.86 —.14 .07 .02liP—C (HI) .02 —.77 .01 —.30 —.07lIP—C (PA) .04 7fl —.11 .34 .16lIP—C (LM) —.20 —.64 —.37 .27 .06liP—C (BC) .19 .48 .03 —.36 —.35Conscientious (NEO—PI) —.06 .00 —.88 —.04 .11Lazy —.20 .05 .02 .03Unreliable .03 —.33 .62 .29 .07Compulsive —.06 —.09 —.59 .25 —.13Impulsive .09 .17 .29 —.24Urge Control .18 .12 .03 .31liP—C (FG) .14 —.08 —.07 —.82 .03lIP—C (DE) .01 .17 —.05 —.76 .00liP—C (NO) —.11 .25 .09 .12Lack of Focusb -.29 .03 -.29 .44 .18147Table 48 continuedOpenness (NEO—PI) —.19 —.15 .01 .10Unimaginative —.05 —.05 .07 .02 —.77Nonconformitya—.42 .20 .27 .06 .53Egocentric —.16 .40 —.16 —.12Note. N = 72. aunipsatized scales. bunexpected loading onnontarget dimension.Studies 1 and 2 were replicated in this analysis, with theexception of the Lack of Focus facet, which loadedanomalously on the dimension marked by liP-C octants. Whilethe anomalous loading is not unexpected given the results ofthe first analysis in Table 48; the highest loading of thefacet is on an interpersonal dimension marked by theIntrusive scale (.44) rather than neuroticism (—.29). Therobustness of the structural relations of the Lack of Focusfacet is challenged by this result and further work isrequired.The remaining facet scales loaded most highly on theirtarget dimensions, although a number of facets show moderatesecondary loadings. Both the Anger Control facet(neuroticism) and the Egocentric facet (openness) havemoderate secondary loadings on an interpersonal dimensionmarked by the lIP-C Domineering and lIP-C Vindictive scales.The Nonconformity facet has a secondary loading on theneuroticism dimension (as expected).148Table 50. Conjoint principal components analysis of NEO-PIdomain scales and IIP—B5 domain scales.FactorsI II III IV VScalesNeuroticism (NEO—PI) 94 —.11 .12 —.13 .09Neuroticism (IIP—B5) .93 —.16 .00 —.19 .02Openness (NEO—PI) —.08 .02 .16 —.13Openness (IIP—B5) —.15 .03 .08 .08Conscientious (NEO—PI) —.04 .09 —.94 .09 —.02Unconscientious (IIP-B5) .07 .13 .93 .03 .06Nurturance (IIP-B5) -.17 .14 -.01 .92 -.07Agreeableness (NEO—PI) -.36 .11 -.13 -.57Extraversion (NEO—PI) -.16 .55 —.05 .61 .40Dominance (IIP-B5) .05 .01 .05 —.01 .95Note. N = 72.It is not possible to directly validate the structuralfidelity of the IIP-B5 with the Dyadic-Interactionalperspective on the FFM using a conjoint principal componentsanalysis of the IIP-B5 domain scales with the FFM domainscales from the NEO—PI. These two instruments operationalizedifferent structural perspectives on the FFM. Because of thestructural relationships between NEO—PI Extraversion andAgreeableness and the interpersonal circumplex, splitloadings for these conjointly factored interpersonal scaleswas expected. A clear five factor solution accounting for14988.4% of the variance was rotated to a varimax criterion.This solution is presented in Table 50.Examination of Table 50 indicates excellent convergentstructure for the new IIP—B5 scales assessing problemsrelated to neuroticism, unconscientiousness, and openness.As expected the NEO—PI Extraversion and Agreeableness scalesexhibited split loadings on the interpersonal dimensionsmarked by dominance and nurturance. Extraversion exhibited amoderate loading on the openness dimension. In light of theanomalous results of the IIP—B5 Lack of Focus facet, thePearson correlations among the IIP—B5 openness facets anddomain scale and the NEO—PI domain scales were examined.These correlations can be found in Table 51.Table 51. Intercorrelations among IIP—B5 openness facets anddomain scales and domain scales of the NEO-PI in apsychiatric sample.NEO-PI ScaleE 0 N C AIIP-B5 OPENNESSUnimaginative —.26 _.58* .03 —.09 —.03Nonconformity 43* _.31* -.08 .00Egocentric —.19 _.36* —.05 .05 —.21Lack of Focus .17 _.30* .17 .12Openness 53* —.26 .06 .14Note. N = 72. *p < .01.150Examination of Table 51 indicates that in this smallpsychiatric sample, the convergent validity of the IIP—85openness problems domain is only moderately supported. TheNonconformity and Lack of Focus facets exhibit significantcorrelations with trait extraversion and trait neuroticism.The IIP—B5 Openness domain is significantly correlated withtrait extraversion. These results are congruent with themoderate secondary loading exhibited by NEO-PI Extraversionon the Openness dimension in Table 50.Reliability. Cronbach’s alpha was calculated for allIIP—B5 facet and domain scales for neuroticism,conscientiousness, and openness. Reliability coefficientsare presented in Table 52. Alpha coefficients for all scalesremained stable in this psychiatric sample, and a number ofscales internal consistency demonstrated modest gains.Neuroticism problem facets and domain alpha coefficients arevirtually identical across samples. The Unreliable facetalpha rose from .53 to .74, as was predicted. The variancein endorsement of unreliable problems increases in apsychiatric sample relative to a university sample.Similarly, the Urge Control facet alpha rose from .55 to .82,indicating significant improvement in a psychiatric sample.The Impulsivity facet alpha unexpectedly dropped from .70 to.64 in this sample. It is possible that theoverrepresentative proportion of dysthymic and depressedpatients in the sample reduced the variance in endorsement ofthe facet items. The remaining unconscientiousness facets’151and domain alpha coefficients remained stable across samples.All Openness facets demonstrated moderate increases in alphaTable 52. Alpha coefficients for IIP-B5 facet and domainscales in a psychiatric sample.AlphaScaleDepression .84Anxiety .89Low Self-esteem .84Anger .86Neuroticism (IIP-B5) .92Lazy .70Unreliable .74Impulsive .64Urge Control .82Compulsive .84Unconscientiousness (IIP-B5) .70Unimaginative .86Egocentric .88Lack of Focus .82Nonconformity .68Openness (IIP—B5) .89Reckless .78Note. N = 72.152in this psychiatric sample, contributing to a significantimprovement in internal consistency for the Openness domainfrom .79 in a university sample to .89 in this sample. TheLack of Focus facet alpha increased from .65 in a universitysample to .82 in this sample. While the structural relationsof this facet were called into question, the internalconsistency of the facet is excellent, as is the Opennessdomain when, in fact, the Lack of Focus facet is included.The Reckless facet also exhibited an increased alphacoefficient, rising from .71 in a university sample to .78 inthis sample.Convergent Validity. The IIP-B5 domain scales werecorrelated with the domain scales of the NEO-PI to assessconvergent validity. These correlations can be found inTable 53. Additional validity correlations between IIP-B5facets and the NEO—PI facets for neuroticism and openness canbe found in Tables 54 and 55.Table 53. Correlations between IIP—B5 domain scales anddomain scales of the NEO-PI.NEO-PI Domain ScalesIIIIP—B5 Scales* * *Neuroticism .88 —.27 — .09 —.36 —.45Openness —.26 .63* .06 •53* .14Unconscientious .16 .11 _77* .06 —.14Note. N = 72. *r p < .01.153Table 54. Correlations between IIP-B5 neuroticism facetsand NEO-PI neuroticisiu facets.NEO-PI Neuroticism FacetsANX HOS DEP SELFCON IMPL VULIIP-B5 FacetsAnxiety .42 .73 .58 .66Depression .62 .47 Q .60 .58Low Self—esteem .61 .33 .82 .58 .57Anger Control .50 .46 .44 .59Note. N = 72. p > .01 deleted. Abbreviations: ANX=anxiety;DEP=depress ion; HOS=hostility; SELFCON=self consciousness;IMPL=impulsive; VUL=vulnerability.Table 55. Correlations between IIP—B5 openness facets andNEO-PI openness facets.NEO-PI Openness FacetsFAN AES FEEL ACT IDEAS VALUESIIP—B5 FacetsUnimaginative -.35 -.61 -.36 - .49Egocentric -.47Nonconformity .36 .34Lack of FocusNote. N = 72. p > .01 deleted. Abbreviations: FAN=fantasy;AES=aesthetics; FEEL=feelings; ACT=actions.Examination of Table 53 indicates that the domain scalesof the IIP-B5 exhibit good convergent validity with domain154scales of the NEO—PI. IIP—B5 Unconscientiousness correlatessignificantly (r = -.77) only with NEO-PI Conscientiousness.IIP-B5 Neuroticism correlates significantly (r = .88) withNEO-PI Neuroticism and exhibits secondary significantcorrelations with NEO—PI Agreeableness (r = -.45) and NEO-PIExtraversion (r = —.36). IIP—B5 Openness correlates mosthighly with NEO-PI Openness (r = .63) and exhibits a secondsignificant correlation with NEO-PI Extraversion (r = .53).This last result is most likely the result of the IIP-B5 Lackof Focus facet correlating significantly with traitextraversion in this sample.Examination of Table 54 indicates good convergentvalidity between IIP-B5 neuroticism facets and NEO-PIneuroticism facets. IIP—B5 Anxiety and Depression correlatemost highly with their NEO-PI counterparts. IIP-B5 Low Self-esteem correlates most highly with NEO-PI Depression. IIP-B5Anger Control correlates most highly with NEO-PI Hostilityand is the only neurotic problems facet to correlatesignificantly with NEO-PI Impulsivity.Examination of Table 55 indicates good convergentvalidity between IIP-B5 openness facets and NEO-PI opennessfacets. IIP-B5 Unimaginative correlates most highly withNEO-PI Aesthetics (r = -.61) and exhibits moderate negativecorrelations with NEO-PI Fantasy, Feelings, and Ideas facets.IIP-B5 Egocentric correlates significantly with NEO-PI Values(r = -.47). IIP-B5 Nonconformity exhibits significant155positive correlations with NEO-PI Aesthetics, Ideas, andValues.Ben-Porath and Wailer (1992b) note that the FFMdimension of conscientiousness and Tellegen’s (1985)constraint dimension of his three dimensional model bothimply control. They are critical of the NEO-PI for includinga facet for impulsivity (the opposite of control) in thedomain of neuroticism. In the current sample, the only liP—B5 neuroticism facet to correlate with NEO-PI Impulsivity wasAnger Control. IIP-B5 Impulsivity (a facet located in theunconscientious problems domain) does correlate significantlywith NEO-PI Impulsivity (r = .40). IIP-B5 Impulsivityexhibits no other significant correlations with any NEO-PIneuroticism facets. However, it does correlate significantlywith the NEO-PI Conscientiousness scale (r = -.35). IIP-B5Impulsivity exhibits no other significant correlations withdimensions of the FFM assessed by the NEO—PI. These resultssupport the IIP-B5, Ben-Porath and Waller (1992b), andTellegen (1985) in conceptualizing impulsivity as a facet of(un)conscientiousness.DiscussionThe psychometric and structural properties of the IIP-B5were examined in a small (n =72) sample of psychiatricoutpatients. Results indicated that the general structure ofthe inventory at the facet and domain levels of the FFMconstruct hierarchy, derived and validated in universitysamples, was replicated in a psychiatric sample. Internal156consistency of problems facets and domain scales improvedoverall in the psychiatric sample as expected. In order toevaluate the structural stability of the IIP-B5, a number oflower and higher order principal components analyses wereconducted. Those results, and examination of convergent anddivergent validity, provide both support for the stablestructure of the IIP—B5 and identify areas where structuraland substantive improvement may be required. The overallsupport for structural stability of the IIP-B5 across bothnormal and psychiatric samples suggests that the dimensionaltaxonomy of normal personality traits referred to as the “BigFive” has important relations to the domain of maladaptive,abnormal, or psychopathological behavior. While not allbehavior seen in clinical personality assessment is subsumedby the FFM (Ben—Porath & Wailer, 1992a; Butcher, 1992),results of this research support a quantitative, dimensionalrelationship between normal and abnormal personality as aviable alternative to a categorical approach to clinicalpersonality assessment which often assumes a qualitativedifference between normal personality and pathologicalpersonality features (see Carson, 1991; Livesley, 1991;Millon, 1991; Gunderson, Links, & Reich, 1991; and Widiger,in press, for extended discussions of this issue).Neuroticism. The IIP—B5 neuroticism facets demonstratestable structure and excellent internal consistency acrossthe derivation, validation, and psychiatric samples.Convergent validity with a number of personality and clinical157inventories suggests that problems of sad affect, low self—esteem, anxiety, and anger are important aspects of behaviorto assess in both clinical and normal personality assessment.The IIP-B5 Anger facet is the only neuroticism facet toexhibit significant relations with other FFM dimensions.This valid and reliable facet demonstrates moderate relationsto the interpersonal dimensions of the FFM. Costa, NcCrae,and Dembroski (1989) distinguish between “neurotic hostility”which is an affective personality trait and “antagonistichostility” which is an interpersonal personality trait.Substantive improvement in the IIP-B5 Anger facet couldinvolve a reduction in interpersonal content by modifying oradding new items.Unconscientiousness. The IIP—B5 unconscientious facetsexhibit stable structure and good internal consistency acrossthe derivation, validation, and psychiatric samples.Examination of convergent validity with typical clinicalmeasures yielded few significant relations. It appears thatboth trait conscientiousness and problems ofunconscientiousness are poorly represented in a number ofclinical inventories, such as the BSI and WAI examined inthis research program, and the MNPI (Ben-Porath & Waller,1992b; Costa et al, 1986; Johnson, Butcher, Null, & Johnson,1984). Given the significant relations between measures ofconscientiousness and conceptions of personality disorders(e.g., Costa & McCrae, 1990; Schroeder, et al, 1992; Wiggins& Pincus, 1989) and the importance of individual differences158in conscientiousness for engagement and efficacy ofpsychotherapy (e.g., Miller, 1991; Shapiro, 1989), problemsof compulsivity, unreliability, laziness, urge control, andimpulsivity may be important aspects of behavior to assess inboth clinical and normal personality assessment.The facet of impulsivity assesses a lack of pianfulnessand attention to the consequences of one’s behavior. Itappears related to the domain of unconscientious problems assuggested by Ben-Porath and Wailer (1992b) and Tellegen(1985). The psychometric and structural properties of theIIP—B5 impulsivity facet remained stable across three samplesand the results clearly support its inclusion in theunconscientious domain, rather than the neuroticism domain assuggested by Costa and McCrae (1985). The assessment ofproblems of unconscientiousness are unique to the IIP—B5 andmay provide substantial incremental gain in understanding andeffectively treating individual patients.Openness. The fifth factor of the FFM is Openness toExperience. This factor has demonstrated a long history oftaxonomic debate among FFM investigators (e.g., Peabody &Goldberg, 1989; Wiggins & Trapnell, in press). It has beenreferred to as Openness (McCrae & Costa, 1985b),Intellectance (Hogan, 1986), Culture (Norman, 1963), andIntellect or Sophistication (Goldberg, 1992). It is mostoften the last factor extracted in structural investigationsof the FFM (e.g., Goldberg, 1990; Trapnell & Wiggins, 1990;Wiggins & Pincus, 1989). Of all problems domains assessed by159the IIP—B5, openness has demonstrated the weakest structuralstability across samples, although the internal consistencyof the facets is stable and excellent.The Unimaginative and Egocentric facets clearly assessproblems of being too closed to experience and these facetsremained structurally stable across the derivation,validation, and psychiatric samples. The convergent validityof these facets was also supported. The facets constructedto assess problems of being too open (Nonconformity and Lackof Focus) exhibit the weakest evidence for both structuralstability and discriminant validity. Nonconformityconsistently split component loadings across openness (+) andneuroticism (—). Lack of Focus deviated from a consistentloading on openness in the derivation and validation samplesto a pattern indicative of a strong relation to traitextraversion in the psychiatric sample. A possible confoundin the substantive interpretation of this facet is thedifficulty discriminating between problems of being toocurious and novelty—seeking (openness) and being toostimulus—seeking (recklessness, sensation—seeking). The fullNEO-PI used in the psychiatric sample includes an ExcitementSeeking facet in assessing domain Extraversion; and,Schroeder et al (1992) demonstrated that the NEO-PIExtraversion scale “showed a strong positive relationship toStimulus Seeking and played a lesser role in the predictionof nine other DAPP-BQ scales” (p. 52). The significantcorrelation between IIP—B5 Openness and NEO—PI Extraversion160suggests that the Lack of Focus facet (and to a lesser extentthe Nonconformity facet) may require substantive fine tuningvia item modifications and or additions to more accuratelyoperationalize problems related to being too open. Miller(1991) suggests several potential openness problems effectingpsychotherapy engagement and outcome. Further investigationsto expand the taxonomy of maladaptive traits expressingopenness is required.161CHAPTER 5An application of the IIP-B5: The Weinberger adjustmenttypology and the five-factor model of personalityINTRODUCTIONThe Weinberger Adjustment Typology. Weinberger and hiscolleagues (Weinberger, 1989; Weinberger, in press;Weinberger, Feldman, & Ford, 1989; Weinberger & Schwartz,1990; Weinberger, Tublin, Feldman, & Ford, in press) haveproposed a typological model of social—emotional adjustmentbased on the superordinate dimensions of distress andrestraint. Weinberger and Schwartz (1990) define distress as“a general measure of individuals’ tendencies to feeldissatisfied with themselves and their ability to achievedesired outcomes. Proneness to anxiety, depression, lowself—esteem, and low well—being are operationally defined assubtypes of distress” (p. 382). They define restraint as“encompassing domains related to socialization and self—control and refers to suppression of egoistic desires in theinterest of long—term goals and relations with others. Thusrestraint is superordinate to tendencies to inhibitaggressive behavior, to exercise impulse control, to actresponsibly, and to be considerate of others” (p. 382).These superordinate and lower order constructs can beassessed by the Weinberger Adjustment Inventory (WAI;Weinberger, 1989).162In their discussion of distress and restraint,Weinberger and Schwartz (1990) propose the followingrelations between the WAI and the FFM. Distress is mostclosely related to trait neuroticism but differs due to theinclusion of a Low Well-being subscale, which they suggest ismost strongly related to trait extraversion. Restraintencompasses agreeableness (WAI subscales for Suppression ofAggression and Consideration of Others) and conscientiousness(WAI subscales for Impulse Control and Responsibility). Allempirical investigations of these hypothesized relations haveinvolved a piecemeal operationalization of FFM dimensions(e.g., Weinberger et al, 1989). The relations between theWAI and a fully operationalized FFM have not been reported.A summary of these hypotheses can be found in Table 56.Weinberger and Schwartz (1990) propose a six celltypology of adjustment styles by crossing restraint(high/moderate/low) X Distress (high/low). They argueagainst the claim that typologies derived from theintersection of dimensions cannot claim to represent“distinct forms” unless significant interaction terms (e.g.,A X B) emerge in the prediction of external variables (Hicks,1984; Mendelsohn, Weiss, & Feimer, 1982). They argue thattraditional analysis of variance interaction terms are mostsensitive to complementary, disordinal patterns, such ascross—overs. Because such techniques are used to describefactors rather than groups, they do not identify nonadditivepatterns among between—groups cells. Weinberger and Schwartz163Table 56. Weinberger and Schwartz’ (1990) hypothesizedrelations between WAI scales and the Big Five.Big Five DimensionsWAIScale N E Q A CAnxiety (+)Depression (+)Low Self-esteem (+)Low Well-being (-)DISTRESS (+) (—)Suppression of Aggression (+)Consideration (+)Impulse Control (+)Responsibility (+)RESTRAINT (+) (+)Note. Abbreviations: N = neuroticism; E = extraversion; 0 =openness; A = agreeableness; C = conscientiousness(1990) argue that “asymmetrical variance is assigned to maineffects, even when group differences are completelyattributable to one cell” (p. 387). They argue that thevalidity of their adjustment typology can be supported by theidentification of nonadditive group differences that departfrom what could be predicted from a strictly dimensionalanalysis of distress and restraint or the use of traditionalinteraction terms.164To assess the validity of their typology, 28 dependentmeasures were assessed and oneway multivariate analyses ofvariance (MANOVA) were performed to ensure reliable groupdifferences in each domain. Following this, the post-hocNewman—Keuls procedure was employed to take into account thenumber of paired comparisons involved. If patterns of pairedcomparisons between adjustment style groups differed fromlinear contrasts indicating main effects for distress,restraint, or a combination of the two, Weinberger andSchwartz (1990) argue that these nonadditive differencesidentify distinctive features of individual groups, hencetheir typology would be validated.Their results indicated that for 26 of their 28dependent variables, main effects for distress, restraint, orboth were significant. The Distress X Restraint interactionwas not significant. The Newman—Keuls comparisons identified21 out of 28 dependent variables where group differences weresignificant and nonadditive. The authors concluded,“prototypic members of each of the six adjustment groups havedistinct personality structures that cannot be fullyexplicated by dimensional analyses” (p. 407). The adjustmenttypology consists of the following types: Reactive (highdistress/low restraint), Sensitized (high distress/moderaterestraint), Oversocialized (high distress/high restraint),Undersocialized (low distress/low restraint), Self—assured(low distress/moderate restraint), and Repressive (lowdistress/high restraint) (see Table 57).165Table 57. The Weinberger adjustment typology.RestraintHigh Moderate LowDistressHigh Oversocialized Sensitized ReactiveLow Repressive Self—assured UndersocializedAdjustment and the Big Five. A major issue ofimportance can be discussed regarding the WAI typology. Therelations between the WAI Distress and Restraint scales andsubscales were proposed to subsume four of the five majordimensions of personality. In doing so, it is unclearwhether Weinberger is asserting that distress and restraintare superordinate to the FFM or whether he has chosen toassess blends of the FFM dimensions. If the former is true,conjoint factor analytic investigations of the WAI and aninstrument operationalizing the FFM would help determine thehierarchical relations between the dimensions of adjustmentand the dimensions of the FFM. If the latter is true,Weinberger is open to the criticism that he has notdemonstrated incremental gain in assessing a particular blendof dimensions which fifty years of personality structureresearch has cumulatively suggested are superordinate andbasic (Costa & McCrae, in press; Digman, 1990; Wiggins, inpress; Wiggins & Trapnell, in press).166Additionally, if the dimensions of distress andrestraint are substantively different from the dimensions ofthe FFM and the typological classifications are valid, itshould be possible to replicate significant adjustment groupdifferences using the dimensions of the FFM as dependentvariables.MethodsSample. A subset of 307 subjects from the sample of 572university students described in Study 2 make up the samplefor the following investigation.Instruments. The Inventory of Interpersonal Problems—Big Five Version (IIP-B5) is a 140 item inventory constructedto assess rigid and maladaptive trait expression from a FFMframework. Two types of items are included: Behaviors onedoes too much (excesses) and behaviors one finds chronicallyhard to do (inhibitions). Subjects rate how much of aproblem each statement has been for them on a 5-point Likertscale ranging from “not at all” to “extremely.” The IIP—B5assesses the interpersonal problems circumplex (Alden et al,1990) and additional problems related to the trait domains ofneuroticism, unconscientiousness, and openness. The IIP—B5possesses good reliability and has demonstrated homologousstructure with the IASR-B5 (Trapnell & Wiggins, 1990). Thisinstrument operationalizes the Dyadic—Interactionalperspective on the FFM.The Weinberger Adjustment Inventory (WAI; Weinberger &Schwartz, 1990) is an 84 item inventory designed to assess167long term functioning rather than short term symptoms. Thetwo primary dimensions are “Distress” (as assessed bysubscales measuring anxiety, depression, low self—esteem, andlow well—being) and “Restraint” (as assessed by subscalesmeasuring impulse control, suppression of aggression,consideration of others, and responsibility). There are twovalidity indicators consisting of a “Denial of Distress”scale, which refers to defensiveness about normativeexperiences of distress, and a “Repression” scale whichrefers to claims of nearly absolute restraint. Subjectsendorse items on a 5—point scale ranging from completelyfalse to completely true. Alpha reliabilities range from .72for the Impulse Control scale to .92 for the compositeDistress scale.Procedures. Data collection procedures proceeded asdescribed in Studies 1 and 2.Analyses. In order to assess the relations between thescales of the WAI and the dimensions of the FFM, zero ordercorrelations between WAI and IIP—B5 scales were examined, andconjoint principal components analyses were performed at boththe subscale/facet and superordinate levels. Additionally,cohorts falling into each of the six WAI adjustment stylegroups were partitioned and their IIP—B5 profiles examinedusing multivariate analysis of variance. A replication ofthe Weinberger and Schwartz (1990) methodology, examiningpair—wise between group comparisons using the IIP—B5 domainscales as dependent variables was examined.168HypothesesIt was hypothesized that the scales and constructs ofthe WAI could be subsumed by the dimensions of the FFM. Theadjustment typology would not produce nonadditive effects onthe dimensions of the FFM when paired comparisons amonggroups were examined.ResultsCorrelations. The Pearson correlations among the scalesof the WAI and the domain scales of the IIP—B5 are presentedin Table 58. All WAI scales had significant correlationswith neuroticism and dominance. All distress scalescorrelated most highly and positively with neuroticism. Allrestraint scales had small, but significant negativecorrelations with neuroticism. In examining the correlationsof the restraint scales, the most evident pattern is apredominant relationship to the interpersonal dimensions ofdominance (negatively) and nurturance (positively).Weinberger and Schwartz (1990) propose that the restraintdimension is a combination of agreeableness (nurturance) andconscientiousness. In particular, they suggest thatSuppression of Aggression and Consideration are related toagreeableness. While Consideration’s highest correlation iswith nurturance (r = .54), Suppression of Aggression is moststrongly related to dominance (r = -.60), suggestingassessment of nonassertiveness. Additionally they suggestthat Impulse Control and Responsibility are related to169Table 58. Correlations between the WAI and the IIP-B5.IIP-B5 Domain ScaleNEUR UNCON OPEN NURWAI ScaleAnxiety .69 -.17Depression .75 .15 —.25 —.16Low Self—esteem .70 .21 -.38Low well-being .54 —.16 -.33 —.39DISTRESS .81 -.34 —.19Supp. of Aggr. -.17 -.60 .27Impulse Control —.23 —.29 —.28 —.27Consideration -.14 —.26 .54Responsibility —.27 —.21 —.26 .27RESTRAINT —.27 —.23 —.48 .35Note. N 307. p > .01 deleted. Abbreviations:WAI=Weinberger Adjustment Inventory; IIP-B5=Inventory ofInterpersonal Problems—Big Five Version; NEUR=neuroticism;UCON=unconscientious; OPEN=openness; DOM=dominance;NUR=nurturance; Supp of Aggr=supression of aggression.conscientiousness. while both these scales correlate between-.21 and -.29 with unconscientiousness, both exhibitcorrelations of equal magnitude with four of the five domainscales of the IIP—B5. Additionally, the Restraint scalecorrelates most highly with dominance (r = -.48) rather thanwith unconscientiousness or nurturance. Finally, weinbergerand Schwartz claim that the Low Well-being scale is relatednegatively to extraversion contributing to the distress170dimension’s substantive difference from trait neuroticism.Low Well-being does correlate significantly with thedimensions of the interpersonal circumplex, however, itshighest correlation is with neuroticism (r = .54).Structural Analyses. The Restraint and Distresssubscales of the WAI and the domain scales of the IIP-B5 weresubjected to a conjoint principal components analysis. Aclear five factor solution emerged which accounted for 77.2%of the variance. The solution was rotated to a varimaxcriterion and can be seen in Table 59.Examination of Table 59 indicates that all WAT Distresssubscales are strongly related to neuroticisiu. The WAI LowWell-being subscale splits its loading, positively onneuroticism (.67) and negatively on nurturance (-.51). WAISuppression of Aggression loads solely on thenonassertiveness factor (.85). WAI responsibility exhibits ahigh positive loading on nonassertiveness (.67) and amoderate negative correlation on the unconscientious factor(-.35). WAI Impulse Control splits its loading, positivelyon nonassertiveness (.58) and negatively on unconscientious(-.53). WAI Consideration splits its loading, positively onthe nurturance factor (.68) and positively on thenonassertiveness factor (.46). No WAI subscales exhibitedprimary loadings on factors marked by IIP—B5 Unconscientiousor IIP-B5 Openness.The WAI superordinate scales for Restraint and Distressand the IIP—B5 domain scales were also subjected to a171Table 59. Conjoint principal components analysis of the WAIsubscales and the IIP—B5 domain scales.FactorsI II III IV VScalesNeuroticism (IIP—B5) .90Depression (WAI) .87Low Self—esteem (WAI) .84Anxiety (WAI) .83Low Well-being .67 -.51Supp. of Aggr. (WAI) .85Dominance (IIP-B5) -.77Responsibility (WAI) .67 —.35Impulse Control (WAI) .58 -.53Nurturance (IIP-B5) .89Consideration (WAI) .46 .68Unconscientious (IIP—B5) .87Openness (IIP—B5) .97Note. N = 307. Loadings < .33 deleted. Abbreviations:WAI=Weinberger Adjustment Inventory; IIP-B5=Inventory ofInterpersonal Problems-Big Five Version; Supp. ofAggr.=suppression of aggression.172conjoint principal components analysis. A clear five factorsolution emerged which accounted for 93.3% of the variance.The solution was rotated to varimax criterion and ispresented in Table 60.Table 60. Conjoint principal components analysis of the WAIRestraint and Distress scales and the IIP-B5 domainscales.FactorsI II III IV VScalesNeuroticism (IIP—B5) .96Distress (WAI) .93Dominance (IIP-B5) .89Restraint (WAI) -.80 .33Nurturance (IIP-B5) .98Unconscientious (IIP—B5) .99Openness (IIP-B5) .99Note. N = 307. Loadings < .33 deleted. Abbreviations:WAI=Weinberger Adjustment Inventory; IIP-B5=Inventory ofInterpersonal Problems—Big Five Version.Examination of Table 60 indicates that the WAIdimensions of distress and restraint are subsumed by the173superordinate structure of the FFM. Distress loads singlyand significantly on neuroticism (.93). Restraint exhibits aprimary negative loading on dominance (—.80) and a smallsecondary loading on nurturance (.33).Assessment of the WAI Adlustment Typology. The samplestandard scores on Distress and Restraint were used topartition subjects into six adjustment groups. A 2 (highdistress/low distress) by 3 (high restraint/moderaterestraint/low restraint) MANOVA was conducted using thestandardized IIP—B5 domain scale scores as dependentvariables. The cell sizes were as follows: Reactive (45),Sensitized (57), Oversocialized (41), Undersocialized (35),Self—assured (65), and Repressive (51). The multivariate Efor the Distress x Restraint interaction was not significant.The multivariate for Restraint was significant andunivariate tests revealed a main effect for Restraint onneuroticism (2, 271) = 6.77, p = .001; unconscientiousness(2, 271) = 6.15, p= .002; dominance (2, 271) = 48.58, p< .001; and nurturance (2, 271) = 23.46, p < .001. Themultivariate for Distress was significant and univariatetests revealed a main effect for Distress on neuroticism F(2, 271) = 245.19, p< .001; and dominance (2, 271) =55.79, p < .001.In order to replicate the procedures employed byWeinberger and Schwartz (1990), the two by three factorialwas reconceptualized as a oneway analysis of variance withadjustment type as the independent variable and the IIP-B5174scales as the dependent variables. When the univariate forAdjustment Group was significant, Weinberger and Schwartz(1990) evaluated all possible pairwise comparisons using theNewman—Keuls test. In the present analysis, pairwisecomparisons were evaluated using the Tukey test to controlfor family wise error at alpha = .05. The Tukey test waschosen rather than the Newman—Keuls on the recommendation ofKeppel (1982), who reviewed Monte Carlo studies comparing theTukey, Duncan, and Newman—Keuls tests and concluded “thecollective evidence seems to support the conclusion that theTukey test is preferred over the other two tests” (p. 157).The Newman—Keuls test does not adequately control for familywise error (Einot & Gabriel, 1975). The results of thesecomparisons are summarized in Table 61.Significant group comparisons were found forneuroticism, I nurturance, and unconscientiousness.However, only the group differences for unconscientiousnessexhibited a distinct pattern not attributable to main effectsfor restraint, distress, or additive main effects for bothrestraint and distress. Univariate effects on openness werenonsignificant. To better understand the distinct pattern ofgroup differences on unconscientiousness, the procedure wasreplicated using the standardized scores for IIP-B5unconscientious facets as dependent variables. The resultsof this analysis can be found in Table 62.Significant group comparisons were found only for theImpulsive facet. However, the pattern of group differences175Table 61. Tukey test comparisons of adjustment groupdifferences on dimensions of the Big Five.High Distress Low DistressIIP-B5 OVER- UNDER- UNIV.Scale REAC SENS SOC SOC ASS REPR F pATaNEUR 1.01 .68 .60 —.35 —.58 —.86 59.03* ND(H) (H) (H) (L) (L) (L)DOM .14 —.30 —.95 1.18 .15 —.22 28.26* ND(L) (H)NUR —.53 .08 .21 —.48 .10 .63 10.04* ND(L) (H) (L) (H)UNC .44 .19 —.08 .27 .06 —.34 3.28* D(H) (H) (H) (L)OPEN —.06 —.01 —.19 .22 .19 —.02 1.08(no significant comparisons)Note. Groups labeled with (“H”) (i.e., High) aresignificantly different than those with (“L”) (i.e., Low)according to the Tukey test.Abbreviations: IIP—B5=Inventory of InterpersonalProblems-Big Five Version; REAC=reactive; SENS=sensitized;OVER-SOC=oversocialized; UNDER-SOC=undersocialized; ASS=selfassured; REPR=repressive; UNIV. F=univariate F; PAT=pattern;NEUR=neuroticism; DOM=dominance; NUR=nurturance;UNC=unconscientious; OPEN=openness; ND=non-distinct;D=distinct.a. “Nondistinct” patterns completely match linear contrastsfor distress and/or restraint. Pattern for neuroticismpurely corresponds to main effects for distress. Pattern fordominance purely corresponds to additive main effects forboth distress and restraint. Pattern for nurturance purelycorresponds to main effects for restraint. Pattern forunconscientious provides distinct information.*p< .01.176Table 62. Tukey test comparisons of adjustment groupdifferences on facets of unconscientiousness.High Distress Low DistressIIP-B5 OVER- UNDER- UNIV.Facet REAC SOC SOC ASS REPR F PATaURGE .30 .23 —.14 .25 —.07 —.15 2.06(no significant comparisons)LAZY .11 .29 .03 —.13 .02 —.16 1.50(no significant comparisons)UNRE .35 —.15 —.11 .41 .06 —.05 2.67(no significant comparisons)CMPL —.22 .00 —.10 —.04 —.09 .24 1.29(no significant comparisons)IMPL .66 .05 —.30 .37 .03 —.38754* ND(H) (L) (H) (L)Note. Groups labeled with (“H”) (i.e., High) aresignificantly different than those with (“L”) (i.e., Low)according to the Tukey test.Abbreviations: IIP—B5=Inventory of InterpersonalProblems-Big Five Version; REAC=reactive; SENS=sensitized;OVER-SOC=oversocialized; UNDER-SOC=undersocialized; ASS=selfassured; REPR=repressive; UNIV. F=univariate F; PAT=pattern;ND=non-distinct; D=distinct; URGE=urge control;UNRE=unreliable; CMPL=compulsive; IMPL:=impulsive.a. “Nondistinct” patterns completely match linear contrastsfor distress and/or restraint. Pattern for impulsive purelycorresponds to main effects for restraint.p < .01.is attributable to main effects for restraint. Thus, thedistinct pattern found for unconscientiousness may best beexplained as a statistical artifact. No evidence was foundthat the distinct pattern of group means for177uncOnscientiousness could be due to nonlinear additiveeffects of the adjustment typology formed by intersecting thedimensions of distress and restraint.DiscussionThe results of this investigation clarify the relationsbetween the WAI dimensions of distress and restraint and thedimensions of the FFM. Weinberger and Schwartz assert thatdistress is most strongly related to trait neuroticism, butdiffers slightly by the inclusion of a Low Well-beingsubscale which is a measure of negative affectivity(Tellegen, 1985) or low extraversion. Correlations betweenthe IIP-B5 domain scales and the Low Well-being scale supportthe contention that this subscale is related to extraversion.The scale’s negative correlations with dominance andnurturance place the scale in the lower left quadrant of theinterpersonal circumplex, indicating a relation with traitintroversion. However, the scale’s highest correlation iswith IIP-B5 neuroticism. Additionally, principal componentsanalysis revealed a strong positive loading on neuroticismand a secondary negative loading on agreeableness. When thesubscales of the WAI distress dimension are aggregated, and aconjoint principal components analysis is performed, WAIDistress loads strongly and solely on neuroticism. Noevidence was found to conclude that this WAI dimension issubstantively different from the neuroticism dimension of theFFM.178Weinberger and Schwartz (1990) propose that thedimension of restraint is a blend of conscientiousness andagreeableness. Specifically, WAI Impulse Control and WAIResponsibility subscales were hypothesized to be related toconscientiousness. While these two scales had negativecorrelations with IIP-B5 Unconscientiousness, they exhibitedsignificant correlations of equal magnitude with three of thefour remaining dimensions of the FFM. WAI Suppression ofAggression and WAI Consideration were hypothesized to berelated to agreeableness. The former subscale correlatednegatively with IIP-B5 Dominance (r = -.60), suggestingassessment of nonassertive problems. The latter subscalecorrelated most highly with IIP-B5 Nurturance, supportingWeinberger and Schwartz’ hypothesis. However, conjointprincipal components analysis indicated that three of thefour restraint subscales loaded most strongly on dominance(negatively). Responsibility and Impulse Control exhibitedsecondary negative loadings on unconscientiousness andConsideration split its loading on nurturance and dominance(indicating the scale assesses an interpersonal construct).When the subscales of restraint are aggregated, WAI Restraintloads strongly on dominance (negatively) and demonstrates aminor secondary loading on nurturance. Hence, the restraintdimension of the WAI is primarily an interpersonal dimensionassessing nonassertive and nurturant traits. It has littlerelation to conscientiousness.179The structural relations between the WAI and the IIP-B5reveal that the dimensions of distress and restraint can beinterpreted clearly within the framework of the FFM. Assuch, it is of questionable necessity to reformulatedimensions of social—emotional adjustment as distinct fromthe basic dimensions of personality. If the five dimensionalspace of the model is to be divided differently, evidence ofthe incremental gain in understanding of personality,predictive utility, or practical application should bedemonstrated. The Dyadic-Interactional perspective indeeddivides the five dimensional space differently than thesimple structure representation assessed by the NEO—PI. Indoing so, the advantages accrued in assessing theinterpersonal circumplex and its associated methodology weredemonstrated. In the case of the WAI, the only advantagecould be the typological categories of adjustment styleproposed by intersecting the dimensions of distress andrestraint.In discussing methods for identifying typologicalcategories, Neehi (1992) suggests that “in constructingassessment devices, the psychometric strategy is different,the distinct technological aims being assignment ofindividuals to a category versus location of individuals on adimension. For the latter task, item difficulties andcorrelations should be chosen so as to disperse scores widelyand discriminate effectively in all regions of the dimension,a very different function from sorting at a best cut so as to180minimize in/out misclassifications” (p. 162). He concludesthat “construction and selection of items to compose aquantitative indicator cannot optimize both dimensional andtaxonic power” (p. 162). Therefore, given the common methodsof personality assessment, Meehl (1992) endorses therequirement of consistency tests applied to suchidentification of typological categories that providePopperian risk of strong discorroboration (Popper, 1959).The validity of the adjustment typology was investigated byreplicating the methods used by Weinberger and Schwartz(1990) to determine if distinct forms could be identifiedusing the FFM dimensions as dependent variables.The results of the analyses of variance replicated thefindings of Weinberger and Schwartz (1990) in that nomultivariate or univariate Distress x Restraint interactionterms were significant, but main effects for both dimensionswere significant. However, none of the between—groupcomparisons deviated from linear main effects for thedimensions, thus typological distinctions were notreplicated.In examining the results of Weinberger and Schwartz(1990), it is apparent that the Newman-Keuls procedure is aparticularly lenient test of group differences. For example,Weinberger and Schwartz reported a distinct, nonadditivepattern of group means for the Capacity of Intimacy scale ofShostrom’s (1974) Personal Orientation Inventory. The meanscore for the entire sample on this scale was 46.28. The181Newman—Keul’s procedure identified a scale score differenceof 2.5 as significantly different. While the standarddeviations of this scale were not reported, it is doubtfulthat individuals demonstrating a difference of 2.5 points ona self-report scale with a mean of 46.28 would be detectablydifferent in their behaviors. The difference betweenstatistically distinct scale scores and personologicaldifferences in kind is an issue not discussed by Weinbergerand Schwartz (1990). This result seems to support Meehl’s(1992) assertion that a good dimensional measure is animperfect typological indicator because just such a resultcan be confused with “distinct features.”The replication attempted here thus used a morestringent test of between-group differences, the Tukey test.Standardized scores on the IIP—B5 domain scales were used asdependent variables. The smallest significant between—groupsmean difference was equal to .53 of a standard deviation, aquantity suggestive of potentially unique personologicalcharacteristics differing between group members. However,even if the analysis had identified nonadditive between groupdifferences, the proposed adjustment typology may not besupported. Taxometric methods (e.g., Meehi & Golden, 1982)to determine best cut points for distress and restraint wouldbe significantly more compelling evidence for the validity ofthe typology. Therefore, the results reported here do notinvalidate the adjustment typology. They do, however, call182into question the methodology used by Weinberger and Schwartz(1990) to suggest typological validity.In conclusion, the results of this investigation suggestthat the reconceptualization of dimensions of social—emotional adjustment as differing from the five basicdimensions of personality is premature. No evidence wasfound that the WAI Distress and WAI Restraint constructs weresubstantively different constructs than the dimensions of theFFM. The typological model of adjustment was not supportedusing the methods employed by Weinberger and Schwartz (1990)and weaknesses in the methodology were highlighted. Finally,the results also suggest the IIP-B5 can be used effectivelyto operationalize the FFM for personality assessmentresearch.183CHAPTER 6An application of the IIP-B5: Clinical personalityassessment in an intensive group psychotherapy programIntroductionThe utility of assessing individual patients with theIIP—B5 can be demonstrated with a case presentation. Costaand McCrae (1992a) proposed six uses of personality data inclinical assessment: understanding the patient, aiding indiagnosis, developing empathy and rapport, providing feedbackand insight, anticipating the course of treatment, andmatching treatments to patients. The following casepresentation briefly demonstrates the use of the IIP-B5 andhighlights how information from a Dyadic—Interactional FFMassessment can aid in understanding the patient’spresentation, developing empathy and rapport in a dynamic—interpersonal therapy group, providing feedback and insight,and anticipating and understanding a patient’s changesthrough a course of psychotherapy.Figure 7 presents the dyadic-interactional five-factormodel profile of interpersonal problems of a 32 year—old malepatient, selected at random, seen for short—term intensivegroup therapy by the author. If information was restrictedto the interpersonal domain of maladaptive and rigidbehavior, it could be assumed (based on his location in theDE octant of the interpersonal problems circumplex) that thispatient may exhibit an inability to express affection andScoresPABCDE.FGHIJi<LMNORawSlandardl3049645547333131Angle:_200.54Vector length:593339S.C.NameM32SexAgeIDA’•DomineeringPA 90ColdDE 180 20323FiveFactorProfileInstrument:IIP—B5OverlynurturanlLMH CD I-I H (31 0 H CD 0 H CD 0 H) H-H CD H) 0 hi (I) C-)80 70 60 50 40 30 20 10 0270 HIH293NonassertiveNEURUNCONSCOPEN185feel love for other people, have difficulty making long-termcommitments to others, and demonstrate a rigid inability tobe generous to, get along with, and forgive others (Alden etal, 1990, p. 528).How would this interpersonal coldness manifest itselfboth in the therapy group this patient was entering and inhis life functioning in general? If the patient wereimpulsive, unconscientious, and relatively free of emotionaldistress (i.e., low neuroticism), his interpersonal coldnesswould have manifested itself in insensitive, hostile, andpossibly antisocial behavior that infringed on the rights ofothers for whom he cared little. In actuality, this was notthe case. The patient was, in fact, over—controlled,emotionally distressed, and conscientious. He was referredto psychotherapy for anxiety and depression related to anobsessive preoccupation with everyday noises which disturbedhim so intensely that he had relocated five times in fivemonths. Additionally, because the patient had so little inthe way of supportive relationships, his major method ofcoping with his distress was the chronic use of marijuana.Figure 8 presents the patient’s problem facet profilefor the IIP-B5. The additional dimensions of neuroticism,unconscientiousness, and openness comprise problem facetswhich assess maladaptive and rigid behavior at a level ofspecificity similar to the lIP-C octant scales. From thisproblem facet profile, it is possible to describe how the186Figure 8. IIP-B5 facet profile for S. C.UNCONSCIENTIOUSLazy Urges Unrehable Impulsive Compulsive (Reckless)8070605040-30________201036 31 27 27 31 29OPENNESS TO EXPERIENCENonconformity Unfocused Unimaginative Egocentric8070 V60NEUROTICISMDepression Anger Anxiety Low sell-esteem80EE187patient’s interpersonal coldness manifested itself within thetherapy group. It can be seen that the patient wasparticularly inhibited and overcontrolled as indicated by hislow scores on the Reckless, Impulsive, and Lack of Focusfacets. Thus his coldness was manifested in a lack ofinteraction and joining, little spontaneity, and difficultyproviding support for others. In the group, he initiallyspoke infrequently, and paused for long periods “to collecthis thoughts” when speaking, inviting interruption fromothers due to evoked frustration and boredom. Thismaladaptive transaction cycle (Kiesler, 1988; Van Denburg etal, in press) furthered his feelings of disconnectedness byreinforcing his covert feelings that others did not care forhim and would dominate and overwhelm him if he were toengage. He presented as distressed but emotionallyconstrained, hence he was neither attacking nor angry, whichis compatible with his below average score on the Angerfacet. Thus, when interrupted or encouraged to expressemotion (as was the evoked group tendency), he would chooseto withdraw and disengage rather than become angry, blaming,or assertive. Thus, his hostile—submissive behavior(withdrawal, silence) reinforced the complementary responsesof hostile—dominance (interruptions, demands that he emoteand be involved) that characterized his engagement with thegroup.Consistent with his low scores on the Lazy andUnreliable facets, his interpersonal coldness did not188manifest itself though lateness, absences, or earlytermination (the group met 4 times a week for 6 weeks).However, his high score on Nonconformity indicated anadditional component of his interpersonal coldness. Hepresented as an artistic recluse, with long hair and rathereccentric clothes. He saw himself as an outsider in society,a fringe person, and he did not wish to lose this componentof his self-concept. In many areas of his life, he didindeed value being different. Again, his interpersonalcoldness was maintained by a resistance to influence fromauthority, society, or other people.Van Denburg et al (in press) describe patient’s currentproblematic transactional patterns as stemming frominternalized early interactions with significant others.Clearly, this patient’s early years played a role in hispresentation when therapy was initiated. Briefly, the grouplearned that his father was an angry, intrusive, anddomineering lawyer who would not tolerate disagreement withhis own views, rules, and desires. This description wascorroborated when the patient’s father attended a weeklyguest night for the group. As a young boy, the patientwitnessed frequent loud, aggressive, and frighteningaltercations between his older brother and his father. Hisresponse to this developmental experience was to withdraw andrepress anger and emotion because of his fear of: (a) beingsimilarly assaulted by his father and (b) becoming like hisfrightening and feared father. At the same time, however,189his coping strategy evoked feelings of inadequacy for beingunable to stand up for himself as he saw his older brother dodespite the frightening consequences. His feelings ofinadequacy were detected by a high score on the Low Self—esteem facet.As group treatment progressed, the interruptions anddomineering complementary responses of the group shifted.They learned more about the patient’s development and hiscovert responses to the group. They began to provide supportand respectful suggestions, as well as displaying genuineaffection for him. At the time of termination, the grouptypically provided acomplementary responses (Kiesler, 1983,1988) of friendly—dominance which evoked increased engagementand emotional expression from the patient. His first stepstoward breaking a maladaptive transaction cycle took placeduring an intensive period of group treatment. He thenentered a continuing therapy group that met once a week.DiscussionThe case just presented is a brief demonstration of theutility of assessment with the five—factor model ofinterpersonal problems. One important point is that theadditional dimensions of the five—factor model will likelyinfluence the interpersonal presentation of individualpatients. Two patients whose interpersonal profiles aresimilar (i.e., both locate in the same circumplex octant) arelikely to present with distinctive patterns of maladaptiveand rigid behavior (problems), depending on their relative190standings on emotional lability and distress (neuroticism),impulsivity, persistence, and reliability(unconscientiousness), and conformity, tolerance, andimagination (openness). This useful information can beobtained efficiently by administering a combined circumplexand five—factor assessment battery such as the TIP—B5.In reviewing his personality information, it waspossible to anticipate the patient’s responses to theinterruptions and impatience which characterized early groupengagement. This aided in providing group processinterpretations which helped to establish rapport andempathy. As treatment continued, the repetition of thepattern and continued group process interpretations werelinked with developmental issues identified over the courseof psychotherapy, providing the patient with insight into hismaladaptive and rigid behavior. As behavioral shifts becameapparent, empathic feedback and process interpretations werereceived with greater self understanding and a sense ofpersonal control over his engagement with others. Whileother useful information was provided by the pretreatmentassessment (including a clinical interview and other self—report measures), the personality information provided by theIIP-B5 was particularly useful in understanding and treatingthis patient in a group psychotherapy context.191REFERENCESAdler, A. (1912). The neurotic character. In H.L.Ansbacher & RR. Ansbacher (Eds.), (1956). Theindividual psychology of Alfred Adler. New York:Harpur.Adler, A. (1964). Social interest: A challenge to mankind.New York, Putnam.Affleck, G., Tennen, H., Urrows, S., & Higgins, P. (1992).Neuroticism and the pain—mood relation in rheumatoidarthritis: Insights from a prospective daily study.Journal of Consulting and Clinical Psychology, Q, 119—126.Alden, L. E. (1991, August). An interpersonal analysis ofavoidant personality disorder. Paper presented at thesymposium on Assessment of Interpersonal Problems:Implications for Treatment and Research, L.E. Alden(Chair). American Psychological Association annualconvention, San Francisco, CA.Alden, L. E., & Capreol, M. J. (1992). Prediction oftreatment response in avoidant personality disorderedoutpatients using the Inventory of Interpersonal ProblemsCircumplex Scales. Manuscript submitted for publication.Alden, L. E., & Phillips, N. (1990). An interpersonalanalysis of social anxiety and depression. CognitiveTherapy and Research, j, 499-513.Alden, L. E., Wiggins, J. S., & Pincus, A. L. (1990).Construction of circumplex scales for the Inventory ofInterpersonal Problems. Journal of PersonalityAssessment, , 521—536.Ailport, G. W. (1937). Personality: A psychologicalinterpretation. New York: Holt.Ailport, G. W., & Odbert, H. S. (1936). Trait-names: Apsycho-lexical study. Psychological Monographs, 41 (1,Whole No. 211).American Psychiatric Association (1980). Diagnostic andStatistical Manual of Mental Disorders (DSM—III).Washington, DC: Author.Anchin, J. C., & Kiesler, D. J. (Eds.) (1982). Handbook ofinterpersonal psychotherapy. New York: Permagon.192Angleitner, A., & Ostendorf, F. (1989). Personality factorsvia self— and peer—ratings based on a representativesample of German trait descriptive terms. Paperpresented at the first European Congress of Psychology,Amsterdam.Angleitner, A., Ostendorf, F., & John, 0. P. (1990).Towards a taxonomy of personality descriptors in German:A psycho—lexcial study. European Journal of Personality,4, 89—118.Arnold, A. P., & Cheek, 3. N. (1986). Shyness, self-preoccupation, and the Stroop color and word test.Personality and Individual Differences, 2, 571-573.Bartholomew, K., & Horowit2, L. N. (1991). Attachmentstyles in young adults: A test of a four—category model.Journal of Personality and Social Psychology, 226—244.Ben—Porath, Y. S., & Wailer, N. G. (1992a). “Normal”personality inventories in clinical assessment: Generalrequirements and the potential for using the NEOPersonality Inventory. Psychological Assessment: AJournal of Consulting and Clinical Psychology, 4, 14-19.Ben—Porath, Y. S., & Wailer, N. G. (l992b). Five big issuesin clinical personality assessment. PsychologicalAssessment: A Journal of Consulting and ClinicalPsychology, 4, 23-25.Benjafield, 3., & Carson, E. (1985). Anhistoricodevelopmental analysis of the circumplex modelof trait descriptive terms. Canadian Journal ofBehavioral Science, 4, 339—345.Benjamin, L. S. (1974). Structural analysis of socialbehavior. Psychological Review, i, 392-425.Benjamin, L. S. (1984). Principals of prediction usingstructural analysis of social behavior. In R.A. Zucicerand A.J. Aronoff (Eds.), Personality and prediction ofbehavior (pp. 121-173). New York: Academic.Benjamin, L. S. (1988). Intrex shortform users manual.Madison, WI: Intrex Interpersonal Institute.Benjamin, L. S. (in press). Brief SASB-directedreconstructive learning therapy. In P. Crits-Cristoph &J. Barber (Eds.), Handbook of short term dynamic therapy.New York: Guilford.193Bentler, P. M., & Kano, Y. (1990). On the equivalence offactors and components. Multivariate BehavioralResearch, 25, 67-74.Birtchnell, J. (1991). The measurement of dependence byquestionnaire. Journal of Personality Disorders, .., 281—295.Bolger, N., & Schilling, E. (1991). Personality and theproblems of everyday life: The role of neuroticism inexposure and reactivity to daily stressors. Journal ofPersonality, , 355-386.Borkenau, P., & Ostendorf, F. (1989). Descriptiveconsistency and social desirability in self— and peer-reports. European Journal of Personality, , 31—45.Briggs, S. R. (1988). Shyness: Introversion orneuroticism? Journal of Research in Personality, 22,290—307.Briggs, S. R., Hogan, R., & Jones, W. H. (Eds.). (in press).Handbook of Personality Psychology. Orlando, FL:Academic Press.Burisch, M. (1986). Methods of personality inventorydevelopment: A comparative analysis. In A. Angleitnerand J.S. Wiggins (Eds.), Personality Assessment viaquestionnaires (pp. 109-120). New York: Springer-Verlag.Buss, A. H. (1989). Personality as traits. AmericanPsychologist, j4, 1378—1388.Buss, D. M. (1991). Evolutionary personality psychology.Annual Review of Psychology, 42, 459-491.Buss, D. N., & Craik, K. H. (1983). The act frequencyapproach to personality. Psychological Review, , 105-126.Buss, D. N., & Craik, K. H. (1984). Acts, dispositions, andpersonality. In B.A. Maher & W.B. Maher (Eds.), Progressin experimental personality research: Normal personalityprocesses (Vol. 13, pp. 241—301). New York: AcademicPress.Buss, D. M., & Craik, K. H. (1986). Acts, dispositions, andclinical assessment: The psychopathology of everydayconduct. Clinical Psychology Review, , 387-406.Butcher, J. N. (1990). MMPI-2 in psychological treatment.New York: Oxford University Press.194Butcher, J. N. (1992). Introduction to the special sectionon assessing personality characteristics in clinicalsettings. Psychological Assessment: A Journal ofConsulting and Clinical Psychology, , 3-4.Carson, R. C. (1969). Interaction concepts of personality.Chicago: Aldine.Carson, R. C. (1982). Self-fulfilling prophecy, maladaptivebehavior, and psychotherapy. In J.C. Anchin & D.J.Kiesler (Eds.), Handbook of interpersonal psychotherapy(pp. 64-77). New York: Permagon Press.Carson, R. C. (1991). Dilemmas in the pathway of the DSMIV. Journal of Abnormal Psychology, jQ, 302-307.Carson, R. C. (in press). Discussion: Can the Big Fivehelp salvage the DSM? Psycholociical Inquiry.Cattell, R. B. (1943). The description of personality.II. Basic traits resolved into clusters. Journal ofAbnormal and Social Psychology, , 476-507.Cattell, R. B. (1945). The principal trait clusters fordescribing personality. Psychological Bulletin, 42, 129-169.Cattell, R. B. (1957). Personality and motivation structureand measurement. New York: World Book.Cattell, R. B. (1973). Personality and mood byquestionnaire. San Francisco, CA: Jossey—Bass.Chapman, L. J., Chapman, J. P., , J. S., Edell, W.S., Carpenter, B. N., & Blackfield, D. (1984). Impulsivenonconformity as a trait contributing to the predictionof psychotic-like and schizotypal symptoms. Journal ofNervous and Mental Disease, 172, 681—691.Cheek, J. M., & Meichior, L. A. (1985, August). Measuringthe three components of shyness. Paper presented at theannual meeting of the American Psychological Association,Los Angeles, CA.Cheek, J. N., & Stahl, S. S. (1986). Shyness and verbalcreativity. Journal of Research in Personality, 2Q, 51—61.Comrey, A. L. (1988). Factor—analytic methods of scaledevelopment in personality and clinical psychology.Journal of Consulting and Clinical Psychology, , 754-761.195Costa, P.T., Jr., (Ed.) (1991). Clinical use of the fivefactor model. Journal of Personality Assessment.(Special Section), 57, 393-464.Costa, P. T., Jr., & McCrae, R. R. (1980). Still stableafter all these years: Personality as a key to someissues in adulthood and old age. In P.B. Baltes & O.G.Brim, Jr. (Eds.), Life span development and behavior(Vol. 3, pp. 65—102). New York: Academic Press.Costa, P. T., Jr., & McCrae, R. R. (1985). The NECPersonality Inventory manual. Odessa, FL: PsychologicalAssessment Resources.Costa, P. T., Jr., & McCrae, R. R. (1986a). Personalitystability and its implications for clinical psychology.Clinical Psychology Review, , 407-424.Costa, P. T., Jr., & McCrae, R. R. (1986b). Clinicalassessment can benefit from recent advances inpersonality psychology. American Psychologist, 4, 10011003.Costa, P. T., Jr., & McCrae, R. R. (1987). Neuroticism,somatic complaints, and disease: Is the bark worse thanthe bite? Journal of Personality, 5, 299—316.Costa, P. T., Jr., & McCrae, R. R. (l988a). From catalogueto classification: Murray’s needs and the five—factormodel. Journal of Personality and Social Psychology, j,258—265.Costa, P. T., Jr., & McCrae, R. R. (l988b). Personality inadulthood: A six-year longitudinal study of self-reportsand spouse ratings on the NEO Personality Inventory.Journal of Personality and Social Psychology, 54, 853-863.Costa, P. T., Jr., & McCrae, R. R. (1989). The NEO-PI/NEOFF1 manual supplement. Odessa, FL: PsychologicalAssessment Resources.Costa, P. T., Jr., & McCrae, R. R. (1990). Personalitydisorders and the five—factor model of personality.Journal of Personality Disorders, 4., 362-371.Costa, P.T., Jr., & McCrae, R. R. (1992a). Normalpersonality assessment in clinical practice: The NEOPersonality Inventory. Psychological Assessment: AJournal of Consulting and Clinical Psychology, 4., 5-13.Costa, P. T., Jr., & McCrae, R. R. (1992b). Reply to BenPorath and Waller. Psychological Assessment: A Journalof Consulting and Clinical Psychology, 4., 20-22.196Costa, P. T., Jr., & McCrae, R. R. (in press). Traitpsychology comes of age. In T.B. Sonderegger (Ed.),Nebraska symposium on motivation: Psycho1oy and aging.Lincoln, NE: University of Nebraska Press.Costa, P. T., Jr., Busch, C. M., Zonderinan, A. B., & McCrae,R. R. (1986). Correlations of MMPI factor scales withmeasures of the five—factor model of personality.Journal of Personality Assessment, , 640-650.Costa, P. T., Jr., McCrae, R. R., & Dembroski, T. M. (1989).Agreeableness vs. antagonism: Explication of a potentialrisk factor for CHD. In A. Siegman & T.M. Dembroski(Eds.), In search of coronary-prone behavior (pp. 41-63).Hilisdale, NJ: Eribaum.Costa, P. T., Jr., McCrae, R. R., & Dye, D. A. (1991).Facet scales for agreeableness and conscientiousness: Arevision of the NEO personality inventory. Personalityand Individual Differences, j, 887-898.Costa, P. T., Jr., McCrae, R. R., & Holland, J. L. (1984).Personality and vocational interests in an adult sample.Journal of Applied Psycholocry, , 390-400.Cronbach, L. J. (1949). Statistical methods applied toRorschach scores: A review. Psychological Bulletin, 4k.,393—429.Cronbach, L. 3., & Meehl, P. E. (1955). Construct validityof psychological tests. Psychological Bulletin, , 281-302.DeJong, C. A. 3., van den Brink, W., Jansen, J. A. M., &Schippers, G. M. (1989). Interpersonal aspects of DSMIII Axis II: Theoretical hypotheses and empiricalfindings. Journal of Personality Disorders, 2, 135-146.Derogatis, L. R. (1977). The SCL-90 manual: Scoring,administration, and procedures for the SCL-90.Baltimore, MD: Johns Hopkins University School ofMedicine, Clinical Psychometrics Unit.Derogatis, L. R., & Melisaratos, N. (1983). The BriefSymptom Inventory: An introductory report.Psychological Medicine, , 596—605.Digman, J. M. (1990). Personality structure: Emergence ofthe five—factor model. Annual Review of Psychology, ,195—214.197Edwards, A. L. (1953). The relationship between judgeddesirability of a trait and the probability that thetrait will be endorsed. Journal of Applied Psychology,az, 90—93.Edwards, A. L. (1966). Relationship between probability ofendorsement and social desirability scale value for a setof 2,824 personality statements. Journal of AppliedPsycholoay, , 238-239.Edwards, A. L. (1967). The social desirability variable: Areview of the evidence. In l.A. Berg (Ed.), Response setin personality assessment (pp. 48—70). Chicago, IL:Aldine.Einot, I., & Gabriel, K. R. (1975). A study of powers ofseveral methods of multiple comparisons. Journal of theAmerican Statistical Association, 7, 574-583.Eysenck, H. J. (1947). Dimensions of personality. London:Routledge & Kegan Paul.Eysenck, S. B. G., Eysenck, H. J., & Barret, p. (1985). Arevised version of the psychoticism scale. Personalityand Individual Differences, ., 21-29.Fagan, P. J., Wise, T. N., Schmidt, C. W., Jr., Ponticas,Y., Marshall, R. D., & Costa, P. T., Jr. (1991). Acomparison of five—factor personality dimensions in maleswith sexual dysfunction and males with paraphilia.Journal of Personality Assessment, i, 434-448.Fenigstein, A., Scheier, M. F., & Buss, A. H. (1975).Public and private Self—consciousness: Assessment andtheory. Journal of Consulting and Clinical Psychology,4, 522—527.Fiske, D. W. (1949). Consistency of the factorialstructures of personality ratings from different sources.Journal of Abnormal and Social Psychology, 4.4, 329-344.Foa, U. G., & Foa, E. B. (1974). Societal structures of themind. Springfield, IL: Charles C. Thomas.Freedman, M. B. (1985). Symposium: Interpersonalcircumplex models (1948-1983). Journal of PersonalityAssessment, 4., 621—625.Freedman, M. B., Leary, T. F., Ossorio, A. G., & Coffey, H.S. (1951). The interpersonal dimension of personality.Journal of Personality, 20, 143-161.198Funder, D. C. (1991). Global traits: A neo-Allportianapproach to personality. Psychological Science, Z, 31-39.Goldberg, L. R. (1981). Language and individualdifferences: The search for universals in personalitylexicons. In L. Wheeler (Ed.), Review of personality andsocial psychology (Vol. 2, pp. 141—165). Beverly Hills,CA: Sage.Goldberg, L. R.. (1982). From Ace to Zombie: Someexplorations in the language of personality. In C.D.Spielberger & J.N. Butcher (Eds.), Advances inpersonality assessment (Vol. 1, pp. 203—234). Hilisdale,NJ: Lawrence Eribaum.Goldberg, L. R. (1990). An alternative “description ofpersonality”: The Big—Five factor structure. Journal ofPersonality and Social Psychology, , 1216—1229.Goldberg, L. R. (1992). The development of markers for theBig—Five factor structure. Psychological Assessment: AJournal of Consulting and Clinical Psychology, .4, 26-42.Guilford, J. P. (1948). Factor analysis in a testdevelopment program. Psychological Review, , 79-94.Guilford, J. P. (1959). Personality. New York: McGrawHill.Gunderson, J. G., Links, P. S., & Reich, J. H. (1991).Competing models of personality disorders. Journal ofPersonality Disorders, , 60-68.Gurtman, M. B. (1991). Evaluating the interpersonalness ofpersonality scales. Personality and Social PsychologyBulletin, 1, 670—677.Gurtman, M. B. (1992). Trust, distrust, and interpersonalproblems: A circumplex analysis. Journal of Personalityand Social Psychology, , 989—1002.Gurtman, M. B. (in press—a). Constructing personality teststo meet a structural criterion: Application of theInterpersonal Circumplex. Journal of Personality.Gurtman, M. B. (in press-b). On the construct validity ofinterpersonal personality measures: The interpersonalcircumplex as nomological net. Journal of Personalityand Social Psychology.199Gurtman, M. B., Fernandez, S. J., & Phillips, L. A. (1991,August). Narcissism from the viewpoint of aninterpersonal problem circumplex. Paper presented at thesymposium on Assessment of interpersonal problems:Implications for treatment and research, L.E. Alden(Chair). American Psychological Association annualconvention, San Francisco, CA.Guttman, L. (1954). A new approach to factor analysis: Theradex. In P.R. Lazarsfeld (Ed.), Mathematical thinkingin the social sciences. Glencoe, IL: Free Press.Harpur, T. J., Hart, S. D., & Hare, R. D. (in press). Thepersonality of the psychopath. In P.T. Costa, Jr., &T.A. Widiger (Eds.), Personality disorders and the fivefactor model of personality. Washington, DC: AmericanPsychological Association.Hase, H. D., & Goldberg, L. R. (1967). Comparative validityof different strategies for constructing personalityinventory scales. Psychological Bulletin, i, 231-248.Henderson, N. B. (1982). Human behavior genetics. AnnualReview of Psycholoy, .33, 403-440.Henry, W. P., Schacht, T. E., & Strupp, H. H. (1990).Patient and therapist introject, interpersonal process,and differential psychotherapy outcome. Journal ofConsulting and Clinical Psychology, , 768-774.Hicks, L. E. (1984). Conceptual and empirical analysis ofsome assumptions of an explicitly typological theory.Journal of Personality and Social Psychology, 4k., 1118-1131.Hirschfeld, R. M. A., Kierman, G. L., Gough, H. G., Barret,J., Korchin, S. J., & Chodoff, P. (1977). A measure ofinterpersonal dependency. Journal of PersonalityAssessment, 41, 610—619.Hofstee, W. K. B. (1990). The use of everyday personalitylanguage for scientific purposes. European Journal ofPersonality, 4., 77—88.Hofstee, W. K. B., & Van Heck, G. L. (Eds.) (1990). Specialissue on personality language. European Journal ofPersonality, 4., 75—173.Hogan, R. (1983). A socioanalytic theory of personality. InM. M. Page (Ed.), Nebraska Symposium on Motivation:Personality - Current Theory and Research (pp. 55-89).Hogan, R. (1986). Hogan Personality Inventory Manual.Minneapolis, MN: National Computer Systems.200Hogan, R., & Johnson, J. A. (1981, August). The structureof personality. Paper presented at the annual meeting ofthe American Psychological Association, Los Angeles, CA.Hogan, R., Carpenter, B. N., Briggs, S. R., & Hansson, R. 0.(1984). Personnel assessment in personnel selection. InJ.J. Bernardin & D.A. Bownas (Eds.), Personalityassessment in organizations (pp. 21—52). New York:Praeger.Hogan, R., Cheek, J. N., & Jones, W. H. (1985).Socioanalytic theory: An alternative to armadillopsychology. In B. Schienker (Ed.), The self and sociallife (pp. 175-198). New York: McGraw-Hill.Homey, K. (1937). The neurotic personality of our time.New York: Norton.Horowitz, L. M. (1979). On the cognitive structure ofinterpersonal problems treated in psychotherapy. Journalof Consulting and Clinical Psychology, 41, 5-15.Horowitz, L. M. (1991, August). The capacity to describeother people clearly: Prediction of outcome in briefdynamic psychotherapy. Paper presented at the symposiumon Assessment of interpersonal problems: Implicationsfor treatment and research, L.E. Alden (Chair). AmericanPsychological Association annual convention, SanFrancisco, CA.Horowitz, L. N., Locke, K. D., Morse, M. B., Waiker, S. V.,& Dryer, D. C. (1991). Self-derogations and theinterpersonal theory. Journal of Personality and SocialPsychology, fl, 68—79.Horowitz, L. N., Rosenberg, S. E., Baer, B. A., Ureno, G., &Villasenor, V. S. (1988). Inventory of InterpersonalProblems: Psychometric properties and clinicalapplications. Journal of Consulting and ClinicalPsychology, j, 885-892.Horowitz, L. M., Rosenberg, S. E., Ureno, G., Kahlezhan, B.M., & O’Halloran, P. (1989). Psychodynamic formulation,consensual response method, and interpersonal problems.Journal of Consulting and Clinical Psychology, i, 599-606.Horowitz, L. N., & Vitkus, J. (1986). The interpersonalbasis of psychiatric symptoms. Clinical PsychologyReview, , 443-469.201Humphrey, L. L. (1989). Observed family interactions amongsubtypes of eating disorders using Structural Analysis ofSocial Behavior. Journal of Consulting and ClinicalPsychology, , 206-214.Hyler, S. E., & Lyons, M. (1988). Factor analysis of theDSM-III personality disOrder clusters: A replication.Comprehensive Psychiatry, , 304-308.Jackson, D. N. (1989). Basic Personality Inventory manual.Port Huron, MI: Sigma Assessment SystemsJackson, D. N.., & Messick, 5. (1958). Content and style inpersonality assessment. Psychological Bulletin, , 243—252.John, 0. P. (1990a). The “Big Five” factor taxonomy:Dimensions of personality in the natural language and inquestionnaires. In L.A. Pervin (Ed.), Handbook ofpersonality: Theory and research (pp. 66-100). NewYork: Guilford.John, 0. P. (1990b). The search for basic dimensions ofpersonality: A review and critique. In P. McReynolds,J.C. Rosen, & G.J. Chelune (Eds.), Advances inPsychological Assessment, 2, 1-37.John, 0. P., Angleitner, A., & Ostendorf, F. (1988). Thelexical approach to personality: A historical review oftrait taxonomic research. European Journal ofpersonality, a, 171-203.Johnson, J. H., Butcher, J. N., Null, C., & Johnson, K. N.(1984). Replicated item-level factor analysis of thefull MMPI. Journal of Personality and Social Psychology,41, 105—114.Johnson, M. E., Popp, C., Schacht, T. E., Mellon, J., &Strupp, H. H. (1989). Converging evidence foridentification of recurrent relationship themes:Comparison of two methods. Psychiatry, , 275-288.Kendler, K. S., & Hewitt, J. (1992). The structure of self-report schizotypy in twins. Journal of PersonalityDisorders, 6, 1-17.Keppel, G. (1982). Design and analysis: A researcher’shandbook. Englewood Cliffs, NJ: Prentice-Hall Inc.Kiesler, D. J. (1983). The 1982 interpersonal circle: Ataxonomy for complementarity in human transactions.Psychological Review, 9.Q, 185—214.202Kiesler, D. J. (1986). The 1982 interpersonal circle: Ananalysis of DSM—III personality disorders. In T. Millon& G.L. Kleriuan (Eds.), Contemporary directions inpsychopathology: Toward the DSM-IV (pp. 571-597). NewYork: Guilford.Kiesler, D. J. (1988). Therapeutic metacommunication:Therapist impact disclosure as feedback in psychotherapy.Palo Alto, CA: Consulting Psychologists Press.Kiesler, D. J. (1991). Interpersonal methods of assessmentand diagnosis. In C.R. Snyder and D.R. Forsyth (Eds.),Handbook of Social and Clinical Psychology: The HealthPerspective (pp. 438-468). Elmsford, NY: Permagon.Kiesler, D. J., & Schmidt, J. A. (1991). The Impact MessageInventory: Form hA octant scale version. Unpublishedmanuscript, Virginia Commonwealth University, Richmond,Virginia.Kiesler, D. J., Van Denburg, T. F., Sikes-Nova, V. E.,Larus, J. P., & Goldston, C. S. (1990). Interpersonalbehavior profiles of eight cases of DSM-III personalitydisorder. Journal of Clinical Psychology, 4, 440-453.Kiesler, D. J., & Watkins, L. M. (1989). Interpersonalcomplementarity and the therapeutic alliance: A study ofrelationship in psychotherapy. Psychotherapy, 2, 183-194.LaForge, R. (1985). The early development of the FreedmanLeary—Cof fey interpersonal system. Journal ofPersonality Assessment, , 613-621.LaForge, R., & Suczek, R. F. (1955). The interpersonaldimension of personality: III. An interpersonal checklist. Journal of Personality, , 94-112.Larsen, R. (in press). Neuroticism and bias in the encodingand recall of physical symptoms: Evidence from acombined prospective—retrospective study. Journal ofPersonality and Social Psychology.Larsen, R., & Kasimatis, M. (1991). Day-today physicalsymptoms: Individual differences in the occurrence,duration, and emotional concomitants of minor dailyillnesses. Journal of Personality, , 387—423.Leary, T. F. (1957). Interpersonal diagnosis ofprsonality. New York: Ronald.203Lewak, R. W., Marks, P. A., & Nelson, G. E. (1990).Therapist guide to the MMPI and MMPI-2: Providingfeedback and treatment. Muncie, IN: AcceleratedDevelopment Inc.Livesley, W. J. (1986). Trait and behavioral prototypes ofpersonality disorder. American Journal of Psychiatry,143, 728—732.Livesley, W. J. (1987). A systematic approach to thedelineation of personality disorders. American Journalof Psychiatry, 144., 772-777.Livesley, W. J. (1991). Classifying personality disorders:Ideal types, prototypes, or dimensions? Journal ofPersonality Disorders, , 52-59.Livesley, W. J., Jackson, D. N., & Schroeder, M. L. (1989).A study of the factorial structure of personalitypathology. Journal of Personality Disorders, 3, 292-306.Loevinger, J. (1957). Objective tests as instruments ofpsychological theory. Psychological Reports (MonographSupplement 9; whole issue).Lorr, M., & McNair, D. M., (1963). An interpersonalbehavior circle. Journal of Abnormal and SocialPsychology, 1Q, 468-472.Lorr, M., & McNair, D. M. (1965). Expansion of theinterpersonal behavior circle. Journal of Personalityand Social Psychology, a, 823—830.McCrae, R. R. (1991). The five—factor model and itsassessment in clinical settings. Journal of PersonalityAssessment, 57, 399-414.McCrae, R. R., & Costa, P. T., Jr. (l985a). Comparison ofEPI and Psychoticism scales with measures of the five—factor model. Personality and Individual Differences, ,587—597.McCrae, R. R., & Costa, P. T., Jr. (l985b). Openness toexperience. In R. Hogan & W.H. Jones (Eds.),Perspectives in personality (Vol. 1, pp. 145—172).Greenwich, CT: JAI Press.McCrae, R. R., & Costa, P. T., Jr. (l989a). Reinterpretingthe Myers-Briggs Type Indicator from the perspective ofthe five—factor model of personality. Journal ofPersonality, , 17—40.204McCrae, R. R., & Costa, P. T., Jr. (1989b). The structure ofinterpersonal traits: Wiggins’s circumplex and the five—factor model. Journal of Personality and SocialPsycholoav, , 586-595.McCrae, R. R., & Costa, P. T., Jr. (1990). Personality inadulthood. New York: Guilford.McCrae, R. R., Costa, P. T., Jr., & Busch, C. M. (1986).Evaluating the comprehensiveness in personality systems:The California Q-Set and the five-factor model. Journalof Personality, , 430—446.McCrae, R. R., Costa, P. T., Jr., & Piedmont, R. L. (inpress). Folk concepts, natural language, andpsychological constructs: The California PsychologicalInventory and the five—factor model. Journal ofPersonality.McCrae, R. R., & John, 0. P. (Eds.). (in press). Journal ofPersonality (Special Issue on the Five—factor model).McLeiuore, C. W., & Benjamin, L. S. (1979). Whateverhappened to interpersonal diagnosis? A psychosocialalternative to DSM—III. American Psychologist, A, 17-34.McReynolds, P. (1989). Diagnosis and clinical assessment:Current status and major issues. Annual Review ofPsychology, 4.Q, 83-108.Meehi, P. E. (1992). Factors and taxa, traits and types,differences in degree and differences in kind. Journalof Personality, Q, 117—174.Meehl, P. E., & Golden, R. R. (1982). Taxometric methods.In P. Kendall & J.N. Butcher (Eds.), Handbook of researchmethods in clinical psychology (pp. 127-181). New York:Wiley.Mendelsohn, G. A., Weiss, D. S., & Feimer, N. R. (1982).Conceptual and empirical analysis of the typologicalimplications of patterns of socialization and femininity.Journal of Personality and Social Psychology, j, 1157-1170.Miller, T. R. (1991). The psychotherapeutic utility of thefive—factor model of personality: A clinician’sexperience. Journal of Personality Assessment, 51, 415—433.Millon, T. (1981). Disorders of personality. New York:Wiley.205Millon, T. (1983). Millon Clinical Multiaxial InventoryManual (third edition). Minneapolis, MN: InterpretiveScoring Systems.Millon, T. (1987). Manual for the MCMI-II. Minneapolis,MN: National Computer Systems.Millon, T. (1991). Classification in psychopathology:Rationale, alternatives, and standards. Journal ofAbnormal Psychology, IQ..Q, 245—261.Mischel, W. (1968). Personality and assessment. New York:Wiley.Mohr, D. C., Beutler, L. E., Engle, D., Shoham-Solomon, V.,Bergan, J., Kaszniak, A. W., & Yost, E. B. (1990).Identification of patients at risk for nonresponse andnegative outcome in psychotherapy. Journal of Consultingand Clinical Psychology, , 622-628.Morey, L. C. (1991). Personality Assessment Inventory:Professional manual. Odessa, FL: PsychologicalAssessment Resources.Morey, L. C., Waugh, M. H., & Blashfield, R. K. (1985).MMPI scales for DSM-III personality disorders: Theirderivation and correlates. Journal of PersonalityAssessment, 42, 245—251.Muten, E. (1991). Self—reports, spouse ratings, andpsychophysiological assessment in a behavioral medicineprogram: An application of the five-factor model.Journal of Personality Assessment, Z, 449-464.Norman, W. P. (1963). Toward an adequate taxonomy ofpersonality attributes: Replicated factor structure inpeer nomination ratings. Journal of Abnormal and SocialPsychology, , 574-583.Ostendorf, F. (1990). Sprache und Persoenlichkeitsstruktur:Zur Validitaet des Fuenf-Faktoren-Modells derPersoenlichkeit [Language and personality structure:Toward the validation of the five-factor model ofpersonality]. Regensburg: S. Roderer Verlag.Ostendorf, F., & Angleitner, A. (1990). On thecomprehensiveness of the five—factor model ofpersonality: some more evidence for the five robustrating factors in questionnaire data. Paper presented atthe fifth European Conference on Personality, Rome.Paddock, J. R., & Nowicki, S., Jr. (1986). An examination ofthe Leary Circumplex through the Interpersonal CheckList. Journal of Research in Personality, .2.Q, 107-144.206Peabody, D., & Goldberg, L. R. (1989). Some determinants offactor structures from personality—trait descriptors.Journal of Personality and Social Psychology, .i, 552-567.Piedmont, R. L., McCrae, R. R., & Costa, P. T., Jr. (1991).Adjective Check List scales and the five-factor model ofpersonality. Journal of Personality and SocialPsychology, fi, 630—637.Piedmont, R. L., McCrae, R. R., & Costa, P. T., Jr. (1992).An assessment of the Edwards Personal Preference Schedulefrom the perspective of the five—factor model. Journalof Personality Assessment, , 67—78.Pincus, A. L., & Wiggins, J. S. (1990a). Interpersonalproblems and conceptions of personality disorders.Journal of Personality Disorders, 4, 342—352.Pincus, A. L., & Wiggins, J. S. (1990b, August).Interpersonal traits, interpersonal problems, andpersonality disorders: Dual circumplex analyses. Paperpresented at the annual meeting of the AmericanPsychological Association, Boston, MA.Pincus, A. L., & Wiggins, J. S. (in press). An expandedperspective on interpersonal assessment. Journal ofCounse1inc and Development.P10mm, R, & Daniels, D. (1987). Why are children in thesame family so different from one another? Behavior andBrain Science, jQ, 1—16.Popper, K. R. (1959). The logic of scientific discovery.New York: Basic Books.Romney, D. M., & Byner, J. M. (1989). Evaluation of acircumplex model of DSM-III personality disorders.Journal of Research in Personality, 21, 525-538.Rushton, J. P., Fulker, D. W., Neale, M. C., Nias, D. K. B.,& Eysenck, H. J. (1986). Altruism and aggression: Theheritability of individual differences. Journal ofPersonality and Social Psychology, Q, 1192-1198.Safran, J. D., & Segal, Z. V. (1990). Interpersonal processin coanitive therapy. New York: Basic Books.Schroeder, M. L., Wormworth, J. A., & Livesley, W. J.(1992). Dimensions of personality disorder and theirrelationship to the Big Five dimensions of personality.Psychological Assessment: A Journal of Consulting andClinical Psychology, 4, 47-53.207Shapiro, D. (1989). Psychotherapy of neurotic character.New York: Basic Books.Shostrom, E. L. (1974). Personal Orientation Inventory.San Diego: EdITS.Sim, J. P., & Romney, D. M. (1990). The relationshipbetween a circumplex model of interpersonal behaviors andpersonality disorders. Journal of Personality Disorders,4, 329—341.Spielberger, C. D. (1983). Manual for the State—TraitAnxiety Inventory. Palo Alto, CA: ConsultingPsychologists Press.Strack, S. (1987). Development and validation of anadjective checklist to assess the Millon personalitytypes in a normal population. Journal of PersonalityAssessment, 51, 572—587.Strack, S. (1990). Manual for the Personality AdjectiveCheck List. Richmond, WA: Pacific Psychological.Strack, S., Lorr, M., & Campbell, L. (1990). An evaluationof Millon’s circular model of personality disorders.Journal of Personality Disorders, 4, 351-361.Sullivan, H. S. (l953a). Conceptions of modern psychiatry.New York: Norton.Sullivan, H. S. (1953b). The interpersonal theory ofpsychiatry. New York: Norton.Talley, P. F., Strupp, H. H., & Morey, L. C. (1990).Matchmaking in psychotherapy: Patient-therapistdimensions and their impact on outcome. Journal ofConsulting and Clinical Psychology, , 182-188.Tellegen, A. (1985). Structures of mood and personality andtheir relevance to assessing anxiety, with an emphasis onself-report. In A.H. Tuma & J.D. Maser (Eds.), Anxietyand the anxiety disorders (pp. 681—706). Hillsdale, NJ:Earibaum.Tellegen, A. (1991). Personality traits: Issues ofdefinition, evidence, and assessment. In D. Ciccetti andW. Grove (Eds.), Thinking clearly about psychology:Essays in honor of Paul Everett Meehl (Vol. 2, pp. 10-35). Minneapolis MN: University of Minnesota Press.Tellegen, A., Lykken, D. T., Bouchard, T. J., Wilcox, K. J.,Segal, N. L., & Rich, S. (1988). Personality similarityin twins reared apart and together. Journal ofPersonality and Social Psychology, .A, 1031—1039.208Trapnell, P. D., & Wiggins, J. S. (1990). Extension of theInterpersonal Adjective Scales to include the Big Fivedimensions of personality. Journal of Personality andSocial Psychology, , 781-790.Tunis, S. L., Fridhandler, B. N., & Horowitz, M. J. (1990).Identifying schematized views of self with significantothers: Convergence of quantitative and clinicalmethods. Journal of Personality and Social Psychology,, 1279—1286.Tupes, E. C., & Christal, R. E. (1958). Stability ofpersonality trait rating factors obtained under diverseconditions. USAF WADC Technical Note, No. 58-61.Tupes, E. C., & Christal, R. E. (1961). Recurrentpersonality factors based on trait ratings. USAD ASDTechnical Report, No. 61-97.Van Denburg, T. F., Schmidt, J. A., & Kiesier, D. J. (inpress). Interpersonal assessment in counseling andpsychotherapy. Journal of Counseling and Development.Wailer, N. G., & Ben—Porath, Y. (1987). Is it time forclinical psychology to embrace the five—factor model ofpersonality? American Psychologist, 4,, 887-889.Watson, D., & Clark, L. A. (in press). On traits andtemperament: General and specific factors of emotionalexperience and their relation to the five—factor model.Journal of Personality.Watson, D., & Pennebaker, J. (1989). Health complaints,stress, and distress: Exploring the role of negativeaffectivity. Psychological Review, , 234-254.Weinberger, D. A. (1989). Social—emotional adjustment inolder children and adults: I. Psychometric properties ofthe Weinberger Adiustment Inventory. Manuscriptsubmitted for publication.Weinberger, D. A. (in press). The construct validity of therepressive coping style. In J.L. Singer (Ed.),Repression and dissociation: Implications forpersonality theory. psychopathology, and health.Chicago, IL: University of Chicago Press.Weinberger, D. A., Feldman, S. S., & Ford, N. E. (1989).Social—emotional adjustment in older children and adults:II. External validation of subdimensions of distress andrestraint. Manuscript submitted for publication.209Weinberger, D. A., & Schwartz, G. E. (1990). Distress andrestraint as superordinate dimensions of self—reportedadjustment: A typological perspective. Journal ofPersonality, , 381—417.Weinberger, D. A.., Tublin, S. K., Feldman, S. S., & Ford, N.E. (in press). Preadolescents’ social—emotionaladjustment and selective attrition in family research.Child Development.Widiger, T. A. (in press). The DSM—III—R personalitydisorder diagnoses: A critique and an alternative.Psychological Inquiry.Widiger, T. A., & Trull, T. J. (in press). Personality andpsychopathology: An application of the five-factormodel. Journal of Personality.Wiggins, J. S. (1968). Personality structure. AnnualReview of Psychology, , 293-350.Wiggins, J. S. (1973). Personality and prediction:Principles of personality assessment. Reading, MA:Addison-Wesley.Wiggins, J. S. (1979). A psychological taxonomy of trait-descriptive terms: The interpersonal domain. Journal ofPersonality and Social Psychology, fl, 395-412.Wiggins, J. 5. (1980). Circumplex models of interpersonalbehavior. In L. Wheeler (Ed.), Review of personality andsocial psychology (Vol.. 1, pp. 265-294). Beverly Hills:Sage..Wiggins, J. S. (1982). Circumplex models of interpersonalbehavior in clinical psychology. In P.C. Kendall & J.N.Butcher (Eds..), Handbook of research methods in clinicalpsychology (pp. 183-221). New York: Wiley.Wiggins, J. S. (1985). Symposium: Interpersonal circumplexmodels (1948-1983). Journal of Personality Assessment,, 626—631.Wiggins, J. S. (1991). Agency and communion as conceptualcoordinates for the understanding and measurement ofinterpersonal behavior. In W. Grove & D. Cicchetti(Eds.), Thinking clearly about psychology: Essays inhonor of Paul Everett Meehi, (Vol. 2, pp. 89-113).Minneapolis, MN: University of Minnesota Press.Wiggins, J. S. (in press). Have model, will travel..Journal of Personality.210Wiggins, J. S., & Broughton, R. (1985). The interpersonalcircle: A structural model for the integration ofpersonality research. In R. Hogan & W. H. Jones (Eds.),Perspectives in personality, (Vol. 1, pp. 1-47).Greenwich, CT: JAI Press.Wiggins, J. S., & Broughton, R. (1991). A geometrictaxonomy of personality scales. European Journal ofPersonality, •, 343—365.Wiggins, J. S., Phillips, N., & Trapnell, P. D. (1989).Circular reasoning about interpersonal behavior:Evidence concerning some untested assumptions underlyingdiagnostic classification. Journal of Personality andSocial Psychology, , 296-305.Wiggins, J. S., & Pincus, A. L. (1989). Conceptions ofpersonality disorders and dimensions of personality.Psychological Assessment: A Journal of Consulting andClinical Psycholoay, ], 305-316.Wiggins, J. S., & Pincus, A. L. (1992). Personality:Structure and assessment. Annual Review of Psychology,4., 473—504.Wiggins, J. S., & Pincus, A. L. (in press). Personalitystructure and the structure of personality disorders. InP. T. Costa Jr. & T. A. Widiger (Eds.). PersonalityDisorders and the Five—Factor Model of Personality.Washington, DC: American Psychological Association.Wiggins, J. S., Steiger, J. H., & Gaelick, L. (1981).Evaluating circumplexity in personality data.Multivariate Behavioral Research, j, 263-289.Wiggins, J. S., & Trapnell, P. D. (in press). Personalitystructure: The return of the Big Five. In S.R. Briggs,R. Hogan, & W.H. Jones (Eds.), Handbook of personalitypsychology. Orlando, FL: Academic Press.Wiggins, J. S., Trapnell, P. D., & Phillips, N. (1988).Psychometric and geometric characteristics of the revisedInterpersonal Adjective Scales (IAS-R). MultivariateBehavioral Research, 2, 517—530.Zuckerman, M. (1971). Dimensions of sensation—seeking.Journal of Consulting and Clinical Psychology, , 45-52.Zuckerman, M., Bone, R. N., Neary, R., Mangelsdorff, D., &Brustman, B. (1972). What is the sensation seeker?Personality trait and experience correlates of thesensation seeking scales. Journal of Consulting andClinical Psvcholoav, , 308—321.211Zuckerman, M., Kuhlman, D. M., & Camac, C. (1988). Whatlies beyond E and N? Factor analyses of scales believedto measure basic dimensions of personality. Journal ofPersonality and Social Psychology, , 96-107.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items