Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The Profile of Narcissistic Dispositions (POND): development and validation Taylor, Candace Margo 1995

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

Item Metadata

Download

Media
831-ubc_1996-091708.pdf [ 10.39MB ]
Metadata
JSON: 831-1.0087680.json
JSON-LD: 831-1.0087680-ld.json
RDF/XML (Pretty): 831-1.0087680-rdf.xml
RDF/JSON: 831-1.0087680-rdf.json
Turtle: 831-1.0087680-turtle.txt
N-Triples: 831-1.0087680-rdf-ntriples.txt
Original Record: 831-1.0087680-source.json
Full Text
831-1.0087680-fulltext.txt
Citation
831-1.0087680.ris

Full Text

T H E P R O F I L E O F N A R C I S S I S T I C DISPOSITIONS  (POND)  DEVELOPMENT AND VALIDATION by CANDACE MARGO TAYLOR  B . A . , The University of British Columbia, 1982 M . A . , The University of British Columbia, 1986  A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F  T H E REQUIREMENTS FOR T H E D E G R E E OF  DOCTOR OF PHILOSOPHY  in  T H E F A C U L T Y OF G R A D U A T E STUDIES  Department of Psychology W e accept this thesis as conforming to the required standard  T H E UNIVERSITY O F BRITISH C O L U M B I A October 1995 © Candace Margo Taylor,  1995  In  presenting  degree at the  this  thesis in  University of  partial  fulfilment  of  N  of  department  this thesis for or  by  his  or  requirements  British Columbia, I agree that the  freely available for reference and study. I further copying  the  representatives.  an advanced  Library shall make  it  agree that permission for extensive  scholarly purposes may be granted her  for  It  is  by the  understood  that  head of copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department The University of British ColumJ Vancouver, Canada  DE-6 (2/88)  i i  ABSTRACT Four studies were conducted with the aim of developing a measure of narcissism that, unlike previous measures, is not inherently pathological. In Study 1, the NPI—the closest approximation to such a measure—was administered to a large-scale sample. T w o separate Principal Component Analyses (PCAs) failed to replicate its reported structure, and revealed further psychometric problems. Use of an alternative item format (Likert ratings) was shown to yield much higher reliabilities than the original forced-choice format. Using the Likert item-format, the Profile of Narcissistic Dispositions ( P O N D ) was developed in Study 2. The items were based on a comprehensive analysis of the literature on "normal" narcissism. Five reliable subscales emerged from an oblique factor analysis. A l l five loaded substantially on the first unrotated principal component. The relationship of the P O N D to established self-report measures was also explored. In Study 3, the P O N D ' s structure was replicated and empirical relations were expanded: In particular, o  the P O N D showed negative correlations with various self-reports of psychopathology. In Study 4, the P O N D was shown to predict peer ratings of narcissism. Further peer-ratings elaborated the character of normal narcissism, that is, an interpersonal style that is marked by a dominant and secure but disagreeable social presence.  iit  Table of Contents  Abstract  ii  Table of Contents  iii  List of Tables  iv  List of Figures  v  Acknowledgement Introduction/ Literature Review Theoretical perspectives The Measurement of narcissism The Present studies Study 1:What can we learn from the NPI?  vi 1 7 22 55 57  Method  62  Results and Discussion  63  Conclusions  76  Study 2 :Scale development and initial correlates  80  Method  83  Results  91  Discussion Study 3: Extending the nomological network  105 110  Method  112  Results and Discussion  114  Study 4: Additional correlates and peer-ratings  130  Method  134  Results and Discussion  137  New self-report variables  140  Peer ratings and the P O N D  142  Conclusion  149  General Discussion Future directions  150 174  References  179  Tables  190  Figures  212  Appendix I  214  IV  List of Tables Study 1. Table la:  Principal components analysis of the matrix of NPI inter-item tetrachoric correlations.  185  Table lb:  Intercorrelations among the NPI (tetrachoric-based) components.  186  Table 2a:  Principal components analysis of the matrix of NPI inter-item phi coefficients.  187  Table 2b:  Intercorrelations among the NPI (phi-based) components.  188  Table 3:  Principal components analysis of the 40 Likert-rated narcissistic  189  alternatives of the NPI dyads Table 4a:  Normative statistics and reliabilities.  190  Table 4b:  Increments in alpha reliabilities over those obtained by Raskin and Terry  191  (1988). Study 2. Table 5:  Principal components analysis of the 65 P O N D - P items: Oblique rotation.  192  Table 6:  Final 52 P O N D scale items.  194  Table 7:  Intercorrelations, means, standard deviations, age and gender correlations and coefficients alpha.  196  Table 8:  Correlations of the P O N D scales with various self-report measures  197  Study 3. Table 9:  Principal components analysis of the P O N D items: Oblique rotation.  198  Table 10:  Descriptive statistics and principal components analysis of P O N D subscales: Unrotated.  199  Table 11:  Correlations of the P O N D scales with various self-report measures.  200  Study 4. Table 12:  P O N D scale intercorrelations, means, standard deviations, and coefficients alpha.  201  Table 13:  Correlations of P O N D scales with selected self-report measures.  202  Table 14:  Correlations of P O N D scales with peer ratings of narcissism.  203  Table 15a:  Correlations of P O N D scales with performance-related ratings.  204  Table 15b:  Correlations of P O N D scales with various interpersonal ratings.  205  V  List of Figures  Figure 1:  Projection of the P O N D scales onto the circumplex.  186  ACKNOWLEDGEMENTS I should like to attempt to estimate what are typically referred to as the "inestimable contributions" of a number of persons; the following order in no way reflects the results of that analysis. First, I'd like to thank Craig McCullough for giving my S O the newest Joshua Redman C D ; as it turns out, he generates excellent music to accompany writing. In a similar vein, I should also like to thank Wolfy, Grieg, Vivaldi (and the other classical guys), Tony Bennett, Marvin Gaye and George Benson (themselves kind of classical guys); also, the Gypsy Kings, who by thoughtfully singing in Spanish, didn't disrupt my lucubration. A n d thanks to Jerry Wiggins, for getting me an electric pencil sharpener; my former quill sharpening was getting a little taxing. Thanks for the patience and general peachiness of Jennifer Campbell and Darrin Lehman. A n d especial thanks to Del Paulhus, for: a) forcing me to learn how to process words (someday, I'm sure my thanks will be less grudging); b) his tremendous assistance with analyses, tables, (usually) judicious redacting and frequent atticisms; and c) most importantly, for his 'stand by your student' approach; when he agreed to be in it for the long haul, he probably didn't suspect that I'd take it literally. Finally, many thanks to H R (no, not Holt Renfrew): For his computer, his mastery of Harvard Graphics, and his endless psychic handholding.  CMT  1  To love oneself is the beginning of a life-long romance. At first blush, Oscar Wilde's observation appears a perfect—and typically a r c h — characterization of narcissism. The more pedantic among us might inquire, however, as to his intended referent for romance. Those familiar with Wilde's cynical rejection of the heartsand-flowers variety—so dear to lesser (and more ingenuous) mortals—would likely suspect that his meaning might be rather more oblique. If one consults the ultimate arbiter of such matters (i.e., the Oxford English Dictionary), several different shadings are revealed: 1) The vernacular language of France (whose citizens, it might be noted, are said to know a thing or two about the subject at hand); 2) A tale in verse, embodying the adventures of some hero of chivalry; 3) A fictitious narrative...of which scenes and incidents are very remote from ordinary life; 4) A Spanish historical ballad; 5) That class of literature which consists of romances; and finally, 6) A n extravagant fiction, invention, or story; a wild or wanton exaggeration; a picturesque falsehood (as in "This is romance—I'll not believe a word on't"; Dryden & Newcastle, 1667; O E D , 1933/71, pp. 766-7.) Once again, it seems, the poets have said (or at least alluded to) it first. For in what one might infer as Wilde's double-entendre use of "romance," particularly definition '6', he closely anticipates the current D S M - I V ( A P A , 1994) definition of narcissism. That decretal source states that narcissists, i.e., those exhibiting a Narcissistic Personality Disorder (NPD) are characterized by, among other things, a grandiose sense of self-importance, an exaggerated sense of their abilities and achievements and a preoccupation with fantasies of unlimited success. Wilde's observation is similarly echoed in another D S M - I V criterion for N P D , i.e., that the relevant behaviors, dispositions, etc., be "characteristic of the individual's current and long-  2  term functioning," in other words, life-long. T o receive a diagnosis of N P D , however, such traits must cause "either significant impairment in social or occupational functioning or subjective distress." In this proposal, I should like to argue that narcissism need not be disabling or dysfunctional, and may, to the contrary, be a highly adaptive state of psychological affairs. This assertion, however, seems a little temerarious when one discovers that historically, it has rarely been construed as tenable.  Consider, to begin, the mythically-derived eponym (f. Gk. Narkissos), familiar initially as a rather ignominious object lesson in which the beauteous youth, overly enamoured with his self-reflection in a pond, lost his soul to the spirits lurking within. The alleged perils of selflove reemerged in the eighteenth century, in the philosopher VonSchubert's (1780-1860) conception of Ich sucht (self-love) as the agent that wrested man from his originally harmonious union with nature. Nineteenth century Neo-Romanticists proved somewhat kinder, however, elevating narcissistic self-absorption to near-paragon status: Worship of the individual, even to the extreme of interpersonal isolation, became the paramount symbol and incarnation of that period (i.e., c. 1885; Ellenberger, 1970). A n d perhaps not too surprisingly, the quiddities of the fin de siecle rendered a reconceptualization of self-love, in which one finds it reduced to the status of a sexual perversion or deviation. Binet, for example, likened fetishists who chose themselves as preferred sex object to the myth of Narcissus. Ellis (1898) used the term similarly to denote a sexual deviation, as did Nacke (1899), in his discussion of the type Narcissimus, in Die sexuellen Perversitaten.  A decade and one half later, Freud (1914) proposed "a primary and normal narcissism" (p. 31), a stage of auto-eroticism intrinsic to normal (albeit potentially precipitating deviated)  3  psychosexual development. Later, Freud (1931) postulated narcissism as one of the "pure" libidinal types, delineating its virtues yet (somewhat sheepishly, one imagines) allowing how its characteristics are, "in the main negatively described" (p. 247). Such an aspect pervades later clinical conceptions of narcissism, viz. the "narcissistic character disorder" (Schafer, 1948), or Mayman's (1955) discussion of the "narcissistic character," each redolent with what Allport (1937) has abjured in his interpretation of "character" as personality evaluated.* Both of these are, however, logical precursors to the current psychiatric view of the N P D noted above.  A n d while Murray (1955) alluded to the "narcistic [sic] core in everyman" (p. 632), the received view of narcissism as intrinsically negative, even aberrant, is the prevailing one in most current literatures (but cf. Block, 1971; Cattell, 1957; Kohut, 1971; Wink, 1992). Some (e.g., Lasch, 1979) go even further, describing (and decrying) a sort of creeping narcissism that is seen as increasingly characteristic of contemporary, self-oriented society (Wallach & Wallach, 1983). In short, the concept of narcissism has, in the main, received rather bad press.  In the DSM-III R introduction to the Axis II (personality) disorders, the writers assert that it is only when "personality traits are inflexible and maladaptive and cause either significant functional impairment or subjective distress that they constitute Personality Disorders" ( A P A , 1987, p. 335). The same source further espouses a view of such traits as forming a continuum, ranging from their normal expression (flexible and adaptive) to those considered abnormal and psychopathological, as with the personality disorders (which, one might note, have a fairly low incidence and prevalence).  In contrast to personality as character devaluated.  4 Yet in the corpus of the literature on narcissism, one is hard pressed to find discussion of the "other end" of the continuum, i.e., what form a "normal" narcissism—should it exist— might take. Thus in order to abstract some sense of the probable nature of such a construct, a kind of backward reasoning is required: A s Freud remarked in On Narcissism, "Once more, in order to arrive at what is normal and apparently so simple, we shall have to study the pathological with its distortions and exaggerations" (1914, p. 39).  While something of a Hobson's choice, Freud's prescription evidently underlies the most visible measure of what putatively, is normal, nonpathological narcissism, i.e., Raskin's Narcissistic Personality Inventory (NPI; Raskin & Hall, 1979; Raskin & Terry, 1988). But although even that measure was designed to gauge individual differences in narcissism as represented in normal, or non-clinical populations, its "conceptual template" (Raskin & Terry, 1988, p. 892) was the DSM-III criteria for the narcissistic personality. Raskin's approach 2  may, faute de mieux, be the only logical route toward understanding non-pathological narcissism. But as Allport (1937) has inquired, "Is the normal personality simply an undistinguished edition of the mentally diseased?" (p. 26). I should like to argue—and ultimately demonstrate—that it is not.  What I propose is, of course, not entirely new. Leary (1957), for example, wrote of an interpersonal style characterized by "competitive self-confident narcissism," one that in its adaptive mode is "a most impressive social maneuver" (p. 332). Similarly, Emmons (1984), in his examination of the correlates of the NPI (e.g., self-esteem), allowed how "one might conclude that being narcissistic isn't so maladaptive after all" (p. 298). But despite a  While the NPI was indeed based upon DSM-III, the criterial alterations to DSM-IV (see p. 16) in no way obviate my criticisms of the DSM-based provenance of the NPI.  2  5  burgeoning literature on narcissism, a view of it as a positive, agentic style has yet to fully emerge. I believe that in spite of its overtly clinical provenance (and concomitant features of maladaptiveness), the interpersonal style captured by the NPI and the alternative inventory proffered herein might instead reflect, for the most part, a normal or non-pathological narcissism, insofar as that construct is exemplified by competent, competitive, but always assured self-promotion; the French call it Vamour propre, a self-esteem that "is ready with its claims" ( O E D , p. 72). The core of this presentation is a description of the development and initial validation of a measure designed to assess such a construct. Toward that end, I begin with a brief overview of the theoretical literature on narcissism. I note that I provide this review not as I intend to focus on etiological issues (i.e., the "causes" of narcissism), but rather to acquaint the reader with the major theoretical perspectives that have given rise to the contemporary received view of the nature of narcissism—pathological or otherwise. A n d further, despite my disclaimer of etiological nonpartisanship, it strikes me that consideration of theoretical issues is important insofar as these same etiological viewpoints have in large part also determined the direction of most descriptive, i.e., assessment-oriented approaches to narcissism (e.g., D S M - I V ) . Even though one may take issue with the undiscerning assumption that clinically-based theoretical approaches are suitable templates for the study of non-psychiatric narcissism, even the most objective personality assessment programmes need not proceed in an atmosphere of blind empiricism. A construct-oriented approach (e.g., Loevinger, 1957) in fact requires that the procedures involved in test construction and validation (e.g., defining the universe of content) must  necessarily be theoretically informed. So while the final arbiter on the relevance of theoretical/etiological viewpoints is, of course, objective-empirical assessment, they must be considered in even what is the ultimately descriptive, albeit construct-oriented approach to narcissism adopted here. Following an examination of the literature on theories of narcissism, I then present some representative examples of attempts to psychometrically capture the construct. A s noted earlier, chief among these is Raskin's NPI, and as such, it merits particularly close scrutiny. I will review available evidence bearing upon the validity of the Raskin measure, paying particular attention to issues concerning its psychometric characteristics. I next describe an extended investigation of those characteristics in which I examine the internal (factorial) structure of the instrument, test the properties of an alternative item format, and gauge the degree of convergence (i.e., structurally) between the original forced-choice and revised Likert-type item formats. The evidence presented here suggests that although the NPI adequately captures one circumscribed view of the construct it purports to measure, unresolved questions pertaining to its degree of structural stability and somewhat limited theoretical scope may ultimately restrict its status to that of an instrument  manque for the assessment of non-pathological narcissism.  Accordingly, the main corpus of the current research is a description of the procedures concomitant to the development and initial validation of what I believe is a promising alternative to the NPI—the Profile of Narcissistic Dispositions  (POND).  7  Theoretical Perspectives Freud When Murray (1938) wrote that "a personality is largely revealed in the objects that it cathects (values or rejects)," he noted that a primary candidate for positive cathection "is the self or Ego—firstly and perhaps lastly beloved" (p. 106). Though differing in their particulars, many prominent theoreticians of narcissism (e.g., Freud, 1911/1914, 1931; Kernberg, 1975; Kohut, 1971; Reich, 1938) conceive of its etiology as emanating from early libidinal investment in the self as love object. Freud's initial formulation (1911/1914), for example, was of narcissism as one possible sequence antecedent to the infant's initial auto-eroticism, a stage in which auto-erotic gratifications are experienced contiguously with "vital functions in the service of self-preservation" (p. 44). Narcissism, he argued, arose when some "disturbance"  3  in libidinal development led to cathection of self, rather than the mother or caregiver, as the love-object. He maintained, however, that all persons share two original sexual/love objects: Oneself and one's mother, and thereby postulated "a primary and normal narcissism in everyone, which may in the long run manifest itself as dominating...object choice" (p. 45). When appropriate object love is not deflected, the course of this primary narcissism is illuminated, he argued, by examining the attitude of "fond parents," an attitude that reflects "a revival and reproduction of their own, long since abandoned [reluctantly, one infers] narcissism" (p. 48).  Thus, according to Freud, parental attitudes toward the once and future narcissist are characterized by overestimation:  The nature of this disturbance is never clearly elucidated.  8  "They are impelled to ascribe to the child all manner of perfections which sober  reflection  would  not  confirm, to  gloss  over  and forget  all  his  shortcomings...they are inclined to suspend in the child's favor the operation of all those cultural acquirements which their own narcissism has been forced to respect, and to renew in his person the claims for privileges which were long ago given up by themselves." (p. 48) This theme of nascent narcissist as the ineluctable result of parental over-valuation, is, as shall soon become apparent, one that emerges frequently in the etiological notions of many writers on narcissism, notably Horney (1939), Millon (1969; 1981) and Murray (1938; 1955). Because Freud postulated primary narcissism as a normal stage in libidinal development, the nature of an adult narcissistic character was not, as such, initially described. Numbered among 1931's Libidinal Types, however, was a second formulation he characterized as narcissistic—not, it should be noted, in terms of a clinical entity, but rather as a "pure" psychological type, one of three from which an entire spectrum of personality "blends" could be derived. In fact, he categorically stated that such types "must not coincide with specific clinical pictures," but rather, "should embrace all the variations which according to our practical standards fall within the category of the normal" (italics added; pp. 247-8). A qualification, however, was that in extreme cases, the types could resemble clinical manifestations, and could thus, "help to bridge the gap which is assumed to exist between the normal and the pathological" (p. 248). A s it turns out, infancy's primary narcissism is father to that of man, marked by the same central focus on self-preservation. In the adult, however, this is manifest in a type that is "independent, and not easily overawed," qualities that strike one as desirable, and certainly culturally valued:  "People of this type impress others as being "personalities"; it is on them that their fellow-men are specially likely to lean; they readily assume the role of  9  leader, give fresh stimulus to cultural development or break down existing conditions." (p. 249) The strongly agentic nature of this formulation is, as Millon (1981) points out, sharply dissonant with the more contemporary psychoanalytic formulations of, for example, Kohut and Kernberg—positions that stress fragile rather than robust self-esteem, weakness rather than strength, and so on. It is, however perfectly consonant with the thesis of this presentation, namely that in its normal complexion, narcissism comprises characteristics of healthy self4  esteem, ascendance, and self-assured dominance. Horney A s I alluded to earlier, however, this focus on adaptive narcissism appears to be the exception rather than the rule in the theoretical literature on the construct. Horney (1939), for example, specifically rejects the Freudian view of both narcissistic and normal self-regard as "desexualized derivatives" of earlier libidinal investment. She argues instead that "narcissistic trends" are, like all other neurotic trends, sequellae of disordered childhood interpersonal relationships (see also Sullivan, 1953). Horney's enumeration of parental attitudes said to foster such trends seems rather a 'mixed bag', among them: Authoritarianism; ostensibly "self-sacrificing" demeanour; delivering direct blows to self-esteem; transfer of (perhaps thwarted) parental ambitions to the child (e.g., "boy as embryonic genius, girl as princess," i.e., overvaluation; recall Freud, p. 4); parental devaluation; ostensive preference for one sibling such that the "unfavored" one must concentrate on outshining the other; and so forth. This rather dizzying array of etiologic candidates is said to have a common outcome, however, in that the child comes to feel that in  As opposed to pathological or clinical characterizations.  10  order to win love and acceptance, he must mold himself to the prescriptions of others. Submission to the omnipotent parental will is the only way to effect this, and as a result, the child comes to entirely lose the capacity to objectively assess his personal value.  The concomitant—and indeed the essence of the narcissistic trend—is neurotic selfinflation qua coping mechanism, where, according to Horney (1939) "Psychic inflation, like economic inflation, means presenting greater values than really exist" (p.89). The child is said to construct a romantic fantasy world, featuring the self as hero (with appropriately heroic qualities); the implicit message being that 'even if you won't love or respect me, you have to at least pay attention to me, and admire me for my outstanding qualities.' The neuroticallysought admiration becomes, according to Horney, a substitute for actual love, and forms the narcissist's primary gratification (see also Machover, 1955). Self-inflation is thus adopted as the preferred security operation as it successfully promotes the illusion that one is strong and omnipotent—above the fray, as it were.  5  So while the narcissistic trend is described as borne of the need to cope with the basic anxiety engendered by dysfunctional early relationships—a trend which, it is argued, can be overcome if early influences "are not too decisive"—an alternative and perhaps more insidious route to narcissism is postulated. Anticipating Lasch's (1979) view of narcissism, Horney notes that  Narcissistic trends are frequent in our culture. More often than not people are incapable of true friendship and love; they are egocentric, that is, concerned with their security, health, recognition; they feel insecure and tend to overrate their personal  This conceptualization is strikingly similar to Reich's (1938) notion of the function of narcissistic "character armour."  5  11  significance; they lack judgement of their own value because they have relegated it to others (p.98). Horney further argues that such "typical[ly] narcissistic features are by no means restricted to persons who are incapacitated by neuroses" alleging instead that striving for status and prestige—pursuits considered by some (e.g., Buss, 1988; Hogan, 1983) as immanent to human behavior—is derogated merely as "a means of overcoming fears and inner emptiness", means she notes that are, nonetheless, "certainly culturally prescribed" (p.98). Yet despite the prevalence of such characteristics, Horney argues—contra Freud and a legion of others who postulate a continuum between normal and abnormal expression of traits—that normal self-esteem and narcissistic self-aggrandizement differ qualitatively rather than quantitatively. She asserts: "True self-esteem rests on qualities which a person actually possesses, while self-inflation implies presenting to the self and to others qualities and achievements for which there is no adequate foundation" (p.99). Thus in Horney's view, the narcissistic person simply does not have the basic goods to back up his self-presentation, or, at very best, has them in lesser inventory than he or she advertises.  Murray The romantic view of self and one's (overvalued) abilities and achievements is also reflected in Murray's (1938, 1955) conception of narcissism (or Narcism, as he preferred)— albeit in a somewhat more benign form. Recall that Murray alluded to the "narcistic core in everyman", an egophilia or self-love whose genesis was a normal, universal concomitant of the urethral-phallic stage of development. Like that of Freud, Murray's stage theory similarly  12  embraced the possibility of development going awry or becoming fixated, failing to progress to more mature modes of relation to the world, i.e., object (other than self) love. In Murray's system, fixation at the urethral-phallic stage could precipitate diverse forms of narcissism (among other complexes), the most generic characterized by " A n unusual attention to one's body, feelings and thoughts and a narrow devotion to one's interests, disregarding the well-being of others" (p. 108). Associated features of narcissism include what Murray terms direct (e.g., self-absorption, self-admiration, superiority feelings and delusions of grandeur) and indirect (e.g., ruthless self-seeking, object deprecation, and egocentricity) manifestations. (Those familiar with the DSM-III R and IV descriptions of N P D might experience a slight deja vu at this). Cynosural narcissism, in which the self is similarly the shining star, differs somewhat but also involves an all-embracing obsession with unsolicited attention and adoration, a "desire to attract and enchant all eyes" (1955, p.633); this is consistent with one clinician's observation that the T A T protocols of narcissists frequently feature themes of unlimited success and applause from the crowd (Holler, 1981; personal communication).  The nascent quasar's emergent need is eloquently wrought by Murray:  "It is first supremely gratified at that epiphanal moment of babyhood when the grandparents and relatives arrive at the cradle, with gifts perhaps, to beam with wonder at this new emergence of pure potentiality, pure Being..." (p.633) [See also Benjamin (1987) on the baby narcissist who has "learned to expect a brass band" everytime he ties his shoes (p.45).] A n d typically, Murray noted, cynosural narcism is "almost inevitably" fused with ascensionism, "the wish to overcome gravity, stand tall, rise"; a wish frequently expressed, emotionally and ideationally, in "passionate enthusiasm, rapid elevations of confidence, exultation [and] likely to be expressed in the imagery of physical  13  ascensionism" (p. 631). A more mundane instantiation of this fusion (a la Horney) is in the seeking of upward mobility, or a fast and impressive increase in prestige. For, as Murray notes, "The way to attract all eyes is to be very tall, to stand erect above the multitude, and best of all to rise in the air like a god" (p. 633). Murray provides the example of "American Icarus", a subject (characterized by an integrate of cynosural narcism, ascensionism and fear of falling, i.e., the Icarus complex) whose existence is a barely contained anticipation of the day when "his 'soul' ignites and his inner fire will send him hurling (two rungs at a time) up the ladder of success" (p.633). A n d while, as Murray notes elsewhere, this subject's rather immodest ambitions are fueled with "little reliance... on will power, discipline, industry, or conscientiousness"—incidentally, a near-identical anticipation of the DSM-III R (and IV) N P D criterion of entitlement—he also notes that this type of narcissistic dynamic  can  precipitate actual achievement. A n d while Murray cites possible negative examples such as messianic fervour and prophecy, he also enumerates physical/athletic achievements such as acrobatics and high jumping (!) and verbal/social skills, among them singing, acting, "ardent romantic poetry" (e.g., Byron, Shelley, perhaps Wilde?), and charismatic oratory and leadership. Thus despite the possibly "unhealthy" provenance of narcissism, Murray's view, not dissimilar to that of Freud on sublimation , and indeed narcissism, is that competence, confidence and agency can be "healthy" narcissistic sequellae of early developmental experience. This is perhaps most serendipitiously illustrated in Murray's own rich and productive life; in a recent obituary of that great personologist, Smith & Anderson (1989) recalled the following:  "Harry Murray liked to refer to his "narcissism," which was certainly real in most senses of that abused word; however, it was not only compatible with, but  14  it also supported, his inspiring example to the many students and colleagues whose lives he touched." (p. 1154) Clearly, Murray's conception and example of narcissism stands in sharp contrast to that of Horney.  Leary That narcissism can be agentic and positive is similarly reflected in Leary's (1957) view of that particular style of interpersonal behavior (see Sullivan, 1953, for the conceptual grounding of Leary's theory). At the same time, however, Leary carefully delineated how initially normal, self-confident narcissism can, when rigidly deployed as an interpersonal reflex, easily degenerate into a maladaptive style. For while, as noted earlier, in its adaptive intensity, it is "a most impressive social maneuver",  "In its maladaptive extreme it becomes a smug, cold, exploitive social role. In this case the adaptive self-confidence and independence become  exaggerated  into a self-oriented rejection of others. The individual is so rigidly tied to his own self-enhancement that he fails to sense the inappropriateness of his own behavior." (p. 332) This chronic failure to sense the nature of one's own interpersonal impact is, of course, characteristic of pathological, not normal narcissism. Note, in contrast, Leary's characterization of the rewards that accrue to self-confident narcissism in its "adaptive intensity": "The person who bases his security on overt independence is comforted by the satisfaction in flexing his muscles, admiring his own strength or beauty or wisdom, and reveling in his advantages over those whom he perceives as inferior. Adaptively self-confident individuals receive considerable admiration and social approval." (Leary, 1957, pp. 333-4) Competition and self-confident independence—be they conventionally appropriate or n o t — are traits that are nonetheless positively sanctioned in contemporary Western society; it is only  15  when the behavioral manifestation of these attributes becomes rigidly inflexible, totally dominating the interpersonal style, that narcissism becomes maladaptive. So according to Leary, the individual who sensitively—and appropriately—deploys competitive, self-enhancing behaviors "wins the admiration and flattering envy of others" (p.334). Conversely, the maladaptively narcissistic person, rather than demonstrating his independent ascendance, is self-enhanced or gratified only through demonstrating his "superiority" relative to "weakness" in others, primarily, Leary suggests, because of an atavisitic fear of dependence and vulnerability. Leary draws a parallel between the managerial personality and the competitive one: Whereas the former "pulls" or elicits a sort of willing submission from others, in that recognition of his strength engenders an easy yielding to his authority, the latter is instead cold and exploitative in his attempts to wrest authority from, and establish superiority over, others. Unlike the managerial type's neutral or beneficent use of acquired power, the narcissist implies that his "status and strength will be used to shame or humiliate the other" (p. 333). This type of interpersonal exploitativeness, achieving self-enhancement only via a "superior, invidious relation to others" seems then, at least in the interpersonal system, to be the sine qua non of maladaptive, exploitative narcissism (see also Emmons, 1987).  One final issue has to do with the veridicality (as it were) of what a competitive narcissist is narcissistic about. While Leary (1957) notes that the particular form of narcissistic expression varies from person to person (i.e., some may focus upon physical strength or beauty, others may stress intellectual superiority, etc.), he makes no reference as to whether such qualities genuinely exist, or rather are merely arrogated to the self. Recent evidence,  however, suggests that "normal narcissists," as identified by the NPI, for example, do tend to be perceived as attractive, good-looking, handsome, etc. (Raskin & Hall, 1988). Similarly, Emmons (1984) cites unpublished data of Fischer (1984), who found that in a normal college sample, subjects rated as highly narcissistic (albeit "sub-clinically", in Fischer's terms) were indeed rated as possessing more positive characteristics than were less narcissistic students. A n d — a t the risk of tipping my own hand—research described later (Study 4) suggests that peers do indeed rate certain narcissists as verbally skilled and smart.  In any case, the range of individual differences in narcissistic expression or display (Leary, 1957) would seem to preclude identifying (or empirically establishing the veridicality of) any single attribute (e.g., beauty or intelligence) as characteristic of all competitive narcissists.  Object Relations  Theorists  Kohut's (1968, 1971) view of narcissism—like that of Leary—incorporates both adaptive and maladaptive expressions. Unlike Leary, however, Kohut's concern with narcissism is not as one dimension of interpersonal behavior, but rather as a developmental phenomenon, specifically, a progression from childhood to adult narcissistic libido. In his concise characterization of Kohut's position, M i l l o n (1981) explains that the early narcissistic libido "does not 'fade away' by being transformed into object libido [a la Freud], but unfolds into its own set of mature narcissistic processes and structures, [including] behaviors such as humor and creativity... and most significantly, it is through this narcissistic developmental sequence that the cohesive structure of 'self ultimately emerges" (p. 161). Narcissistic pathology, in contrast, stems from an early failure to successfully resolve difficulties surrounding either the "grandiose s e l f , i.e., the early infantile omnipotence that  17  must eventually countenance the illusion-shattering reality of actual inadequacies, or from failure to integrate the "parental imago," i.e., the early perception of parents as idealized, omniscient beings. Early experience of parental aloofness or capriciousness precludes the development of the self-cOnfidence that is concomitant to being loved and valued; the result, according to Kohut, is an insistently neurotic search for narcissistic recognition, one that persists throughout adulthood. Similarly, early parental rejection or indifference prevents the necessary idealization of parental figures; such failure is said to engender depression, emptiness and a consequent adult repetition compulsion, a quest for surrogate parental figures who possess the omnipotence (relative to self) that the narcissist never experienced. The concomitant is a weak, self-deprecating interpersonal style, one characterized by a willing submission to others thought to be potential providers of the ideal. Thus narcissism, as such, is typically a normal developmental progression; "injuries in the course of narcissistic maturation, however, can precipitate pathological states ranging from chronically low selfesteem to borderline or even psychotic states" (Millon, 1981).  In a sense, Kernberg's (1975) psychoanalytic view of narcissism is similarly characterized as normal development gone awry, but only insofar as it invokes the critical role of early experience with parental figures. Narcissism, according to Kernberg, is in the service of defending against "oral rage" that emanates from the child's inability to depend on "internalized good objects." Characteristic patterns of parental coldness, and indifferent or subtly aggressive attitudes toward the child are asserted to be common in a typical history of the adult narcissist, as is a form of double binding, where the child is concurrently granted an additional familial status such as budding genius, or viewed as the beloved cynosure, a much-  18  exalted only child. The latter role is postulated by Kernberg as one frequently sought by the adult narcissist, and is presented as the only effective anodyne for the pain inflicted by early rejection at the hands of basically unloving parents. The sequela is, of course, a "character pathology", comprising egocentricity, lack of empathy, an inordinate need for attention and admiration, grandiose self-inflation, remorseless interpersonal exploitation and a fundamental distrust/depreciation of others. Kernberg fails to elucidate if non-pathological narcissism—should it exist—would differ in kind, or only degree.  Millon M i l l o n (1969, 1981), on the other hand, is quite explicit that his "bio-social-learning" theory of personality (disorder) development is meant to apply to the normal as well as the pathological personality. He asserts that while "Pathology results from the same forces as involved in the development of normal functioning," it is critical differences in the "character, timing and intensity" (1981, p. 9) of those influences that render some individuals likely to manifest pathological traits, while others' development culminates in those that are adaptive and non-pathological. In any case, both developmental paths are said to be directed by the familiar principles of social learning theory (e.g., observation, imitation, reinforcement, etc.).  Not too surprisingly then, M i l l o n (1981) somewhat superciliously rejects notions of "libidinal regressions" and '"disappointments' with formerly omnipotent parents", arguing instead that the provenance of narcissism is parental overvaluation of the child's objective value, "that is, creating an enhanced self-image that cannot be sustained in the 'outer' world" (p. 165). When, according to Millon, the child cannot realize the ideals of "now internalized  19  parental illusions of self-worth", the consequence is the constellation of behaviors that typify the N P D . A s the D S M - I I I R conception of N P D is essentially coterminous with that of M i l l o n (as is that of D S M - I V ) , and because it and derivative operationalizations of narcissism are presented in a subsequent section, I now turn to a summary of the preceding positions.  Commonalities and Disagreements Arguably, then, each of the foregoing viewpoints shares the commonality of regarding the etiology of narcissism as some form of disturbance in interpersonal relationships in early childhood, be it specifically borne of parental aloofness, rejection, or capriciousness (Freud, Horney, Kernberg, Kohut, Leary) or parental overvaluation of the child and his capacities ([again] Freud and Horney, Millon, and Murray). Which type of parental behavior (i.e., over- or de-valuation) is the ultimate catalyst to narcissistic development is, of course, difficult to ascertain. Granted, each of the theorists' views are based on the undoubtedly rich sources of clinical lore and observation; similarly, it may also be the case that the frequency of both commonly observed/reported experiential themes in the histories of narcissistic patients may lend credibility to their status as etiologic candidates (for example, Leary reports that such patients are consciously disidentified with their parents and tend to recall them as "relatively sadistic"; 1957, p.339). However, it must also be acknowledged that data so derived are far from unequivocal, fraught as they may be with such potential confounds as memorial or self-presentational biases. Thus in the absence of data bearing upon actual in vivo parent-child interaction (ideally, longitudinally; see, for  20  example, Benjamin, 1987), one can assume at minimum, dual potential paths that culminate in the development of narcissistic dispositions. Another contentious point emerging from the foregoing analysis involves two separate, yet related issues: First, whether there is a continuum between normal and abnormal self-regard, and second, if what one might describe as universal, infantile narcissism or self-regard emanates from the same source as that which eventually debouches into the pathological. Freud—as was his wont—'refined' his opinions on these topics several times, alluding variously to primary and normal narcissism, its perverted form (i.e., in a psychotic-like withdrawal of libidinal and ego instincts from external objects to the self) and finally to the healthy narcissistic libidinal type. Nonetheless, he wrote more generally of universal selfpreservative instincts originating in childhood which, in varying degrees, pervaded each conception. A n d further, that "a high degree of this self-love constitute[s] the primary and normal state of things" (1922, p. 133).  Kohut is somewhat more explicit about the developmental progression from normal infantile to later healthy and pathological narcissism. He argues that while they originate from the same fountainhead, whether the course remains true—flowing into healthy, mature narcissistic expression or instead diverging into pathology in adulthood—is determined by the tides of certain critical early experiences. Horney similarly emphasizes the adult consequences of early experience, but only insofar as it gives rise to neurotic trends; narcissism is seen as only one refuge from the basic anxiety engendered by early negative parent-child interaction. A n d while Horney alludes to the child's developing sense of self, a normal stage of narcissism, per se, is not acknowledged. Indeed, as was noted earlier, she specifically abjures the notion of  21  a continuum between normal and narcissistic self-regard; in her view, the former represents a reasonably veridical assessment of one's worth whereas the latter is mere arrogation of value or achievements, a triumph of style over substance, as it were. Kernberg similarly rejects the concept of a normal narcissism, seeing it rather as a pathological "adjustment" to early traumatic experience. B y contrast, both Leary (1957) and Millon (1981; as well as Carson, 1969) are quite explicit about a continuum from normal to pathological narcissism. Each views narcissistic behavior as a sort of coping mechanism, a dispositional style learned first in response to problematic relationships experienced in childhood, deployed later as a 'tried and true', or at least effective, mode of interpersonal behavior. Both theorists (and DSM-III R and IV) view the difference between normal and abnormal narcissistic behavior as one of degree, not kind. Finally, there appears to be striking agreement among theorists as to the manifestations of pathological narcissism, i.e., an overweening interpersonal style characterized by selfabsorption and admiration, superiority and arrogance, egocentricity and cold exploitativeness. Yet by all excepting Freud and Leary (and to a lesser extent, Millon and Murray), the nature or even existence of non-pathological narcissism is given particularly short shrift. If there is a shared motif among the preceding conceptions of narcissism, it involves variations in self-evaluation and regard, be they veridical or merely verisimilar. O n the topic of what might be described as the care and feeding of the self, Harry Stack Sullivan once offered the following:  "Each of us comes to be possessed of a self which he esteems and cherishes, shelters from questioning and expands by commendation, all without much regard to his objectively observable performances which include contradictions and gross inconsistencies." (1948, p. 248).  22  Thus analogic to the One Genus Postulate, we may all be much more narcissistic than otherwise; the demarcation between normal and abnormal, adaptive and maladaptive, exists not on a substantive plane but rather one more stylistic, i.e., "the accent, the prominence, the misuse" (Sullivan, cited in Mullahy, 1952, p. 77) of characteristic forms of narcissistic expression. In conceptualizing a normal form of that which has historically been circumscribed to only the realm of the abnormal, the spectre of the pathology error looms large. One cannot presume that adaptive behavior is merely the absence of pathology; as Leary has forcefully asserted: " A science of malfunction cannot precede a science of function" (1957, p. 17). T o view normality merely as degraded (or "less distinguished', in Allport's terms) abnormality can precipitate, more often than not, "some curiously one-sided conceptualizations" (Leary, 1957, p. 12). In the case of contemporary conceptions of narcissism, this is evident in its characterization as a syndrome of overwhelming grandiosity, fragile self-esteem, entitlement and an unrelenting need for attention and admiration. While these elements might render an accurate portrait of clinical narcissism, their untempered transfer to the medium of non-pathological narcissism may produce only a misleading caricature. In the next section, current psychometric representations of normal, non-pathological narcissism are evaluated in this light.  The Measurement of Narcissism Insofar as better than two dozen measures of the construct have been developed in the last 50 years, the measurement of narcissism is not exactly terra incognita. While Murray (1938), for example, employed his characteristic multiform assessment to gauge "narcism" and its associated features, other approaches reflect more circumscribed conceptions, such as Blum's  23  (1950) projective method for assessing a narcissistic psycho-sexual stage. Several instruments are designed to measure DSM-III-derived conceptions of personality disorders, including the Narcissistic Personality Disorder (NPD), such as Millon's (1982) Clinical Multiaxial Inventory ( M C M I ) , and Morey, Waugh and Blashfield's (1985) M M P I Personality Disorder Scales; Ashby, Lee & Duke (1979) have also developed an MMPI-derived measure of personality disorders. Finally, based upon the N P D criteria described by Akhtar and Thomson (1982) as well as those of D S M III, Wink and Gough (1990) have constructed both M M P I and California Psychological Inventory (CPI; Gough, 1987) -based narcissism scales.  In the less-well-charted territory of non-pathological narcissism, however, relatively few reasonably comprehensive measures of the construct exist. The first such instrument developed was the Interpersonal Check List (LaForge & Suczek, 1955), designed to assess each of the eight hypothesized dimensions of the circumplex model of interpersonal behavior (see Leary, 1957, for an extended presentation of the model) as well as the range of intensityincluding the clinical or "maladaptive" end—of each dimension. The model upon which the I C L is based comprises 16 dimensions, each representing a distinct interpersonal style (e.g., B-Narcissistic, C-Competitive). These dimensions are in turn ordered in a circumplex array around the orthogonal coordinates of dominance-submission (status) and hostility-friendliness (affiliation). Adjacent sixteenths may be combined into the (more reliable) octant scores, where, for example, summing B and C scores yields a measure of a Competitive-Narcissistic interpersonal style. In the original form, competitive-narcissistic (BC) behavior is assessed (by, of course different items at different levels) at intensity level 1 (a mild or necessary amount of the trait) by the descriptors "self-respecting" (B) and "able to take care of self" (C);  at level 2 (a moderate or appropriate amount of the trait), by "independent", "self-confident" and "self-reliant and assertive" (B), and "can be indifferent to others", "businesslike" and "likes to compete with others" (C); at level 3 (a marked or inappropriate amount of the trait) by "boastful", "proud and self-satisfied" and "somewhat snobbish" (B), and "thinks only of himself [sic]," "shrewd and calculating" and "selfish" (C); and at the extremity of level 4 by "egotistical and conceited" (B), and "cold and unfeeling" (C). Actual 'diagnosis' with the I C L is relatively straightforward: subjects with B C scores one standard deviation or more below 6  the mean are said to be of the adjustive interpersonal type, i.e., the Competitive personality, while those one standard deviation or more above the mean, conversely, are described as maladjustive, i.e., a Narcissistic personality (Leary, 1957, p. 219).  In the rather complex Leary system, the I C L is used in the description of self or other (or by other) at Level II-C, that of conscious communication. Its psychometric characteristics are rather difficult to evaluate, primarily as few data from which to judge them are adduced. Among the actual data provided are two-week test-retest correlations, for example, values of .68, .74 and .76 for the B and C sixteenths and the B C octant score, respectively, (obtained from a sample of 77 obese women). With respect to normative data, only the sketchiest are provided («=212) for a six-month intake at the Kaiser Permanente Psychiatric Clinic.  7  These  and related considerations led Wiggins (1965, 1982) to remark that the assessment methods presented in the 1957 book constitute more a compelling case for the form interpersonal  "Complete or functional" interpersonal diagnosis, on the other hand, is far more complicated as it involves assessment with different instruments (e.g., ICL, TAT), on different levels (e.g., conscious vs. unconscious communication), and with different informants (e.g., self and other) yielding five different levels of diagnosis. In fairness, however, see also LaForge (1977) for some additional, albeit not definitive, psychometric data.  6  7  25  assessment could take "rather than a description of a wholly validated and psychometrically sound set of existing procedures" (Wiggins, 1982, p. 185). Indeed, while Wiggins (1982) has rather generously described the I C L as still, to some extent, the "standard psychometric device" for assessing the dimensions of the Leary system, it is also true that Wiggins' own Interpersonal Adjective Scales (most recently, IAS-Revised, Wiggins, Trapnell, & Phillips (1988 ) have all but supplanted the I C L as the measure of choice—on grounds of both 8  reliability and validity—for assessment in the domain of interpersonal behavior. Recently, Paddock and Nowicki (1986) demonstrated that on structural and psychometric grounds, Wiggins' IAS scales "are the current measure of choice for evaluating interpersonal styles in the Leary (1957) tradition" (p. 140). Despite the rather substantial methodological shortcomings of the I C L (and concomitantly, its measurement of narcissism), the conceptually rich contributions of the system it operationalizes cannot be underestimated, if only for its not insubstantial heuristic influence (see Wiggins, 1985). A n d with specific reference to narcissism, development of the I C L represented not only a delineation and formalization of how to appropriately conceptualize and measure narcissism as a maladaptive interpersonal reflex, but also permitted a rather shockingly normal narcissism to emerge from the conceptual closet, postulating for the first time an adaptive mode of competitive and independent interpersonal behavior.  The NPI and the PACL Turning then to measures designed specifically to gauge individual differences in normal narcissism, one finds, that with the exception of Wiggins, Trapnell and Phillips' (1988) I A S - R  IAS-R will be discussed more fully in a subsequent section.  26  B C (Arrogant-Calculating) scale (discussed in detail in Study 2, Method, below), there are, essentially, only two such instruments: Strack's (1987) Personality Adjective Checklist ( P A C L ) , and Raskin's Narcissistic Personality Inventory (NPI, Raskin & Hall, 1979; Raskin & Terry, 1988). Even the P A C L is not circumscribed to the measurement of narcissism, but includes it among the 11 scales developed as "normal" measures of the conceptions of personality disorders derived from Millon's social learning model. The NPI is similarly proffered as a measure of narcissism in "non-clinical" populations, and is described by its authors as a "viable and perhaps promising measure for the general construct of narcissism" (Raskin & Terry, 1988, p. 892). Just what comprises that "general construct" is, of course, the issue in this presentation, for while the NPI was specifically designed to assess normal, non-pathological narcissism, its constructors note that they were guided by an "original conceptual rational [sic] of developing an item pool that would fully reflect the DSM-III criteria for the narcissistic personality" (Raskin & Terry, 1988, p.892). One objection to such a conceptual rationale is based on substantive concerns: O n what grounds can it be presumed that a personality  disorder—by  definition a dysfunctional style of interpersonal behavior—provides an appropriate template for its presumably adaptive and functionally normal counterpart? In teaching anatomy, for 9  example, the dissection of a cancerous organ can be a vivid illustration of how pathology alters the organ's structure; the utility of this technique is diminished, however, if the student does not have an appropriate comparison standard, i.e., knowledge of the organ's structure in its healthy, non-pathological state. Thus in the absence of a clear comparative standard, i.e.,  Fischer's (1984) conception of "sub-clinical" narcissism is similarly redolent with implicit 'pathology potential.'  9  27  knowledge of the structure of non-pathological narcissism, the utility of "dissecting" the DSM-III criteria for the N P D remains to be demonstrated. While it is possible and indeed, probable that certain elements of the narcissistic personality (e.g., exhibition) are continuous with those of normal narcissistic expression, it is also probable that the latter could differ from the former in ways substantive as well as stylistic. Because the P A C L and the NPI are the only putative measures of normal narcissism (as an independent construct), examining their empirical relations to measures of N P D , or pathological narcissism, could intimate the nature of such substantive divergence should it in fact exist. Similarly, both instruments are derived from the DSM-III [R] definition of the N P D — t h e NPI directly, the P A C L somewhat more obliquely in that while derived from Millon's conceptions of the personality disorders, these same conceptions and the Axis II criteria do in fact show "close correspondence" (Strack, 1987, p. 572) in their definitional criteria. Thus the actual D S M - I V (see note, below) description of the N P D warrants detailed examination. A s indicated below, a diagnosis of N P D from D S M - I V requires that the following features  10  must be a) evident in the individual's contemporary and long-term  functioning; b) should not be circumscribed to only "episodes of illness"; and c) must cause the individual "significant impairment in social or occupational functioning or subjective distress":  Changes in diagnostic criteria from DSM-III to DSM-III R (and indeed, to the current DSM-IV) do not significantly alter their substance, save beyond making them slightly more 'behavioral' and less inference-based. The rewording of several inclusion criteria in DSM-IV reflects these changes. 10  28  A pervasive pattern of grandiosity (in fantasy or behavior), need for admiration, and lack of empathy beginning by early adulthood and present in a variety of contexts, as indicated by at least five or more of the following: (1) has a grandiose sense of self-importance (e.g., exaggerates achievements and talents, expects to be recognized as superior or "special" without commensurate achievements) (2) is preoccupied with fantasies of unlimited success, power, brilliance, beauty, or ideal love (3) believes that he or she is special and can only be understood by, or should associate with, other special or high status people (or institutions) (4) requires excessive admiration (5) has a sense of entitlement, i.e., unreasonable expectation of especially favorable treatment, or automatic compliance with his or her expectations (6) is interpersonally exploitative, i.e., takes advantage of others to achieve his or her own ends (7) lacks empathy: unwilling to recognize or identify with the feelings or needs of others (8) is often envious of others or believes that others are envious of him or her (9) shows arrogant, haughty behavior or attitudes  Diagnostic criteria for 301.81 Narcissistic Personality  Disorder  Thus implicit in both measures' operationalization of normal narcissism are dispositional tendencies and behaviors emanating from a grandiose sense of self-importance, exhibitionism, entitlement, interpersonal exploitativeness and a lack of empathy. How well then, do the NPI and P A C L capture these variables? (The issue of whether they should at all will be addressed shortly).  Perhaps not too surprisingly, few studies have directly assessed the relation between these measures and actual diagnoses of N P D based on DSM-III R. The reason for this is rather straightforward: The psychiatric/clinical literature on personality disorders and their diagnosis and the personality assessment literatures have not, until recently (e.g., Wiggins & Pincus, 1989), 'cross-pollinated.' In Strack's (1987) paper on the development of the P A C L , for  29  example, no information on the narcissism subscale's relation to clinical narcissism measures is reported. Similarly, Raskin & Terry (1988) failed to examine the NPI's correlations with more clinically-derived measures of N P D . A n d while this is not particularly egregious given that they were developed as measures of normal narcissism, the fact that each conception of narcissism is considered as essentially continuous with the non-normal, pathological pole of their respective systems would seem to beg the question of their degree of convergence. Addressing this issue, Mullins and Koppelman (1988) have correlated the NPI with three clinical (albeit not DSM-III-derived instruments) measures of narcissism: The MargolisThomas Measure of Narcissism (Margolis & Thomas, 1980); Ashby, Lee & Duke's (1979) Narcissism Personality Disorder Scale; and Serkonowek's (1975) MMPI-derived (Scale 5 — Maculinity-Femininity) Narcissism-Hypersensitivity scale. Interestingly, (and perhaps in support of the NPI's "normality") none of these measures correlated with NPI scale scores (rs = .06, -.09, .00, in the order as listed above).  Emmons (1987) also investigated the link between an earlier, 54-item" version of the NPI (substantively similar to the current 40-item measure) and selected clinical measures of narcissism: Solomon's (1982) MMPI-derived Narcissism Personality Disorder Scale (NPDS) and the narcissism scale of Millon's Clinical Multiaxial Inventory ( M C M I ; 1982), a measure which, as noted earlier, is closely related to DSM-III Axis II personality disorder criteria. With respect to the former, although NPI total scores did not correlate with those on the N P D S (r=.12, ns) one NPI component scale, Exploitativeness/Entitlement (actually two separate scales on the current NPI) was related (r=.32, p <.01) to N P D S scores. A s I shall touch upon  Details of NPI scale construction and revision will be presented in a subsequent section.  30  shortly, this result is especially intriguing given that those facets are interpreted by some (including Emmons, 1984, 1987) as among the most toxic (as it were), or pathological aspects of the spectrum of narcissistic behavior. NPI relations with the M C M I , on the other hand, indicated some degree of convergence, with NPI full-scale scores correlating a modest r = .27 (p < .05) with M C M I Narcissism scores. A m o n g the component scale scores, Self-Absorption/Self-Admiration—a substantive facet considered highly diagnostic to virtually all conceptions of narcissism (including that of DSM-III)—did not relate to M C M I narcissism scores, while Leadership/Authority and Exploitativeness/Entitlement showed correlations of .25 and .31 (both ps < .05), respectively, with that scale. Finally, the NPI Superiority/Arrogance component score correlated a not unsubstantial .48 (p < .01) with M C M I Narcissism. These results are not inconsistent with the present thesis, namely that while the components of normal narcissism are not entirely unrelated to clinical narcissism, neither are they isomorphic with clinically-derived, DSM-III-type notions of what the narcissistic syndrome comprises . One possible example is entitlement, or what DSM-III R describes as the "expectation of special favors without assuming reciprocal responsibilities" (DSM-III R, 1987, p. 335): Few adequately functioning persons could reasonably hold such expectations, if only as raw experience would likely fail to confirm them. That exploitativeness/entitlement should be the only NPI component to consistently correlate with clinical measures of narcissism supports not only Emmons' (1984, 1987) interpretation that this facet represents the "maladaptive and possibly pathological" (1984, p. 298; see also Gurtman et a l , 1991)  aspect  of NPI narcissism but also the thrust of the current presentation, again, that non-pathological  31  narcissism is largely uncontaminated by pathology, instead revealing itself in self-assured, agentic independence. This idea of non-clinical narcissism as manifesting itself largely in 'healthy', agentic interpersonal functioning is further supported by Strack's (1987) reports of other correlates of the P A C L narcissism scale. Consider, for example, the following P A C L correlates with the California Psychological Inventory (CPI; Gough, 1957, 1983)—an instrument, incidentally, among the most widely-used and best validated, it might be added): Social Presence ("poise, spontaneity and self-confidence in personal and social interaction") , rs -.41 12  and .49 for men  and women, respectively, and Dominance ("leadership ability, dominance, persistence and social initiative") r= .39 for men, not reported for women. Raskin & Terry (1988) report even stronger relations between the NPI and the same variables, i.e., Social Presence, r = .62 and Dominance, r = .71 as well as Adjective Check List ( A C L ; Gough & Heilbrun, 1965/1980) scale correlates such as Self-Confidence (r = .65), Dominance, (r = .63), Autonomy (r = .57), Deference ("to seek and maintain subordinate roles in relationships"; r = -.67), Abasement ("to express feelings of inferiority through self-criticism, guilt or social impotence"; r = -.63) and Exhibition (r = .62). Strack (1987) similarly found moderate to strong (i.e., high forties to low fifties) correlations between his measure and A C L Exhibition, Deference (-) and Abasement (-); these relations could, however, be somewhat inflated by common method (adjective check list) variance.  Still, the picture that emerges from these normal personality correlates of narcissism is consistent: Individuals scoring high on the NPI are, as Raskin possibly intended, those whose  12  Scale descriptions are from the CPI and ACL manuals.  32  self-presentation is one of strong, self-confident, and dominant social presence. So despite the clinically-influenced provenance of the P A C L and particularly, the NPI, these measures appear to be tapping (an) additional substantive area(s) that lies outside the realm circumscribed by narcissistic pathology and the universes of content defined by their authors. Some clue as to the nature of that substance may be provided by considering the correlates of the NPI subscales, in particular, their differential relation to external variables. Since the subscales themselves are based on the previously identified component structure of the NPI, a brief discussion of the apparent structure of the instrument is in order.  Structure of the NPI Development of the NPI Raskin and Hall first described the development of the 54-item NPI in 1979, and over the next few years, they and other researchers presented evidence said to demonstrate its potential validity in the measurement of narcissism as a normal personality trait. Again, however, it should be noted that item development and selection was based on the attempt to reflect the DSM-III criteria for the N P D (noted earlier). In terms of external correlates, Raskin and Hall (1981), for example, reported positive relations between the rationally-developed NPI and Eysenck's Extra version and Psychoticism scales, while Raskin (1980) found a small positive correlation between the NPI and scores on a measure of creativity. The same author reported a positive association between narcissism scores and the use of first person singular pronouns; first person plural pronouns were found to be negatively related to NPI scores (Raskin & Shaw, 1988). Emmons (1981) found that NPI scores were positively related to a subset of sensation-seeking subscales, while L a V o p a (1981) reported that narcissism was correlated  33  with Machiavellianism—in females only—and wncorrelated with Marlowe-Crowne social desirability. Thus the scale development strategy adopted by the authors of the NPI seems best captured by what Gough (1965) has termed "primary evaluation", or the empirical validation process seen as requisite to even considering issues pertaining to an instrument's dimensional structure. It was left for Emmons (1984) to undertake the relatively more controversial procedures involved in "secondary evaluation", or the attempt to explore "and hence illuminate the underlying dimensionality that is inherent in any test or measure possessing primary utility" (Gough, 1965, p.295).  Emmons (1984, 1987) Emmons (1984) performed a principal components analysis on a 54X54 matrix of phi coefficients, extracting four components which he then rotated to an oblique solution (oblique because he reasoned that it is the "combination [of hypothetical factors] which defines the trait of narcissism", p.292); the solution accounted for 72% of the variance. Emmons' component labels (and sample items and their loadings) are as follows: I. Exploitativeness/ Entitlement (e.g., "I find it easy to manipulate people"; .50); II. Leadership/Authority (e.g., "I would prefer to be a leader", .68); III. Superiority/Arrogance (e.g. "I can read people like a book; .48); and IV. Self-Absorption/Self-Admiration (e.g., "I like to look at myself in the mirror", .53). Examining the matrix of loadings offered in Emmons (1984), one sees a few apparent anomalies. First, fully 22% (12) of the 54 items fail to load on any component at a level exceeding .35 (Emmons' own criterion for substantive interpretation), indicating some ambiguity concerning their relation (if any) to the construct tapped by the NPI (actual item-  34  totals are not reported). Second, a number of items do not load univocally; while this is perhaps to be expected with oblique factors, in some cases, as little as .02 separates the double loadings such that an item's assignment to one component scale or another seems almost arbitrary (e.g. "I can usually talk my way out of anything" loads .33 on Exploitativeness/Entitlement and loads .35 on Superiority/Arrogance; the item is assigned to the subscale defined by the latter factor). In examining factor intercorrelations, Emmons reports factor-based scale score intercorrelations, these ranging from .13 (Factors I & III), to .40 (Factors II & III), with a mean of .32. He also noted a full-scale coefficient alpha of .86, with subscale coefficients of .74, .79, .69 and .69 for Factors I through IV, respectively.  In 1987 Emmons attempted replication of these results, albeit with a different factoring technique (i.e., principal axes factoring). Using the same extraction criteria, four factors were retained and rotated obliquely, this time accounting for 70% of the variance. Although the order of emergence varied, the loading pattern is, apparently "virtually identical" (Emmons, 1987, p. 13) to that obtained in Emmons (1984); some anomalies, however, again emerge along with the components. As with the first analysis, some 31% of the NPI items fail to load at .35 or better. Additionally, seven items that load at .35 (or greater) in the first analysis fail to replicate in second, i.e., they do not load at all in the solution, while six items load significantly in the second analysis that do not emerge in the first. These details aside, however, it does appear that the two factor solutions are indeed comparable in general.  35  Raskin and Terry (1988)  13  --and some methodological  issues  With reference to Emmons' (1984, 1987) factor analytic work, Raskin and Terry (1988) noted several cases in which items loading on the same factors appeared to be tapping "somewhat different conceptual dimensions" (p. 892). From this observation, they drew two conclusions: First, that Emmons used "a conservative selection criterion" in deciding to retain only four factors (thus under-extracting and possibly, obscuring the true dimensionality), and second, that his use of inter-item phi coefficients (as opposed to tetrachorics) may have yielded a somewhat ambiguous (or at least less clear) picture of the underlying item structure (p. 893); each of these assertions deserves comment. First, Emmons' "conservative selection criterion" was the joint use of the Kaiser-Guttman eigenvalues-greater-than unity rule and examination of the solution's scree plot (Cattell, 1966). While factor extraction/retention criteria are still relatively controversial (e.g., Hakstian, Rogers & Cattell, 1982), the criteria that Emmons used are easily among the most common; the Kaiser-Guttman rule is the default on many statistical packages (e.g., SPSS). More to the point, however, is the fact that use of the Kaiser-Guttman rule—far from being a conservative retention criterion—may precipitate over-extraction of factors (e.g., Briggs & Cheek, 1986). Thus Raskin and Terry's (1988) allegation of conservatism in Emmon's (1984, 1987) decision criteria seems questionable.  Further, Raskin and Terry's (1988) assertion concerning the superiority of tetrachoric (r{) over phi coefficients—because the former "can often provide a clearer view of the latent item structure that underlies the inter-individual differences in the response characteristics of a test"  A detailed summary of the following and subsequent sections on the NPI and its nomological network is provided on p. 29 ff. 13  36  (p. 893) is not a universally accepted dictum. Controversy as to which is the preferred method for the analysis of dichotomous test data centers around their relative robustness with respect to departures from bi- or multivariate normality (Lord & Novick, 1968), and how such departures influence the degree of over or under-estimation of the product moment (PM) correlation. Glass and Stanley (1970), for example, assert that while phi will underestimate (relative to r{) the P M correlation one would obtain if one in fact had data from a normal distribution of a continuous latent variable that underlies the observed dichotomous one, they also note the possibility of "substantial errors" (p. 167) in overestimation by the use of rj- (in a positive relation) if proportions depart greatly from .50 for each response; these latter, however are said to be superior if a normal distribution can in fact be assumed. Carroll (1961) similarly holds that neither phi nor phi/phi max have the desirable invariance properties attributed to the tetrachoric.  Nunnally (1978), however, argues that "There are very strong reasons for not using...r^, among them that the accuracy of  is highly dependent upon strict assumptions of normality,  such that departures from same can yield (usually over) estimates off by as much as 20 points of correlation (p. 136-7); (McNemar, 1968, similarly notes that under even ideal conditions, i.e., .50/.50 proportionality with a relatively low correlation, the standard error of  can be up  to 5 0 % higher than a comparable P M correlation, p. 224) and perhaps more germane to the present discussion, that r is not highly mathematically tractable. Nunnally (1978) argues: t  "strictly speaking, it is illegitimate to use such estimates in...multivariate analysis [because in] employing... r in multivariate analysis, the experimenter can claim t  that these are estimates of the P M formula and thus can be used like P M coefficients. In a strict sense this is not mathematically proper and should be tolerated only in the study of mathematical models related to psychometric theory and not in analyzing the results of actual research on people" (p. 137).  37  He, seemingly rather disbelievingly, also notes how "one still sees instances of r employed in t  factor analysis" (p. 137). The relative superiority of  over phi thus seems, at least according  to Nunnally (1978), to be less than clear-cut. Such issues are, of course obviated if one has data on variables actually measured continuously rather than estimated from dichotomously scored variables merely assumed to be continuous.  14  Returning then to Raskin and Terry's (1988) investigation of dimensionality of the NPI, those authors report the results of their own principal components analysis of the NPI. Having reduced the original 54-item pool to 47 (owing to nonmonoticity of 7 items relative to fullscale NPI scores) they subjected the matrix of inter-item tetrachoric coefficients to a principal components analysis. B y a variety of criteria (e.g., a minimum of 3 univocal loadings for each component), seven components were extracted and rotated to a (weighted promax) oblique solution. Owing to unreflectable significant negative loadings of 3 items in the pattern matrix, and failure of 4 additional items to load in the solution at all, these items were dropped, the resultant matrix reanalyzed, and seven components extracted and rotated as above; this solution accounted for 52% of total NPI variance.  Raskin and Terry (1988) interpreted the components as follows (the keyed alternative of sample items and their loadings are in parentheses): I. Authority ("I would prefer to be a leader"; .83); II. Self-Sufficiency ("I rarely depend on anyone to get things done"; .61;); III. Superiority ("I am an extraordinary person"; .69); IV. Exhibition ("Modesty doesn't become me"; .69; V . Exploitativeness ("I can read people like a book"; .71); VI. Vanity ("I like to look at my body"; .87); and VII. Entitlement ("I will never be satisfied until I get all that I deserve";  14  I proffer such continuous measurement in a subsequent section.  38  .70). With respect to component and full-scale reliability estimates, rather than K R 2 0 , Raskin and Terry (1988) report Guttman lambda 3 values (a measure Lord & Novick, 1968, report as "typically...modestly better...and occasionally...substantially better" than coefficient alpha; p. 94). While the full-scale coefficient is a respectable .83, the corresponding component-based scale values are considerably lower, ranging from a low of .50 (Entitlement and SelfSufficiency) to a high of .73 (Authority); the median value is .54. Raskin and Terry (1988) argue that these low reliability estimates are due to the relatively small number of items that the scales comprise; belying their explanation is the estimate for the Vanity scale, i.e., .64, a value second only to that of Authority, and comprising the smallest number of items—three.  The scales themselves are somewhat intercorrelated, ranging from . 11  (Self-Sufficiency  and Vanity) to .42 (Authority and Exhibitionism); the mean intercomponent correlation is .27, a reasonable value given the conceptual relatedness yet relative independence one might expect to characterize a "syndrome of diverse yet interdependent behaviors" (Raskin & Terry, 1988, p. 893).  With respect to the last result, it is important to note that Raskin and Terry (1988) did indeed obtain some evidence indicating an underlying general factor of what they interpreted as narcissism: A l l but four of the 40 NPI items loaded positively and above .33 on a first unrotated principal component (FUPC) which accounted for 22.7% of total NPI variance. Similarly, a P C A of the  component intercorrelations yielded an F U P C accounting for 35% of  the total intercomponent variance; component loadings ranged from .33 (Vanity) to .76 (Authority). Thus in evaluating the apparent dimensionality of the NPI, a few observations might be noted. First, it seems that most of the substantive action (as it were) lies in what  39  Raskin & Terry (1988) label Authority; marked by items such as "I would prefer to be a leader", "I see myself as a good leader", "I will be a success", "People always seem to recognize my authority", "I have a natural talent for influencing people", "I am assertive" and "I like to have authority over other people" (not to mention "I am a born leader"), the scale is redolent with content reflecting generally, surgency and dominance, and more specifically, sub-components of leadership, authority, ambition, etc. None of which, it might be noted, have direct conceptual counterparts in the DSM-III N P D criteria upon which the NPI item pool is said be based (Raskin & Terry, 1988, p. 892). A n d of the 17 NPI items loading most highly (i.e.> .49) on its F U P C , 8 of these are from the Authority scale.  In sum, these structural studies of the NPI are reasonably consistent. I observed that while Emmons (1984) extracted four components accounting for 72% of NPI variance, over one fifth (22%) of NPI items failed to meet his retention criterion (and, incidentally, indicated fairly strong multivocality). His attempted replication (1987) yielded similar results, with, this time, almost one third (31%) of items failing to meet his retention criterion. Although other minor problems existed, his two solutions were, in general, comparable.  Again, Raskin and Terry (1988) claimed that through Emmon's dual use of certain factor retention criteria (i.e., the Kaiser-Guttman rule and examination of the scree elbow; surely among the most commonly employed) and calculation of phi rather than tetrachoric correlation coefficients, he probably ambiguated the structure of the NPI. Their solution was to drop 12 items and extract seven (as opposed to Emmons' four) components from which they built scales with a median Lambda 3 value of .52. Most items did, however, load on a first (unrotated) principal component, and analysis of scale intercorrelations yielded a component  40  accounting for some 35% of observed NPI variance. There are, however (as we shall see) considerable problems—methodological and substantive—with the NPI.  Nomological Network of the NPI External  Correlates  The pattern of external (self-reported) correlates of the full-scale NPI further substantiates the interpretation of the NPI as tapping primarily dominance and surgency. The strongest CPI correlates, for example, are Dominance (r=.71), Sociability (r=.66) and Social Presence (r=.62); the Authority subscale correlates .79 with CPI Dominance, and .57 with Sociability; while Raskin and Terry do not provide the corresponding value for Social Presence, one suspects that it would be similarly substantial. Relations between the full scale NPI and selfreported A C L scales are consistent with the above: Among the highest correlations are those with nDominance (r=.63) and Self-Confidence (r=.65); the corresponding values for the Authority scale are even stronger, (i.e., r = .68 for both nDominance and Self-Confidence.  That the NPI assesses an overt behavioral style characterized by dominant and surgent behavior is convergently validated by Raskin and Terry's (1988) Institute of Personality and Assessment Research (IPAR) live-in assessment data. After observing 57 subjects in a variety of formal (e.g., Leaderless Group Discussions; L G D ) and informal (e.g., breakfast) contexts, 10 I P A R observers rated subjects on trait rankings (e.g., evaluativeness), and completed a California Q-Set ( C A Q ) and A C L observer reports for each. While few of the correlations are remarkably strong, the consistency and coherence of their pattern suggests the commonality of agency/dominance-related characteristics. Among the trait rankings, for example, high NPI scorers, as well as being rated as narcissistic (r=.47), were also seen as sensation-seeking  41  (r=A2; a reliable correlate of surgency), evaluative (r=.37), dominant (r=.36) and extraverted (r=.35). C A Q items positively correlated with NPI scores were similarly consistent with a dominant interpersonal style: Characteristically Pushes and Tries to Stretch Limits (r=.50); Behaves in an Assertive Fashion (^.46); Perceives Many Contexts in Sexual Terms (r=.42); Expresses Hostile Feelings Directly (r=.41) and Feels Satisfied with Self (r=.38). Negative correlates reinforce this picture: Arouses Nurturant Feelings in Others (r=-.43); Is Concerned With Own Adequacy as a Person (r=-.42); Is Genuinely Submissive (r=-.41); Seeks Reassurance From Others (r=-.40) and Tends to A v o i d or Delay Action (r=-.40). Finally, A C L ratings of high scoring NPI subjects are substantively consistent with those of the C A Q : SelfConfident (r=Al);  Rude (r=.40); Aggressive (r=.38); Autocratic (r=.38); Self-Centered  (r=.37); Assertive (r=.37) and Ambitious (^.36); negative correlates (e.g., Submissive, r=.57) are in the main, antonymous with the above.  Briggs and Cheek (1986) have argued that the components of a multiscale inventory are valid and/or important only to the extent that they are differentially related to other measures (p. 111, see also Carver, 1989; West and Finch, in press). How then, might one assess the "importance" of the NPI components on this criterion? While Raskin and Terry (1988) assert, quite reasonably, "some evidence" (p. 897) for differential correlation among the components (see pp. 896-7), they also note, particularly with respect to the A C L observer ratings, "a substantial degree of overlap" in the component correlates. For example, both Authority and Self-Sufficiency correlate at levels •> .25  15  with the same 12 of the highest 25 A C L observer  ratings; Authority and Entitlement have the same pattern for 10 of the A C L ratings; Authority  Although the authors omit this information, a value of >-.25 is required for a p<.05 level of significance with sample size of 57. 15  42  and Exhibition for 8 of the A C L ratings, Exhibition and Self-Sufficiency for 9, etc. Although these values are not overwhelming, the striking and consistent overlap in the pattern of component correlates may render their practical independence questionable . A n d while the 16  Vanity component does indeed show a wholly differentiated correlational pattern (being the only scale to correlate with the A C L items Attractive, Good Looking and Handsome), Superiority shows no pattern whatsoever, as it is uncorrelated with any of the 300 possible A C L ratings. Buss and Chiodo (1991) recently obtained a similar result, finding no significant correlations between the NPI Superiority component and their act-based measures of narcissism. Neither the Vanity nor Superiority component scale scores correlate significantly with the trait ranking for Narcissism.  In fairness, however, most of component scales do have some conceptually-consistent rating correlates. For example, Exhibitionism correlates with the L G D rating for Exhibitionism (r=.36), the Trait Rankings for Extraversion (r=.36), Sensation-Seeking (^.40) and the A C L scale nExhibition (r=.62). Exploitativeness is positively related to the C A Q items Tends to be Rebellious and Non-Conforming (r=.43), Characteristically Pushes and Tries to Stretch Limits (r=.39) as well as the A C L items Aggressive (r=.29) and Unscrupulous (r=.27). Correlates of Entitlement include the C A Q ratings for Expresses Hostility Directly (r=.36) and Is Power Oriented (r=.32); A C L items correlating with the scale include Ambitious (r=.46) and Tough (r=.38). Self-Sufficiency also has some meaningful correlates, for example, the C A Q items Behaves in an Assertive Fashion (r=.40) and Feels Satisfied with  An alternative interpretation of this pattern is, of course, that the overall similarity in the separate scales' correlates is due to the impact of an underlying shared variance. As the reader shall see, a similarity of correlational pattern is also shared by some POND scales; in the case of the POND, however, this is strictly by design. I don't think that the same is true of the NPI 16  43  Self (r=.39); A C L items correlating with Self-Sufficiency include Autocratic (r=.42) and Independent (r=.33). The only rating correlates of Vanity were the C A Q items Regards Self as Physically Attractive (r=.39) and Is Physically Attractive (r=.37), and the A C L observer ratings Attractive (r=.41), Good Looking (r=.37) and Handsome (r=.35). A n d while Raskin and Terry (1988) note that Superiority correlated with a whopping 1% of the observational criteria, none are listed.  Thus with respect to the Briggs & Cheek (1986) criterion of meaningfully differential patterns of component correlates, the NPI fares only reasonably well: While several components (e.g., Authority, Entitlement and Exhibition, Self-Sufficiency and Exploitativeness) appear reasonably empirically valid by virtue of their relations to conceptually-related self-report and observational criteria, there is also evidence of substantial overlap in several of those relations. At the same time, Superiority and Vanity appear to have few meaningful correlates at all. While Raskin and Terry (1988) argue that the overlap is reasonable given the relative independence of dimensions that are nonetheless subsumed by a higher order construct, one could just as readily assert redundancy of measurement owing to the fractioning of dominant/surgent content into superfluously independent scales.  Similarly, those authors argue that the relative saliency of the various component-related behaviors could have precluded certain relations from emerging. For example, they argue— not unreasonably—that behavior relating to Authority and Exhibition would be much more observable in the I P A R setting than would that arising from vanity. How this argument could be logically extended to Superiority—a highly observable interpersonal style—is somewhat unclear. A n d it is also strikingly at odds with the observation by Emmons (1984), who, in  commenting on his own results pertaining to the relations between ratings and NPI scores (albeit the 54-item version), asserted that "the NPI reflects behaviors which manifest themselves publicly and are highly interpersonally observable" (p. 298). Consider also the variables that did correlate with Entitlement, i.e., A C L ratings of Ambitious, Independent, Tough, Enterprising; CPI self-report items Self-control (-) and Tolerance (-) as well as C A Q items Expresses Hostility Directly and Is Power Oriented. Far from being merely "peripherally relevant" (Raskin & Terry, 1988, p. 900), these correlates are instead highly conceptually meaningful insofar as indices of  normal personality would be  expected to correlate with a clinical concept, which, for all intents and purposes, entitlement is. Unlike many of the C A Q items which were, indeed, derived in a clinical context, both the A C L and the CPI (despite its partially clinical, i.e., MMPI-derived, provenance) were specifically developed with reference to dimensions of normal personality, or "folk conceptions" of same. I noted earlier how both Emmons (1984, 1987) and Raskin and Terry (1988) have suggested that exploitativeness and entitlement may represent the more toxic or maladaptive aspects of narcissism; insofar as one wishes to develop an index of normal narcissism, and insofar as the NPI accomplishes that goal (thus far an open question) then the fact that a clinically-derived concept should have a less-than-rich nomological network among constructs of normal personality is not particularly surprising. Interestingly, many of the variables with which Entitlement does correlate are also related to Authority and involve dominant-type content —certainly within the realm of normal personality. One can only conclude that either the NPI does not measure Entitlement reliably and validly, or it simply  45  does not belong in the spectrum of normal narcissistic behavior; this of course, is an empirical question (and one that will be addressed presently). Raskin and Terry (1988) report one additional validity study in which they attempted a structural location of the NPI within the Leary (1957) circumplex model of interpersonal behavior. Somewhat curiously, however, the measure they chose to index the latter was the Interpersonal Check List (ICL; LaForge & Suczek, 1955). Because Wiggins has developed a validated and highly psychometrically sound set of interpersonal assessment devices (i.e., the current Interpersonal Adjective Scales—Revised; IAS-R, Wiggins, et al,  1988), Raskin and  Terry's (1988) selection of the less-sophisticated I C L seems rather curious indeed, if only as far more psychometrically-advanced interpersonal assessment techniques—such as the IAS-R— exist (see also Paddock and Nowicki, 1986). However, it is indeed true that the I C L was meant to gauge a range of intensity of interpersonal behaviors, in that points (scores) falling relatively nearer to the center of the circle are though to fall within a normal range (e.g., competitive), and those lying closer to the circle's perimeter are understood as more pathological (e.g., narcissistic).  In contrast to the I C L , the IAS were developed, normed and validated entirely on nonclinical (i.e., normal) populations: Octants and their associated items reflect this difference in orientation. For example, I C L Managerial-Autocratic is equivalent to IAS Assured-Dominant, Aggressive-Sadistic corresponds to Cold-hearted, while Competitive-Narcissistic is equivalent to IAS Arrogant-Calculating. While some of these traits are perhaps less than highly socially desirable (e.g., arrogant, cold-hearted, calculating, etc.) they nonetheless reflect traits that are both normally distributed in non-clinical populations and are relatively free of pathological  46  content; I C L terms such as sadistic, autocratic, even narcissistic, can probably not lay claim to such a distinction. These considerations aside, the method employed by Raskin and Terry (1988) was to have subjects (an n-\21  subset of the NPI factorial study sample) complete the 128-item I C L under  separate real-and-ideal-self instructional sets (the retest interval is not noted) and then calculated octant and axis (dominance and nurturance) scores. The authors do not provide information as to the relative intensity of subjects' I C L scores (e.g., where scores u 1 standard deviation from the mean are considered within the normal range of interpersonal intensity), which is a considerable omission, given that the intensity of I C L scores  defines that range of  normality. They do report that high NPI scorers presented themselves (under "real self" conditions) as Competitive and Narcissistic (r=.47), Managerial and Autocratic (r=.45); and Aggressive and Sadistic (r=.30); the correlation between NPI scores and those on the Dominance axis was r=.47. With respect to NPI component scores, all were related—albeit modestly, i.e., all r's < .33—to the Competitive-Narcissistic octant score, ranging from .17 (Entitlement) to .33 (Superiority).This latter correlation is interesting in light of Raskin and Terry's (1988) finding that the Superiority scale showed the  weakest pattern of relation to the I P A R observer ratings  (e.g., vis-a-vis the A C L ) .  In contrast, the "real" (vs. "ideal") self-report condition in the I C L study may have exerted a press toward more "honest," less self-presentationally-oriented self-report; subjects responding under these conditions may have thus been more likely to admit to being shrewd, calculating, proud and self-satisfied, etc.  47  Finally, other NPI component scores that showed significant relations with I C L variables included: Authority with Managerial and Autocratic (r=.49); Entitlement with Rebellious and Distrustful (r=.27); Exploitativeness with Sadistic and Aggressive (r=.25); and Cooperative and Over-Conventional (r=-.29); and Self-Sufficiency correlated negatively with Docile and Dependent (r=-.39). The explicitly interpersonal presentational style of the high NPI scorer is thus very much consistent with that emerging from other self-reported (and other-rated) personality variables: While he describes himself as Competitive and Narcissistic, he also presents as highly dominant (i.e., I C L Managerial and Autocratic), the actual strength of these respective relations diverging by a mere two points of correlation. Interestingly, the strongest interpersonal correlate of the NPI and its components is that of I C L Managerial and Autocratic with NPI Authority, a scale whose content, as was noted earlier, is almost purely dominance.  Summary of NPI research Having considered a wide range of the correlates of the NPI and its components, a summary of the characteristics of the nomological net of the construct apparently tapped by that instrument is in order. Recall that Raskin and Hall (1981) explicitly stated that the conceptual template for their measure of normal narcissism was that of abnormal narcissism as delineated in the DSM-III criteria for the N P D . The appropriateness of this model aside, it is nonetheless possible to gauge the degree to which the two formulations appear to converge-or diverge, as the case may be.  A s noted earlier, the few studies that have directly assessed relations between the NPI and clinically-based measures of narcissism, have demonstrated either no relation to NPI full-scale  48  scores (e.g., Mullins & Koppelman, 1988) or at best, modest levels of correlation (e.g., Emmons, 1987, finding of r=.21  between NPI full-scale scores and those on the M C M I  Narcissism scale. Correlations on the componential level, however, reveal a somewhat different picture, with ostensibly exploitativeness/entitlement-type  content consistently  correlating with clinical measures (including the M C M I ) . So while some portion of NPI entitlement/exploitativeness-related content may be related to those variables in their clinical expression, their NPI correlates indicate that the associated behavioral manifestations in nonclinical samples (as rated by trained observers) may be agentic rather than maladaptive.  In an earlier section on theoretical conceptions of narcissism, I noted an apparent substantive commonality among them, namely variations on the themes of self-evaluation and regard. The actual DSM-III R definition of the N P D echoes this theme in its criteria relating to grandiosity with respect to self-importance and uniqueness, or in other words, abnormally high and possibly non-veridical self-regard. This variable—be it labeled self-regard, self evaluation or self-esteem—is one of the most reliable and powerful correlates of a fundamental interpersonal variable, i.e., agency or dominance, itself a trait evincing one of the richest nomological nets in the realm of ordinary language conceptions of personality (see for example, Wiggins & Broughton, 1985; Wiggins & Broughton, 1991). That the densest area of the nomological network of the construct tapped by the NPI should be woven of relations among dominance/agency and its associated constructs (e.g., authority, leadership, power, ambition, self-confidence, etc.) is thus not especially surprising—and perfectly theoretically consistent. Consistent insofar as in order to be successfully agentic, one must have a firm substratum of confidence in one's competence and effectance, and this, according to high-  49  scoring NPI subjects, is precisely what they have: "I see myself as a good leader" "I will be a success" "I am assertive", etc. The high NPI scorer is convergently rated by observers as being rude and tactless, but also as ambitious, aggressive, independent, assertive, self-confident and strong. In short, the NPI narcissist presents himself as self-confident, ambitious and assertive—and observers corroborate those impressions. The NPI thus appears to be assessing a form of narcissism fundamentally different than that evoked by the usual clinical conceptions, insofar as these qualities are not merely arrogated to the self, but are reliably recognized and acknowledged as veridical by observers.  Raskin and Terry (1988) provide the following summary of their research:  "The observational and self-report correlates of the full-scale NPI found in the IPAR  sample portray the high NPI scorer as being relatively dominant,  extraverted, exhibitionistic, aggressive, impulsive, self-centered,  subjectively  self-satisfied, self-indulgent and non-conforming. This portrait of the narcissist is highly congruent with the one we would expect to find in non-clinical manifestations of narcissism and is entirely in accord with clinical observation" (p. 899). pin part, clinical observation was what gave rise to the DSM-III behavioral criteria for N P D and these, as noted earlier, underlie the conceptual rationale of the NPI; as Emmons (1984) has noted "the assumption is that when exhibited in less extreme forms, these behaviors are reflective of narcissism as a personality trait" (p. 292). The issue then, is whether what is measured by the NPI is, in fact, a "less extreme form" of the N P D criteria. One approach to addressing this question is to ascertain whether these criteria actually have directly valid counterparts in the NPI, specifically, between the latters' components and the comparable criteria.  Examining each in turn, it appears that N P D Exhibitionism is indeed conceptually continuous with the correlates of its NPI counterpart: Observer ratings of self-indulgence, selfdramatizing, exhibitionistic, impatient, outgoing, extraverted and show-off, as well as selfreported descriptors indicating exhibitionism, social presence, lack of self-control (or impulsivity) and sociability are logical equivalents of clinical exhibitionism in its "less extreme form". Thus the substance underlying this component does indeed seem continuous with more clinical conceptions of such behaviors.  A s to the N P D criterion entitlement, again, defined as the expectation of special favors without reciprocation, the comparable NPI component does seem to tap a sense of seeing the self as deserving and indeed requiring recognition (e.g., "I want to amount to something in the eyes of the world"; "I have a strong will to power"). It also seems, however, to be associated with behavior that is oriented toward self-determined achievement of that recognition: The highest observer ratings associated with Entitlement are ambitious, independent, determined, enterprising and tough, as well as power-oriented, expressive of hostility, and intolerance of others. A n d while this sort of "What makes Sammy run?" syndrome is perhaps less than optimal interpersonally, it is also often associated with actual achievement, something that generally keeps people outside of, rather than in the clinic.  If narcissism is indeed a continuous trait, then one would expect that the N P D criterion of exploitativeness—as we saw earlier, central to many theoretical conceptions of the construct—should be evident in NPI narcissism. The I P A R rating correlates, as with Entitlement, are not particularly warm and fuzzy, e.g., rebellious, non-conforming, rude, tactless, hostile, aggressive, and trying to push and stretch limits. Yet associated items (e.g., "I  51  can read people like a book", "I find it easy to manipulate people" or "I can usually talk my way out of anything") indicate characteristics that far from being maladaptive liabilities (albeit perhaps interpersonally offensive) are instead considered by many as effective assets in our achievement—driven society. I suspect that Exploitativeness may indeed fall on a continuum from normal to abnormal, the difference between the two being that the effective narcissist is one who appreciates the utility of selective and subtle deployment of this interpersonal maneuver rather than using it indiscriminately, or rigidly and inflexibly (see also Leary, 1957).  Finally, there are two additional N P D criteria that seem inextricably intertwined, namely grandiosity with respect to self-importance and uniqueness, and a preoccupation with fantasies of power, unbounded success, beauty, etc. The NPI does indeed have content that appears to tap such substance, most notably in the Superiority component. Items such as "I am going to be a great person", "I think that I am a special person", "I know that I am good because everybody keeps telling me so", and "I am an extraordinary person" seem to bespeak a selfconcept more than a little inflated, and one that is clearly substantively consistent with the above DSM-III criteria, as well as the usual clinical descriptions of the narcissist. In fact (as noted earlier) NPI Superiority is the component correlating most highly with the I C L Narcissistic/Competitive scale (although even that value is a modest .33). Still, one is hard pressed to characterize the nature of this normal grandiosity (if you will) insofar as with the exception of NPI Vanity, it is the only other component evincing a nomological net that is at best, somewhat threadbare. NPI Superiority has no significant correlates among the A C L observer ratings nor among the C A Q ratings (or at least, Raskin & Terry do not note them) and the only related self-report items were CPI Capacity for Status, and the A C L scales Ideal  52  Self, Free Child and Self-Confidence. With respesct to the A C L Ideal Self scale, Gough and Heilbrun (1980) report that while high scorers on the A C L Ideal Self scale tend to exhibit some "elements of narcissistic ego inflation" they are concomitantly characterized by interpersonal effectiveness and goal-attaining ability. So once again, it appears that while the DSM-III R superiority/grandiosity criterion and the corresponding NPI Superiority component are at least nominally continuous, the few empirical correlates of the latter intimate (however tentatively) that this putative aspect of normal narcissism might instead be more achievementoriented than otherwise.  In sum, the assumption that the behaviors said to be gauged by the NPI are merely "less extreme forms" of those associated with the N P D criteria appears not wholly justified. While the substance of Exhibitionism and its external correlates provide an almost mirror-like image of the corresponding N P D criterion (i.e., #7, DSM-III R, p. 351), the other components provide considerably less clear reflections of their ostensibly-related criteria. Even apparent topographical similarity—as with entitlement and the desire/need for recognition apparently tapped by the N P I — i s belied by empirical evidence for the wherewithal (as indicated by observer ratings) to legitimately achieve same, something that the pathological narcissist either can not, or does not, demonstrate.  A s for the apparent conceptual relation between grandiosity and NPI Superiority, the facts that a) the component as a whole shows, at best, a meagre pattern of empirical correlates, and b) that those self-reported variables to which superiority is related are those tapping mainly adaptive dimensions, not only render interpretation of the component unclear but concomitantly preclude evaluation of its conceptual continuity with any of the N P D criteria.  53  Once again, either the NPI fails to adequately measure "normal" grandiosity/superiority, or else the "conceptual template" of the N P D criterion is simply inappropriate and/or irrelevant in the assessment of normal, adaptive dominance and self-confidence which, from its few significant correlates, seems to be what NPI Superiority taps. NPI Vanity, in contrast, seems almost certainly continuous with the comparable N P D criterion, i f only because high NPI scorers are perceived by raters as regarding themselves as physically attractive. But because judges also say that those same subjects are, in fact, physically attractive, good looking and handsome, this seems rather a veridical convergence having little to do with criteria of pathology—be the manifestations "less extreme" or otherwise.  One is harder pressed to deny the apparent continuity between NPI Exploitiveness and the similarly-named N P D criterion. Component content even seems to bespeak a kind of cynical pride in this ability to manipulate others, and as noted earlier, observer ratings indicate that high scorers on this component can and do engage in such behavior. Emmons' (1984) speculation thus seems justified, i f only as such less-than-socially-desirable behavior would likely have the concomitant of disrupting interpersonal relationships. It may also be the case, however, that the social skills of the normal narcissist include, as suggested earlier, the percipience to detect when this interpersonal maneuver can or should be most discriminatingly deployed, an ability that is not, by definition, characteristic of the behavioral inflexibility associated with the N P D . Finally, it is especially telling that what appears to be the strongest NPI component (i.e., with respect to explainable variance and convergent validation) is the one that bears no  54  apparent relation to any of the N P D criteria. Far from reflecting "the invariably fragile selfesteem" that is characteristic of the N P D , the strongest correlates of NPI Authority are instead the precise opposite, i.e., dominance, aggression, assertiveness, leadership, self-acceptance and self-confidence. While one could argue—as indeed many do (e.g., Millon, 1981)—that such apparent self-presentational bravado belies a hidden insecurity, the impressive convergence between observer and self-reports on these characteristics strongly suggests otherwise. A n d similarly, because Authority was (in the normal I P A R sample) the best predictor of trait ranking on Narcissism, it follows first, that what seems to contribute the most substantive variance to NPI narcissism is authority and dominance, two adaptive characteristics not encompassed by the N P D and second, that therefore, the issue of N P I - N P D continuity may be essentially irrelevant.  So while I can at least partially concur with the assertion that NPI "reflects some of the more important themes that have been historically attributed to the narcissism construct" (Raskin & Terry, 1988, p. 899), I also believe that by relying exclusively on the more negative and maladaptive conceptions typical of those same themes, Raskin and Terry have produced an instrument that only partially reflects what form a normal narcissism might take. A n exclusive focus on content redolent of cold exploitation, for example, eschewing that which taps the self-confident competitiveness described by Leary (1957) serves to all but guarantee a derivatively "undistinguished edition" of the pathological. The one thing that is distinctive about this edition is that despite its conceptual provenance in the N P D criteria, the first and strongest component to reliably emerge in NPI analyses is pure dominance, a factor, it bears repeating, that has no clear counterpart in those same criteria.  55  The Present Studies M y efforts in developing the Profile of Narcissistic Dispositions (POND) were aimed at delineating (and reliably measuring) the nature of normal narcissism, that is, a type of narcissism that is not inherently pathological. Based on the few theoretical treatments that even considered this possibility, conjectures about 'normal' narcissism were typified by seven themes: Forceful group presence; self-assurance, flaunting one's perceived attributes (e.g., vanity, exhibition); a drive to be first even among equals (e.g., competitiveness); a need to have one's talents recognized, lofty ambition, and—given the desire to be at centre stage—the necessity to manage one's fellow actors accordingly (e.g., manipulation).  It is difficult to say to what degree the NPI gauges these qualities. They certainly dd not match the NPI factors one-to-one. The NPI item selection appears to have been haphazard and, for purposes of measuring normal narcissism, seems based on the wrong conceptual template—i.e., the D S M III criteria for the Narcissistic Personality Disorder. Moreover, there is reason to question it on purely psychometric grounds.  Nonetheless, as the closest existing  approximation to a measure of normal narcissism, I used the NPI as a starting point. In Study 1, I assessed the structural characteristics of the instrument and evaluated how an alteration to the forced-choice format, i.e., to a Likert-type format, might improve reliabilty.  In part informed by those data, I then went on to develop an alternative inventory of 'normal' narcissism, the Profile of Narcissistic Dispositions (POND). I examined the P O N D ' s characteristics in a series of three studies. In Study 2,1 assembled items based on the seven themes isolated from the literature, and used principal component analysis to substantiate and clarify the factors. The first (unrotated) principal component was used to represent the core of  normal narcissism. In order to clarify the POND's meaning, I then assessed a number of its empirical correlates. Study 3 was designed to replicate the POND's componential structure and extend its nomological net. A s well as replicating these results, Study 4 was designed to see whether peer ratings could provide convergent validity for those self-reports correlates identified in Studies 2 and 3. A wide variety of peer ratings were explored to flesh out an image of how normal narcissists appear to others.  57  STUDY 1: What Can We Learn from the NPI? In addition to the several conceptual/interpretive problems beclouding the validity of narcissism as assessed by the NPI, the instrument is perhaps even more assailable on purely psychometric grounds; grounds which may be the very source of those same ambiguities. Consider even the fundamental issue of reliability: While the full-scale NPI shows an acceptable level of internal consistency (i.e., Guttman Lambda 3 = .83), the same cannot be said of the component scales in that four of the seven scale reliabilities are only <.54. Insofar as the NPI evinces measurement error on so basic a level, solutions emerging from structural analyses of the instrument could be misleading, if not actually spurious. This unreliability may in part explain why the number of NPI factors has been reported as low as four (Emmons, 1984, 1987) and as high as seven (Raskin & Terry, 1988). Thus the first purpose of Study 1 is to examine the NPI's structure in a large sample.  Forced-choice  vs. Likert-type format  Problems with the reliability of the NPI could be, in part, a function of the measure's item format. The 1988 NPI comprises 40 forced-choice dyads; although Raskin and Terry (1988) provide no rationale for the selection of this item format, a reasonable assumption is that it was adopted in order to control for social desirability. B y matching the desirability of the two alternatives, a forced-choice format can solve two problems. First, it eliminates the possibility that some or all respondents are answering according to the desirability rather than the actual content of the item. Second, it reduces the kind of skewed endorsement frequencies that can emerge when response choices vary in social desirability, as N P I items most certainly do. For example, the item "I am an extraordinary person" might reasonably be  58  expected to be positively skewed, owing to the relative undesirability of admitting to such sentiments. A critical assumption underlying the use of the forced-choice format, however, is that the items in the dyad be of comparable, if not fully equated, social desirability (Edwards, 1970, pp. 202-203); no such evidence is provided with respect to the NPI items. Other possible reasons for opting for a forced-choice format include practical considerations, such as simplicity of instructions for subjects, reduced administration time, and easier scoring. A potential concomitant of this scaling format, however, is a loss of information: Delimiting respondents' options to only two choices can preclude more finelygrained distinctions with regard to item self-descriptiveness. A forced-choice format may also be problematic in terms of subject comfort and acceptance: Jones (1968), for example, found that subjects preferred a multi-step over a two-choice format. A n d although Franken, Gibson and Rowland (1989) reported that subjects did not perceive a forced-choice format negatively, those authors' results are inconclusive insofar as no alternative was comparably evaluated. Velicer, DiClimente and Corriveau (1979), on the other hand, found that while subjects described two-point rating scales as easier to respond to, they actually preferred sixpoint scales, reporting that they could describe themselves more accurately with that format. In short, respondents prefer multi-step (i.e., Likert-type) scales to forced-choice formats.  A critical issue, however, is whether a forced-choice format suppresses or circumscribes valid true-score variance and therefore, reliability. Although Guilford (1954) concluded that individual scale reliability is a monotonically increasing function of the number of steps employed, others (e.g., Cronbach, 1950; Peabody, 1962) have argued that any ostensible increase so obtained owes more to the inflation of response style-related (i.e., error)  59  variance, rather than veridical refinement of measurement. Nunnally (1978) has noted that while error variance does indeed increase as does the number of scale points, true score variance increases at an even more rapid rate (p. 596). However, while Nunnally also concludes that, in terms of psychometric theory, the advantage "always is with using more rather than few steps" (1978, p. 595), the empirical evidence does not necessarily bear him o u t . Although Ferguson (1941) and Jahoda, Deutsch and Cook (1951) demonstrated 1  increasing reliability as a function of the number of scale points employed, Mattell and Jacoby (1971) found that both reliability and validity were independent of the varying numbers of scale points tested.  A study by Komorita and Graham (1965), however, revealed the potentially moderating influence of inter-item structure, or covariance, on reliability, finding that with a relatively homogeneous sample of items (i.e., 24 Semantic Differential evaluation items from Osgood, Suci and Tannenbaum, 1957), reliability was independent of the number of scale rating points (i.e., two-point scales were as reliable as those with six). With relatively heterogeneous items (i.e., 24 items from the CPI Socialization scales; Gough, 1957), in contrast, reliability did in fact increase (by .12 points) as scale points went from two to six.  Lissitz and Green (1975) pointed out that reconciling such results was impeded by two facts, viz., the differing methodologies employed in previous research and the fact that each prior study examined different instruments. In an attempt to circumvent these problems, they applied a Monte Carlo approach, constructing a mathematical model that systematically  Nunnally (1978) observes, however, that this advantage tends to level off at 7 points. However, I should note that selection of a 6-point scale for my instrument (derivation is described in Study 2) facilitates dichotomization (should analysis require it); neither Lissitz and Green's (1975) 5 nor Nunnally's (1978) 7 affords this possibility. 1  60  varied the parameters of item variance-covariance structure and number of scale points. Replicating Komorita and Graham (1965), they too found increased reliability as a function of the number of scale points, particularly when item structure was characterized by relatively low covariance. Lissitz and Green (1975) also reported, however, that such increases tended to level off at five scale points, prompting them to conclude that while an appropriate number of scale points is, of course, dictated by the nature of the particular measurement problem, the utility in employing scale points greater than five is minimal (p. 13).  Given this convergence between the Komorita and Graham (1965) and Lissitz and Green (1975) results, it seems reasonable to expect that with relatively heterogeneous content at least, increases in scale points up to five or six places should yield higher reliability than would two point measurement. With particular regard to the NPI, even though the full-scale alpha is acceptable, the only moderate reliability levels of several of the component scales suggest that any potential avenue toward increasing their values should be explored.  A s well as potentially increasing the reliability of the NPI and its subscales, the change from a forced-choice to six-place Likert item format could have another concomitant: A change in the componential structure of the instrument. Velicer and Stevenson (1978), for example, argued that such a change could underlie the apparently conflicting results in the "number of scale points" controversy, in that alterations in the format of the instrument could produce commensurate changes in its variance-covariance structure. Insofar as multi-point versions of an instrument's constituent items yield a structure differing from the forcedchoice format, then internal consistency indices calculated on the original form might be  61  inappropriate, particularly if the new structure is more complex (Velicer & Stevenson, 1978), or if items are reallocated across scales (Oswald & Velicer, 1980). Velicer and Stevenson (1978) examined these issues by comparing the relative merits of the standard two-choice format of the Eysenck Personality Inventory (Eysenck & Eysenck, 1969) vs. a 7 place Likert version of that measure.Their results (limited though they may be by a small sample size and a relatively 'unproven' component extraction technique and retention criterion i.e., Velicer's own) indicated that not only can a multiple-response format change an instrument's structure, but the more finely-grained response options may yield increased reliability of measurement (i.e., as indicated by higher eigenvalues) as well as accounting for significantly greater proportions of observed variance. With some alterations in method (including a larger sample size), Oswald and Velicer (1980) replicated the previous study, concluding, once again, that a Likert-type format does indeed permit more differentiated (i.e., finely-grained) responding by subjects and associated superior measurement of the factors underlying such inventories (p. 288). _  Therefore, the second goal of Study 1 is to compare the reliabilities that result from alternative item formats (i.e., forced-choice vs Likert-type). The findings of Velicer and his colleagues suggest that adopting a multiple-response item format could be a potentially useful vehicle for not only refining the reliability of the instrument but also providing a clearer view of the structure that underlies it.  62  Method Subjects Subjects were 654 lower level undergraduates (264 men and 390 women) who agreed to participate in a "personality questionnaire" study in exchange for course credit. Ages ranged from 19 to 37 years ( M = 21.3; S D = 3.1).  Instruments Narcissistic  Personality Inventory (NPI; Raskin & Terry, 1988). During the first testing  session (see Procedure, below) subjects received either the standard forced-choice  version of  the NPI (designated hereafter as NPI-FC) or an eighty Likert-type item version (NPI-L) in which each alternative comprising the NPI forced-choice dyads was rated separately on a scale ranging from "extremely inaccurate" (1) through "extremely accurate" (6). Order of presentation of the resulting 80 items was randomized. A copy of each instrument is in Appendix A .  Procedure Subjects participated in two separate testing sessions in groups of 5 to 15. In the first 50minute session, subjects received a test battery comprising several different personality measures  2  in addition to either N P I - F C or NPI-L. During the second 50-minute session  (approximately one week later), another battery was administered; the alternative NPI form not received in the first session was completed at this time. Order of presentation of the NPI forms was randomized, such that one half of the subjects received N P I - F C in session 1 and N P I - L in session 2; the remaining subjects were administered the forms in reverse order. The positions of the NPI forms remained constant within their respective batteries. Following  2  Only results pertaining to the NPI-FC and NPI-L forms are reported here.  63  completion of session 2, subjects' questions concerning the nature of the study were addressed and they received a written debriefing.  Results and Discussion Structural and Component  Analyses  Analysis 1: Tetrachoric treatment of forced-choice  items  Assessing the stability of the seven-component solution reported by Raskin & Terry (1988) required a replication of their methodology, including the extraction/retention of seven components. The first structural analysis of the N P I - F C was a principal components analysis of the inter-item tetrachoric correlation coefficient matrix. Following their procedures, seven principal components were extracted (using the Alberta General Factor Analysis Program; Hakstian & Bay, 1977; incidentally, 12 factors had eigenvalues > 1) and rotated to a weighted promax oblique solution, that accounted for 23.54% of total N P I - F C variance (vs. Raskin & Terry's reported 52%). The component structure, eigenvalues and component variances and intercorrelations are shown in Tables l a and lb.  Insert Tables l a and l b about here  I interpreted the seven components (attempting, where possible, congruence with those of Raskin and Terry, 1988) as follows: I. Leadership/Authority; II. Vanity; III. Entitlement; IV. Grandiosity/Arrogance; V . "Manipulativeness"; VI. "Immodesty"; and VII. Exhibition; components V . and VI. have no direct counterpart in Raskin and Terry, 1988; their original  64  results and interpretations are reported on pps.23-24). There are several noteworthy features of this solution. First, as Raskin & Terry (1988) reported, there appears to be reasonable evidence for a general component of narcissism, in that 36 of the 40 N P I - F C narcissistic alternatives load positively and nontrivially (i.e., at levels > .33) on an unrotated first principal component; the remaining four items load at levels ranging from .10 to .32 (M = .25). Second, with respect to Raskin & Terry's (1988) component retention criterion of "a minimum of 3 items loading univocally (approximately .50 and higher)" (p. 893) on a given component, one sees that in the solution obtained here, only 6 of the 7 components actually fulfill this requirement. Specifically, one component—emerging sixth in the current solution— comprises only a doublet of two Raskin & Terry (1988) Superiority items; while one of these items is indeed univocal, the second has loadings >-.33 on three additional components. Third, the N P I - F C structure obtained here is somewhat ill-defined, insofar as relatively few items evince clear univocality: 4 8 % (19) of the N P I - F C items load at levels >-.40 on up to three additional components. If one employs a more common (albeit less conservative) level of interpretation for a multiple loading (i.e., >-.33), fully 87.5% of the N P I - F C items are "multivocal." Using the more conservative .40 criterion, however, one sees, 3  specifically, that among the items that are not univocal, seven load on one additional component, nine have loadings on two other components, and three items load on three components other than those which they ostensibly mark. Nunnally (1978, p. 418) has referred to criteria of "strong structure":  In Raskin and Terry's (1988) report, only three items fail to load univocally as per the .33 criterion.  3  65  In terms of factor analytic products, this means that each factor has some variables that load mainly on it alone, and there are at least four tests [variables] with loadings above .50. Unless a factor is at least that strong, it would be best to ignore it. Examination of the solution reproduced in Table l a suggests that in terms of these strong structure criteria, at least three of the seven N P I - F C components could profitably be "ignored." A s to the conceptual correspondence of the present components with those reported by Raskin and Terry (R & T ; 1988) one sees again, some significant divergences. Considering each component in turn, it appears first that Leadership/Authority (I) and Vanity (II; e.g., "I like to look at my body", .86) are the only components to retain most of their original item complements. However, although Leadership/Authority is marked by 7/8 of the original R & T (1988) Authority items (e.g., "I would prefer to be a leader," .86), two R & T (1988) Entitlement items (e.g., "I have a strong will to power"; .71) also drift to this component. Tapping content appearing to relate to ambition, the drift of these items lends a more poweroriented flavour to the component. A s noted above, Vanity replicates perfectly; Entitlement, while retaining three of its designated items (e.g., "I expect a great deal from other people"; .63) is reduced by half by the drift of three designated items to Leadership/Authority. The fourth component—interpreted here as Grandiosity/Arrogance, is an amalgam of 4 R & T (1988) Self-Sufficiency items (e.g., "I am going to be a great person"; .79), two Superiority items (e.g., "I am an extraordinary person"; .75) and two Exploitativeness items ("I can read people like a book"; .61). The drift of these items seems conceptually consistent, the apparent commonality being a generalized sense of self-inflation. The fifth component emerging in this solution has no direct counterpart in the structure reported by R & T (1988).  66  Comprising two Raskin & Terry Exploitativeness items (e.g., "I find it easy to manipulate people"; .76) and two Exhibition items (e.g., "I would do almost anything on a dare"; .66), the content of this "Manipulativeness" component is reminiscent of the cool manipulativeness and impulsivity said to characterize the sociopathic personality. The sixth component is a doublet of two R & T (1988) Superiority items (e.g., "I know that I am good because everybody keeps telling me so"; .78) and seems best characterized as immodesty or conceit. The seventh and final component retained clearly reflects Exhibition (e.g., "I like to be the center of attention", .85), retaining 5 of the original seven R & T (1988) Exhibition markers; the items that drift from Exploitativeness ("Everybody likes to hear my stories"; .52) and Superiority ("I wish that someone would someday write my biography"; .51) are similarly consistent with the need for attention and/or recognition that is suggested by the other items marking this component. In sum, the seven component solution reported by R & T (1988) was not fully recoverable in a large (N = 654) sample. It appears that item clusters reflecting Leadership/Authority and Vanity are most stable, and Exhibition-related items are also fairly cohesive albeit less so in comparison to the first two components. O n the basis of these data, I concluded that the 7 component solution emerging from analysis of the inter-item tetrachoric correlation matrix is unstable and probably overextracted. Further, it seems clear that most of the substantive and empirical variance is attributable to items reflecting leadership and authority or, more generally, surgent content.  Analysis 2: Forced-choice  items and phi coefficients  67  A second attempt to replicate the NPI structure reported by R & T (1988) was undertaken, this one utilizing the matrix of inter-item phi coefficients (instead of tetrachoric correlations) as the input for a principal components analysis (PCA). Using the joint criteria of the Kaiser-Guttman rule and the elbow of the scree plot, seven components were extracted and rotated to an oblique solution that accounted for 42.6% of observed N P I - F C variance. Although structurally similar to the P C A results described above, there were some significant divergences in component composition. The component structure, eigen-values and component variances and intercorrelations are shown in Tables 2a and 2b.  Insert Tables 2a and 2b about here  The Leadership/Authority (I) component obtained in the tetrachoric analysis very nearly emerged intact here, with the exception of one Raskin & Terry (1988; R & T ) Authority item drifting to the sixth component (see below). R & T's (1988) Vanity (II) component again replicated perfectly here; the Entitlement component obtained in the first analysis also emerged intact. The fourth component, Grandiosity/Arrogance, was a near replication, losing only one item to the fifth component. This latter cluster, here termed Superiority, has no direct counterpart in the first analysis. Comprising three R & T (1988) Exhibition items (e.g., "Modesty doesn't become me"; .63), one R & T (1988) Superiority item ("I know that I'm good because everybody keeps telling me so"; .48), two R & T (1988) Exploitativeness items (e.g., "I find it easy to manipulate people"; .47) and one R & T (1988) Self Sufficiency  68  item, this component appears to be an amalgam of arrogant self-inflation and brash selfconfidence (e.g., "I'm more capable than other people"; .43). In this analysis, the sixth component, Manipulativeness, is reduced to a triplet comprising two R & T (1988) Exploitativeness items (e.g., "I can make anyone believe anything I want them to"; .52) and one R & T (1988) Authority item ("I have a natural talent for influencing people"; .39). The commonality is clearly manipulative content. The seventh and final component, Exhibition, is a clear reproduction of that obtained in the first analysis, similarly marking need for attention and/or recognition. While this P C A provides a fair replication of the seven component solution observed in the first analysis, the structure tends to destabilize at the 5th to 7th components (i.e., items tend to be especially multivocal, content clusters are less well-defined, etc). This P C A of the phi coefficient matrix does, however, account for considerably greater variance (i.e., 19 %) than does that based on the matrix of tetrachoric correlation coefficients. Comparable to the first analysis, 4 8 % (19) of the N P I - F C items show double loadings; these are, however, typically lower. Leadership/Authority is more clearly defined here, with only three marker [i.e., R & T (1988) Authority] items showing substantial loadings on additional components; no such double loadings exceed .39. In sum, P C A s on two different types of correlations yielded structures that were relatively comparable. At the same time, however, that comparability (i.e., in terms of order of emergence, marker items and loadings, etc.) appears limited to the first four components.  69  A n d as compared to the structure described by R & T (1988), only Vanity fully replicates; Leadership/Authority and Exhibition are relatively structurally stable.  4  Analysis 3: Likert items The third structural analysis undertaken was a P C A of the 40 narcissistic alternatives  5  of  the Likert-rated NPI items (NPI-L). This analysis represented an attempt to determine whether the potential reliability increase often associated with a Likert rating format would have a concomitant influence on the observed structure of the NPI. Because my intent was to attempt to recover a structure similar to that reported by R & T (1988), the 40X40 matrix of inter-item product moment correlations was submitted to a P C A , and 7 PCs were extracted and rotated to an oblique solution that accounted for 53 % of cumulative N P I - L variance. The component structure, eigenvalues and component variances and intercorrelations are shown in Table 3.  Insert Table 3 about here  Examining the individual components in turn, one sees first, that as with my tetrachoric and phi coefficient analyses, Leadership/Authority remains relatively stable, retaining 6/8 R & T (1988) Authority items (vs. 8/8 and 7/8 in the first two analyses, respectively). Four  In a separate analysis, I attempted to recover Emmons' (1984, 1987) four component NPI solution; the only component to approach replication was Leadership/Authority, retaining 8 of 8 of Emmons' designated Leadership/Authority items but also picking up five items from the other three components. I note that as I construe the original scale developers' results/interpretation more directly 'definitive', I focus primarily upon their findings as opposed to those of Emmons. That is, the alternatives of the forced-choice pairs that are keyed in the narcissistic direction.  4  5  70  non-Authority R & T (1988) items drift to this.component however, among them two R & T (1988) Exhibition items (e.g., "I really like to be the center of attention:"; .68), one R & T (1988) Self-Sufficiency item ("I like to take responsibility for making decisions"; .55) and one R & T (1988) Entitlement item ("I have a strong will to power"; .50). Despite their original R & T (1988) designations, these drifted items seem more conceptually consistent with a general surgency, or self-assured ascendancy component. A s in the first two P C As, the only component to fully maintain its structural integrity is that tapping Vanity, comprising 3/3 of the original R & T (1988) items. The structure of the third component, Exploitativeness, is, for the first time in this series of analyses, a near replication of that reported by R & T (1988). Five of the original five R & T (1988) Exploitativeness items define this component; in addition, however, one R & T (1988) Authority item drifts here ("I have a natural talent for influencing people; .70) as does a R & T (1988) Exhibition item ("Modesty doesn't become me"; .40). The fourth component is a triplet of R & T (1988) Self-Sufficiency items (e.g., "I always know what I'm doing"; .64) and one Superiority item, while the fifth seems best described as Impulsive Exhibition, marked by two R & T (1988) Exhibition items (e.g., "I would do almost anything on a dare"; .59) and one R & T (1988) Superiority item ("I can live my life any way I want to"; .50). The sixth component, best interpreted as Superiority/Arrogance, has no clear counterpart in either the previous analyses or in Raskin & Terry's (1988) reported structure. Instead, it is an amalgam of two R & T (1988) Superiority items (e.g. "I think that I am a special person"; .77), two R & T (1988) Self-Sufficiency items (e.g., "I am going to be a great person"; .75),  71  one R & T (1988) Authority item ("I will be a success"; .71), and one R & T (1988) Entitlement item ("If I ruled the world, it would be a better place"; .42). The seventh and final component corresponds fairly closely to Raskin & Terry's (1988) Entitlement cluster, retaining 4 of those 6 original items; two additional items drift here, one R & T (1988) Exhibition item ("I'm apt to show off if I get the chance"; .43), and one Superiority item ("I like to be complimented"; .41). In addition to the alterations in structure noted above, the P C A solution from this analysis of the N P I - L Narcissistic alternatives has several other noteworthy features. Once again, there is reasonable evidence for what might be interpreted as a general component of narcissism, insofar as 36/40 items load >• .33 (and often considerably higher, e.g., .59, .55, .51, etc.) on a first unrotated principal component. What is perhaps most striking about the rotated solution however is its ill-defined structure: fully 70 % (i.e., 28) of the N P I - L items load at levels >• .33 on at least two components, nine on three components, and seven items load comparably on four components. While some proportion of double loadings might be anticipated in an oblique solution, this pattern of double, triple and even quadruple loadings seems inappropriate insofar as it is clear that several items are not uniquely identified with their ostensive or designated components. For example, the R & T (1988) Entitlement item "I have a strong will to power" has its highest loading (i.e., .50) on the Leadership/Authority component in this solution; it also loads, however, on three other components: .47 on Exploitativeness, .47 on Impulsive Exhibition, and .35 on Superiority/Arrogance; note that it does not load on the Entitlement component. In short, this item, while perhaps a good marker of authority-related content, shows absolutely no discriminant validity with respect to  72  the other components. A n d I reiterate that this absence of univocality characterizes at least 70% of the NPI items. This third attempt to recover the Raskin & Terry (1988) seven component solution for the NPI thus met with limited success. A s in the first two P C A s , the structures of the Authority and Vanity-related components are fairly well preserved, the latter again replicating fully. A n d while Exploitativeness and Entitlement each corresponded fairly closely to the item clusters reported by R & T (1988), the remaining components were fractionated and unstable. Once again, the most replicable aspects of the putative NPI structure are those relating to Authority and Vanity; these components also contribute the most empirical (and substantive) variance, i.e., 26.5 and 7.2 % respectively, together accounting for almost 34 % of the 53 % of total observed N P I - L variance. (Parenthetically, I note a similar result in analysis of the N P I - L 80, in which the first two components - one a clear Authority, the second a cluster of Vain Exhibition-related items - together accounted for 25 % of cumulative variance , i.e., 42%).  -  Taken together, the results of the three P C A s indicate that the seven component solution that R & T (1988) propose for the NPI is is difficult to replicate in a large-scale sample. Analyses of the tetrachoric and phi coefficient inter-item correlation matrices yielded relatively comparable structures, although in both analyses, the Superiority and Manipulativeness components tended to fractionate. It is also true, however, that with the exception of the Vanity component—the only one to replicate fully across formats and  73  methods of analysis—even the relatively stable components did not precisely match those described by R & T (1988).  Reliability  Analyses  A s a check on the comparability of the present sample to that of R & T (1988), I first calculated the usual normative statistics, including means, standard deviations and intercorrelations for the full scale N P I - F C , the original seven sub-scales and of course, their reliabilities; these results, along with the comparable statistics reported by R & T (1988) are shown in Tables 4a and 4b.  Insert Tables 4a and 4b about here  Several features of these comparative tables are of note, most basic among them the fact that normatively, the present sample appears fully comparable to Raskin & Terry's; while there are some slight differences between component and full scale means, they are trivial insofar as the highest discrepancy is one half of a point. Similarly, while the present sample appears to have slightly higher variance, the largest difference between component scale standard deviations is + .37. With respect to actual item intercorrelations, those obtained in the present sample range from -.09 to .71 ( M = .15); R & T (1988) do not report their mean inter-item correlations. Component and full scale intercorrelations, on the other hand, are considerably higher in the present sample: While the range of component scale intercorrelations is comparable (i.e., .11 - .42 vs. .13 - .48 in the R & T (1988) and present samples, respectively), the mean value  74  here is somewhat higher (i.e., .33 vs. R & T's .27). Similarly, the mean component scale 6  full scale correlation calculated from R & T's (1988) Table 2 is .44, the comparable value here is a substantially higher .65 . Thus despite the apparent normative comparability of the two samples, it is evident from these data (and those of the component analyses) that the inter-item covariance structure observed in the present sample is significantly different (i.e., showing greater homogeneity) than that reported by R & T (1988). I note that while the present sample (n=553) is considerably smaller than that described by Raskin & Terry (1988; n=1008) it is certainly sufficiently large to permit a reliable multivariate replication (i.e., being at least 150 Ss  greater  than  that suggested by the Nunnally algorithm of  ten times as many subjects as variables). In short, it does not seem reasonable to argue that the apparent differences in heterogeneity can be attributed to sample size effects. A possible alternative explanation is that the lower heterogeneity in this sample is, in part, an artifact of the product moment correlation coefficient estimates employed in the present analysis of the dichotomously scored N P I - F C , i.e., the phi coefficient; as noted earlier, R & T (1988) advocate the use of the tetrachoric estimate. Recall, however, my earlier discussion of the relative merits of the two coefficients, in which I observed that questions regarding their relative superiority primarily concern the level of under-or overestimation of the P M coefficient as a function of departures from bi or multivariate normality (Lord & Novick, 1968). While positions differ as to which estimate is more robust in the face of such departures, most of those reviewed (e.g., Carroll, 1961; Glass & Stanley, 1970) hold that the phi coefficient will tend to underestimate the P M  I note that Raskin & Terry (1988) report two different values for the average intercomponent correlations: .22 on p.893, .27 on p.895; from the values Raskin & Terry provide, I calculate .27.)  6  75  coefficient that would be obtained with data from normally distributed continuous variables, while the tetrachoric will tend to overestimate those same relations (Glass & Stanley, 1970; Nunnally, 1978). Note, however, that irrespective of any possible departures from normality in the underlying distributions, the component scale intercorrelations reported here are based initially on phi coefficient matrices and thus might be expected to underestimate the intercorrelations among NPI items; the relatively higher (vs. R & T's reported values) component intercorrelations shown in Table 4a would thus seem to belie the methodological artifact explanation. The remaining explanation for the relatively higher scale homogeneity in the present sample is simply, that the "latent item structure" that Raskin & Terry (1988, p.893) attributed to the NPI (and claimed would be more clearly articulated by the use of tetrachorics), is less clear and its component-based scales more interdependent than was originally reported. One final point regarding the phi vs. tetrachoric controversy: Both coefficients are, of course, based on the relations between dichotomously scored variables assumed to be continuous; disputes concerning the relative superiority of the coefficients are, of course moot, if the variables are in fact, measured continuously. A second aim of Study 1 was to measure the NPI items accordingly, and thus gauge the effect, if any, of scale rating format on estimated reliability.  In their 1988 report, Raskin and Terry calculated Guttman's Lambda 3 in estimating the internal consistency of the component-based scales. Because Lambda 3 and coefficient alpha (Cronbach, 1951) are functionally empirically equivalent as lower-bound estimates of scale reliability, I preferred to adopt coefficient alpha—the standard, and better understood, estimate of internal consistency.  7  76  A s can be seen in Table 4a, substituting a six-place Likert scale for the two point forced choice item option does indeed yield increments in scale reliability. For example, while the 8  R & T (1988) N P I lambda 3 values range from .50 to .73 (M = .58), alpha coefficients for the comparable N P I - L scales range from .64 to .88, with a substantially higher mean value of .72; Raskin & Terry's full scale lambda 3 (i.e., .83) is similarly surpassed by the full scale alpha of .92 for the 40 item NPI-L.  Conclusions The results of the series of structural analyses described above indicate some doubt as to the structural integrity of a seven component solution for Raskin and Terry's (1988) Narcissistic Personality Inventory. It is, of course, true that some characteristics of the present group may have differed from those of the original sample; it is also true, however, that age and sex breakdowns were comparable, and scale normative statistics were highly similar (see General Discussion).  A n d I note once more that the instability observed characterizes  structures emerging from component analyses of both tetrachoric and phi correlation coefficient matrices. While components comprising items reflecting Leadership/Authority, Exhibitionism, and to a lesser extent, Entitlement and Exploitativeness, are relatively stable across analyses, Vanity is the only component to consistently retain structural integrity, i.e., full replicability. Evidence for the univocality of NPI items is often striking in its absence: Double, triple and even quadruple item loadings appear to be the rule rather than the exception in that  Parenthetically, it is of note—and consistent with the apparently higher inter-item homogeneity noted above—that even the reliabilities of the original NPI-FC scales are slightly higher here than in Raskin & Terry's (1988) sample.  77  across methods and analyses, a range of 47 to 87.5% of items evince loadings at levels >• .33 on at least one additional component. Such "multivocality" similarly characterizes the structural solution emerging from the analysis of the product moment correlation matrix of the 40 Likert-rated narcissistic alternatives of the NPI items: Again, fully 70% of those items loaded at levels •> .33 on at least two components. A n d as noted earlier, while some proportion of double loadings should be expected in oblique solutions, the pattern of multiple loadings obtained across methods and item formats raises serious doubts concerning the discriminant validity of a disturbingly large proportion of the NPI items. The test of an alternative item format (i.e., the six-place Likert scale) yielded a number of interesting concomitants. First, consistent with the findings of Velicer and his colleagues, substituting a Likert for the NPI's usual forced-choice item format did indeed alter the componential structure of the instrument: The observed structure was more complex and less differentiated (e.g., Velicer & Stevenson, 1978) than that reported by Raskin and Terry (1988), and a significant proportion of items were reallocated across scales (e.g., Oswald & Velicer, 1980). These latter authors have also suggested that a multiple response format facilitates increasingly differentiated or more finely-grained responding by subjects and thus improved measurement of the factors underlying Likert-type instruments (i.e., additional variance is accounted for). While my seven component solutions typically yielded lower variance accounted for than that reported by Raskin and Terry (1988), it is noteworthy that the P C A of Likert-rated items was the only analysis in which variance accounted for met or exceeded their 52% figure. It is similarly noteworthy that the seven component solution  78  extracted from the Likert-based product moment matrices—while evincing considerable divergence from that of Raskin and Terry (1988)—was nonetheless the one most structurally similar (i.e., in terms of item allocation, comparability of component composition, etc.) to the 7 component structure described by those authors. Again, as Velicer and colleagues might have predicted, component-based scales built from the Likert-rated narcissistic alternatives of the NPI were considerably more reliable than the comparable forced-choice scored scales, yielding a mean alpha of .72 as compared with Raskin & Terry's (1988) mean lambda 3 of .58; the full-scale alpha was similarly nine points higher (i.e., .92) than that reported by those authors. These results seem ironic given that they obtain  despite the relative nonreplicability of the component solution upon which  the scales are based. In evaluating the results of the three principal component analyses described above, the most tenable conclusion is that a seven component solution for the NPI artificially fractionates or divides common substantive variance in the instrument. A n d further, the single result that replicates across methods and formats is that components reflecting Leadership/Authority and Vanity are, without question, the most stable, and additionally, contribute the lion's share of the common substantive and empirical variance (e.g., almost 34% of the 53% of the observed variance in the P C A of Likert-rated items). In my earlier discussion of theoretical conceptualizations of the construct of narcissism, I noted that a common feature was the relative centrality of characteristics reflecting social dominance. The present study shows that such content consistently forms the strongest and  " '  79  most replicable NPI component —across both methods of correlation and item formats 9  (note that this result obtains despite the fact that DSM-III R criteria—the "conceptual template" for the N P I — m a k e no ostensive reference to dominant or surgent behaviour). Thus, given the conceptual/theoretical and empirical convergence on these points, the evidence available strongly suggests that some variety of dominance-related behavior forms a significant proportion of the universe of content of normal narcissism. Those characteristics do not, however, exhaustively define that same universe; other traits are likely equally important. Study 2 represents an attempt to more fully define that universe.  I wish to make it clear that I do not dispute that the NPI gauges what I (and a number of theoreticians) construe as an important aspect of non-pathological narcissism, i.e., social dominance. I just don't think that it measures it (or other narcissism-related traits) especially well. 9  80  STUDY 2: Scale Development and Initial Correlates The aims of this study were two-fold. First, I began the derivation of my new instrument, the Profile of Narcissistic  Dispositions-Preliminary  form (POND-P) by assembling a large pool of  items designed to gauge the seven features of non-pathological narcissism. These included what my Study 1 data indicated were the more stable NPI items as well as original items more fully reflective of the facets that might comprise 'normal' narcissism. Given my demonstration that item format has a significant impact upon reliability of measurement, I adopted Likert-type measurement. Although I anticipated that narcissism would be a relatively unitary construct (as evinced by all items loading on an unrotated first principal component), I felt that advancing more specific predictions (e.g., regarding potential component configurations) would be premature. Similarly, owing to expansion of the universe of content, advancing specific predictions regarding the component-based scales of the P O N D - P was considered premature. However, because of the nature of that expansion, items corresponding to the seven following facets (as described more fully in Item Development, below) were included and therefore could form separate factors: Dominance/leadership, self-assuredness, manipulation, competition, vanity/exhibition, desire and/or need for recognition/admiration and ambition.  The second aim of this study was to provide initial validation of the resulting measure by mapping its nomological net of relations with a variety of self-report measures designed to assess interpersonal and other personological variables. Therefore, initial consideration of its correlates seems the preliminary step to understanding the scope of the P O N D - P .  81  T o examine its structural validity, the first step was to determine whether all the subscales loaded upon a first (unrotated) principal component that represents the core of narcissism, i.e., the variance that the facets share. The second step was to project the measure (and its hypothesized subscales) onto interpersonal space, i.e., that defined by the two circumplex axes. Because of the influence of Leary's (1957) conceptualization on my selection of items, the P O N D - P and its subscales were expected to fall within the the upper left quadrant of the circumplex. Inclusion of the B i g Five inventories (i.e., the Five Factor Inventory and the I A S - R B5) permits examination of the relations between the P O N D - P , its anticipated facets, and what are construed by many as the fundamental dimensions of personality, i.e., the familiar B i g Five. Because of the substantive commonalities between many provisional P O N D - P items and certain B i g Five dimensions, I expected certain general trends to obtain: Extroversion:  There should a positive association between the P O N D and extraversion.  This follows because dominance and leadership-related content as well as vanity and/or exhibition-type content is moderately to strongly related to the concept of extraversion (McCrae & Costa, 1989). Agreeableness: Given the coldness inherent in some manipulativeness items, one would expect moderate to strong negative relations to obtain here; the same would likely hold for competitiveness-related content. With respect to the remaining B i g Five (i.e., Conscientiousness, Neuroticism and Openness to Experience), specific predictions were difficult. A s regards Conscientiousness, for example, although a narcissist might be expected to be conscientious with respect to  82  certain kinds of personal goals, for example in situations fostering competition or potential personal aggrandizement, social conventions regarding scrupulousness, dependability and predictability are quite likely to be ignored. Given this indeterminacy, predictions specific to this and the remaining two dimensions were not advanced.  Several additional measures were employed to clarify what is assessed by the P O N D - P , among them instruments designed to gauge reported interpersonal problems, and interpersonal dependency. While some of these measures were selected primarily as marker variables, other were included on more substantive grounds. A s regards dependency, for example, as that and insecurity are supposed by some (e.g., Millon, 1981) to underlie the narcissist's putative self-aggrandizement, those variables are assessed with the Hirschfeld et al. (1977) Interpersonal Dependency Inventory (IDI). Here, I anticipated positive relations between the scales and the expected need for attention/admiration items. In contrast, I expected moderate to strong negative relations between the IDI and the dominance/leadership items.  A n d finally, results relating to the Inventory of Interpersonal Problems (IIP-C) were hoped to be especially informative in terms of the interpersonal style gauged by the P O N D and its components. One might expect, for example, that individuals endorsing items reflecting manipulative/competitive content might be expected to report problems related to Vindictiveness ( B C ; e.g., "I want to get revenge against others too much"). Similarly, persons scoring high on ascendance-type items might endorse Domineering (PA) problems (e.g., "I want to control others too much"). Alternatively, individuals scoring high on these dimensions might not perceive such behaviours as problematic.  83  Method Subjects 660 lower-level undergraduates (259 men and 401 women) served as subjects in exchange for course credit. Their ages ranged from 18 to 36 years ( M = 22.2, SD = 3.3). Instruments (a copy of each measure may be found in the Appendix) Five Factor Inventory (FFI; e.g., Costa & McCrae, 1989). The F F I is a 60-item short form of Costa and McCrae's (1985) Neuroticism, Extraversion and Openness to Experience Personality Inventory (NEO-PI), and provides self- and/or other-report measures of the B i g Five dimensions. Subjects rate each statement (e.g., "I like to follow a strict routine in my work") on a five-place Likert scale ranging from "strongly disagree" (1) to "strongly agree" (5). The instrument yields domain scores for the orthogonal dimensions of Neuroticism (N), Extraversion (E), Openness to Experience (O), Conscientiousness (C) and Agreeableness (A). Costa & M c C r e a (1989) report acceptable reliability for the individual scale scores, and present an impressive array of evidence attesting to the validity of the measure. Revised Interpersonal Adjective Scales—Big  Five version (IAS-R B 5 ; Trapnell & Wiggins,  1990). Representing an extension and revision of the Revised Interpersonal Adjective Scales (IAS-R; Wiggins, Trapnell & Phillips, 1988), I A S - R B5 comprises 124 adjectives that together, mark the B i g Five factors and the eight dimensions of interpersonal behavior. Subjects rate the self-descriptive accuracy of each item (e.g., "dominant") on an eight-place Likert scale ranging from "extremely inaccurate" (1), through "extremely accurate" (8). From these ratings, two sets of scale scores may be derived: Interpersonal octant scores, and B i g Five scale scores.  Scores for the interpersonal variables (e.g., Assured-Dominant, P A ; Arrogant-Calculating, B C , etc.) are calculated from the eight, eight-item scales of the original IAS-R, an instrument shown to have acceptable reliability and meaningful correlations with other self-report measures (e.g., Wiggins & Broughton, 1985) as well as non-test behaviors (e.g., Gifford & O'Connor, 1987). B i g Five domain scores (i.e., Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness) may be calculated from the I A S - R B 5 ; coefficients alpha are all highly reliable (i.e., ranging from .87-.93) and Trapnell and Wiggins (1989) report that I A S - R B5 has good convergent and discriminant properties vis-a-vis other B i g Five measures as well as having the added advantage of providing reliable and valid measures of the eight dimensions of interpersonal behavior. Results pertaining to the IAS B5 scales are not reported here (but are available from the author); they were, however, consistent with those of the FFI.  Interpersonal Dependency Inventory (IDI; Hirschfeld, Klerman, Gough, Barrett, Korchin, & Chodoff, 1977). The IDI is a 20-item, four scale inventory that is said to reliably assess cognitive dependence upon others. For brevity's sake, full IDI results are not reported here (they are, of course, available from the author) but representative correlations are noted below. Inventory of Interpersonal Problems - Circumplex version (IIP-C; Alden, Wiggins & Pincus, 1990). The IIP-C is a 64-item, 8 scale measure for the assessment of a wide range of interpersonal problems. Originally derived from presenting problems of outpatient psychotherapy intake interviews, the items include behaviours that are "hard for you to do" and those that "you do too much." Each item is rated on a 5 place scale ranging from "not at all" (0) to "extremely" (4). The eight scales (each of which corresponds to the appropriate octant on the interpersonal circle) include Domineering (PA; e.g., "I try to control other people too much"); Vindictive ( B C ;  85  e.g., "It is hard for me to trust other people"), etc. Scale reliabilities are acceptable, ranging from .72 (Intrusive) to .85 (Socially Avoidant). The Profile of Narcissistic  Dispositions—Preliminary  Form (POND-P). Thirty-six items  culled from the 80-item Likert version of the NPI were interspersed with 67 original items (see Item Development, below) to yield the initial form of the P O N D - P . Items were rated on a 6-place Likert-type scale ranging from, "strongly disagree" (1) through, "strongly agree" (6). Half of these items were written to reflect the presence of the disposition of interest, half to reflect its absence, i.e., an attempt was made to balance self-ascription vs. denial of the trait. Cheek Shyness scale-Revised (1983). Cheek Shyness is a 13-item instrument designed to measure social inhibition and anxiety. Items are rated on a 7-point scale ranging from (1) "not true" to (7) "very true". The measure is reported to have acceptable reliability and validity.  Item Development and Selection for the POND-P If one examines the substance of the narcissistic alternatives of the NPI dyads, several items emerge as potential markers of what, theoretically, one might expect to be some of the facets central to what normal narcissism might comprise. For example, the items "I am a born leader" and "I am assertive" seem to tap the sense of authority and willingness to assert the self that are characteristic of self-assured ascendance, while items such as "I find it easy to manipulate people" reflect self-ascription of the interpersonal skills necessary to achieving that end. Similarly, statements such as "I like to look at my body" and "I am an extraordinary person" capture the narcissist's tendency to "admir[e] his own strength or beauty or wisdom" (Leary, 1957, p.333); the complementary item "I get upset when people don't notice how I look when I go out in public" reflects the frank irritation when, surprisingly, observers' admiration is not  86  similarly forthcoming. Conversely, items such as "I am going to be a great person" reflect a kind of imperial ambition that would dwarf even an unusually high need for achievement, while the statement "I always know what I'm doing" is a kind of neurotically blanket denial of even the most trivial uncertainties that, surely, almost everyone experiences from time to time (even Caesar occasionally consulted the Pantheon.) In short, while the NPI does seem to include several likely markers of non-pathological narcissism, many items are, in my view, somewhat grandiose or simply inappropriate in trying to adequately capture the construct.  At the same time, however, it might be equally grandiose to outright reject the more useful results of the efforts of previous item developers (viz., the NPI item "I am more capable than other people"). A n d much as it was justifiable for Gough to use the "seed" of established items with known empirical properties (i.e., 'borrowing' 38% of the CPI items from the M M P I ) , so too did it seem appropriate to include in the initial P O N D - P pool those NPI items that appear to strongly and univocally mark factors that have reliable and substantively consistent relations with external variables; for the most part, those items tapping leadership and vanity/exhibition-related behaviors. Selection of N P I - L items from the provisional pool was guided by several considerations, foremost among them substantive consistency with the literature on normal narcissism. Reflecting the common conceptual themes in both Freudian and Neo-Freudian views on narcissism, N P I - L items that appeared to gauge the following seven characteristics were provisionally selected: Dominance/authority, self-assuredness, willingness to strategically deploy interpersonal manipulation techniques, competitiveness, exhibition/vanity, desire and/or need for recognition and finally, ambition.  87  Item selection was further informed by multivariate and psychometric features of the present data set, among them whether a potential item showed consistently strong (i.e., > .40) loadings across methods and solutions, evinced a reasonable degree of univocality (i.e., loadings > .33 on non-designated scales), and further, that the components upon which they loaded retained a reasonable degree of structural integrity across those same methods and solutions. The strongest candidates approaching these inclusion criteria were, for the most part, markers of the leadership/authority and vanity/exhibition-related components; additional items tapping content reflecting manipulativeness, need for recognition and self-assuredness were also selected for the provisional pool.  The initial item pool required substantial (and substantive) expansion because the NPI did not fully cover the seven content areas thought to more fully reflect normal narcissism. I should like to stress here that in contrast to the genesis of some other inventories, that of the P O N D was not limited to derivative versions of diagnostic criteria, blind empiricism or serendipity. Instead, I exhaustively reviewed the available literature and sought its commonalities. Turning then to the specific nature of the expansions proposed for each of the facets of normal narcissism listed above, I note that given their interconnections, it makes substantive sense to discuss them under four general headings: I first considered authority-related behavior.  Dominance/Leadership NPI Authority seems to primarily gauge a perception of the self as an authoritative and skilled leader; as such, an NPI item like "I see myself as a good leader" should be an adequate marker of this aspect of dominance. What might be uniquely narcissistic about this ascendance, however, is, in addition to the perception of self as an effective or authoritative leader, one might  88  express a need to be valued or recognized as such by others—or at least believe that one is so perceived. A n item such as "People seem to instinctively trust my judgment" might be expected to tap this self-assured assumption of others' positive valuation.  Competition/Manipulation Part of being an effective leader (and indeed, sustaining narcissism) is, quite simply, some degree of ability to get others to do what you want them to. A n d requisite to the actual exercise of that ability are the beliefs that a) people can be manipulated and b) it is acceptable to do so. Raskin and Terry call this Exploitiveness, and measure it by items such as "I find it easy to manipulate people" (vs. "I don't like it when I find myself manipulating people"). But again, however, what their items perhaps miss are the dynamics that underlie the effective narcissist's interpersonal manipulation, i.e., their drive for competitive self-enhancement (e.g., Leary, 1957). Provisional items like "I'll do anything within reason to get ahead", or "I'm not above flattering people in authority" indicate a willingness to deploy manipulative, self-enhancing strategies while sentiments such as "I believe in the survival of the fittest" and "I really believe that it's a dog eat dog world" reflect the (albeit cynical) philosophy of independent achievement of one's goals that underlies the use of those strategies. A n d similarly, this independence and confidence in the ability to take care of oneself might engender the egocentric assumption that others are—or should be—equally able to do the same, i.e., "If people let you manipulate them, it's their own damn fault". A normal narcissist should, however, have at least some degree of insight into his or her own competitive manipulativeness, a recognition that might be indicated by endorsing items such as "I sometimes take advantage of other people" or " M y friends would probably describe me as competitive". They might also be expected to endorse the kind of flat-out competition-related  89  items that the NPI does not include, for example, "I can't stand losing at games" or "I really resent it when someone else does better than I do."  Vanity/Exhibition Narcissistic individuals are, according to most theoretical and descriptive treatments, highly appreciative of their own qualities (i.e., vain); when those same stellar qualities are displayed (i.e., exhibition), it is expected that comparable recognition should be forthcoming from others. Conceptual—and indeed empirical—separation of the two characteristics is somewhat difficult, however, for while the attractive, successful, good looking (and probably vain) narcissist always has an appreciative audience of at least one, it is applause from the crowd (cf. Murray, 1955) that might be expected to be especially gratifying, if not actually necessary to the service of maintaining high self-evaluation. Thus I tried to formulate items reflecting not only a high opinion of one's attributes (e.g. "I think that I'm pretty good looking", "I think that I have a pretty good body"), but also the tendency to behave in such a way (e.g., "I'm probably a bit of a flirt", "I like to dress in trendy fashions") to elicit the desired or required need for apparent admiration of one's perceived attributes (e.g., "I like it when members of the opposite sex 'check me out'" or "I like to be complimented").  "Entitlement" or Need for  Recognition/Attention  Closely related to the vanity/exhibition dynamic is what Raskin and Terry (1988) and indeed, most clinical conceptions of narcissism describe as one of its central attributes, i.e., entitlement. A number of Raskin and Terry's items reflect the usual sense of entitlement, i.e., the determined insistence to receive the special attention or treatment seen as one's due (e.g., "I will never be satisfied until I get all that I deserve" or "I expect a great deal from other people"). What they  90  don't reflect is what may be a simple need for recognition, i.e., "It's important to me to get the respect I deserve." vs. Raskin and Terry's rather more imperious "I insist upon getting the respect that is due me". Similarly, while many of us can no doubt be satisfied with a quiet pride in our own unsung achievements, there is nothing especially unusual about the need for consensual validation: The item "It bothers me when people don't appreciate my accomplishments" reflects, I think, a reasonable (albeit potentially somewhat insecure) response to not receiving such plaudits.  In addition to the expansion of the universe of content sketched above, I also generated four experimental items intended to retrospectively gauge perceptions of parental child rearing attitudes (e.g., warm and supportive vs. cool and indifferent, see Theoretical Perspectives, p. 4ff.). Results pertaining to the 'fate' of these items are described in the General  Discussion  (below).  Finally, I selected/developed items with an eye to balancing keying such that approximately half reflected self-ascription of the trait and the remainder its denial. These procedures yielded a preliminary item pool comprising 36 from the N P I - L and 67 newly generated. Certain exclusion criteria (described below) were applied, and the remaining pool reanalyzed. The items that survived the initial 'cut' (along with their P C A , also described below) are shown in Table 5.  Procedure While the current test battery (see Instruments, above) differed from that of Study 1, the Procedure remained the same.  .  91  Results Structural and Factor Analyses The first analysis of the structure of the P O N D - P was a principal components analysis of the inter-item product-moment correlation coefficient matrices. Retention criteria comprised joint consideration of eigenvalues greater than unity (and their associated scree plot), a minimum number of items (i.e., 5) loading non-trivially (i.e., >.50) on a given component (see Nunnally, 1978, p. 414, on criteria of strong structure), and the substantive considerations noted above. Because marker items from each of the Raskin and Terry (1988) component scales were included in the preliminary item pool, a seven component solution was first examined. This solution yielded 7 components that together accounted for 37.4% of total P O N D - P variance. Despite the inclusion of markers from each Raskin and Terry (1988) component scale (several of which were ear-marked for provisional P O N D - P scales divergent from their original Raskin and Terry designates; e.g., the Self-sufficiency item "I like to be complimented" was expected to mark the anticipated Need for Recognition factor—it did), the 7 component solution that emerged here bore (quite reasonably, given the additional content) little resemblance to that of Raskin and Terry. It did, however, indicate some convergence with the conceptual position described above. Some characteristics of this structure, its components, eigenvalues, variances, interpretations and sample items and their loadings are as follows: I  Assured Leadership:  (Eigenvalue [EV]= 14.36; variance'accounted for [ V A F ] = 13.9%).  Marked by such items as "I am a born leader" (.74), "I have a natural talent for influencing people" (.73) and "I'm not always sure of myself when it comes to making decisions" (R; .61), this component is redolent of confident, strong, self-assured  leadership. The Raskin and Terry (1988) Authority items load here, as do a number of their Self-Sufficiency, Entitlement and Exploitativeness items that are apparently better construed as markers of the ascendance (e.g., "I have a strong will to power") and interpersonal skills (or impression management strategies) associated with self-assured and effective leadership (e.g., "I can usually talk my way out of anything"). The experimental items "It's pretty easy for me to get other people to do what I want them to" (.59), and "People seem to instinctively trust my judgement" (.51) reflect an easy confidence in one's leadership skills as well as the assurance that others similarly recognize those skills.  II  Manipulative/Competitive  ( E V =7.41; V A F =7.2%). This component reflects a somewhat  cynical manipulativeness (e.g., "I sometimes take advantage of other people", .71; "If I have to, I'll lie to get out of a sticky situation"; .55) as well as pure competitiveness (e.g., "I can't stand losing at games"; .41 and "I'll do anything within reason to get ahead"; .41). UJ Vain Exhibition  ( E V =5.05; V A F =4.9%). Items that cluster here tap a very keen  appreciation of one's personal attributes (e.g., "I think that I'm pretty good looking"; .61; "I'm extraordinary", .56) as well as a pleasure in others' admiration (e.g., "I like to be 'checked out' by members of the opposite sex", .45; "Compliments embarrass me", R, .53). IV Ambition  ( E V =3.38; V A F =3.3%). This component appears to gauge a kind of power-  oriented ambition (e.g., "I'll never be satisfied until I get all that I deserve", .51 and "I'm not particularly ambitious", R; .48).  93  V  Parental Aloofness  ( E V =2.98; V A F =2.9%). This four-item component comprises the  experimental parental attitude items (e.g., "When I was growing up, my parents never really appreciated me"; .83; "When I was growing up, my parents were usually cool and indifferent towards me"; .72). V I Unassured Submission ( E V =2.82; V A F =2.7%). This cluster is marked primarily by reverse keyed items intended to tap exhibition (e.g., "I don't mind blending in with the crowd," R, .56) and competitiveness (e.g., "In an argument, I'll sometimes give in, even though I know I'm right", R, .55). Vn Appearance  Conscious/Need for Recognition  ( E V =2.55; V A F =2.5%). Items loading  highly here include those reflecting public self-consciousness (e.g., "I spend a lot of time on my appearance", .71 and "When I pass a mirror, I usually check myself out", .56) and an exhibitionistic need for others' positive recognition (e.g., "I like to be the center of attention, .52 and "I get upset when people don't notice how I look when I go out in public", .60). Although the exploratory analysis of the structure of the P O N D - P yielded seven (more or less-) interpretable components, other considerations dictated that a more parsimonious solution was preferable. The angle of the scree elbow, for example, indicated that a 4 to 5 component solution could adequately account for the observed variance, and there was some evidence of fractionating of substantively similar content that suggested over-extraction. Prior to considering an alternative structural solution, however, several features of the structure originally obtained suggested that the provisional item pool could be substantially reduced. First among these was the fact that 11 items failed to load at levels >-.32 on any  94  observed component; accordingly, these were eliminated. Second, because of the relatively large number of items slated to mark the provisional components/scales, I was similarly able to eliminate several evidently weak markers of the observed components; no item loading <.41 survived this initial cut. Third, a number of items were dropped owing to insufficient evidence of discriminant validity, i.e., loading at levels exceeding .39 on component(s) other than those to which they seemed most substantively related. Finally, there were a number of superfluous items: For example, the items "I am a born leader", "I would prefer to be a leader" and "I probably wouldn't make a very good leader" (R) intercorrelated at levels >.67; their near synonymity suggested that at least two could be eliminated without compromising substantive coverage. Through these and similar exclusion criteria, 37 items were excluded from further analysis.  The remaining item pool was then reanalyzed, and the matrix of interitem correlations submitted to a principal components analysis. Examination of the scree elbow indicated that a five component solution was most judicious and so was extracted and rotated to an oblique (Oblimin) solution that accounted for 37.2% of observed P O N D - P variance. Component intercorrelations range from -.001 (II, Manipulativeness and IV, Vain Exhibition) to .23 (I, Assured Leadership and V , Competitive Ambition). Although several of the components strongly resemble those that emerged from analysis of the entire 103 item pool, the elimination of the more poorly performing items produced some significant structural alterations. Selected characteristics of the solution given in Table 5 are as follows:  Insert Table 5 about here  95  I  Assured Leadership  (Eigenvalue [EV]= 10.78; variance accounted for [VAF]= 16%).  With the exception of the items eliminated via the procedures described above, this component is virtually identical to that obtained in the analysis of the 103 item P O N D - P . Again, the content here is indicative of the strong self-confidence, interpersonal skills and social presence that are characteristic of the assured and effective leader. II  ( E V =5.19; V A F =7.9%). In this analysis, the competitiveness items  Manipulativeness  drop out, rendering this component one gauging attitudinal acceptance of, and willingness to deploy self-enhancing strategies of interpersonal manipulation (e.g., "If I have to, I'll lie to get out of a sticky situation"). The egocentric assumption that it is necessary to "look out for number one"—even at the expense of numbers two and three—is the theme of this cluster (e.g., "If people let you manipulate them, it's their own damn fault"). HI Public Recognition  1  ( E V =3.99; V A F =6.1%). This component is virtually identical to  the seventh that emerged in the initial item pool analysis. Again, the items here reflect a striking public self-consciousness—primarily with respect to appearance—but also strongly imbued with a need for recognition both of one's physical stimulus value (as it were) and, to a lesser degree, one's generalized worth (e.g., "It bothers me when people don't appreciate my accomplishments"). IV Vain Exhibition ( E V =2.55; V A F =3.9%). While the third component reflects a nearinsecure need to be recognized, items here indicate both a frankly positive self-  To avoid delimiting this component's conceptual meaning to its theoretical provenance (i.e., Murray and others' notion of need for recognition), I decided to change its name to one more conceptually neutral; "public recognition" implies the same substance, and imparts additional tongue/user friendliness. 1  96  evaluation of one's attributes (e.g., "I think that I have a pretty good body") as well as an exhibitionist's pleasure in others' recognition of those same attributes (e.g., "I like it when members of the opposite sex 'check me out'"). Items indicating what some might term narcissistic self-inflation (e.g., "I'm an extraordinary person") also load here. V  Competitive Ambition ( E V =2.17; V A F =3.3%). The common content that fractionated between two components in the first analysis coalesces here into a tight cluster of recognition of one's flat-out competitiveness (e.g., " M y friends would probably describe me as competitive") and a complementary driven ambition (e.g., "It's my ambition to somehow 'leave my mark' upon the world"). Having thus obtained a solution approximating fairly clear simple structure and satisfactory  conceptual cohesiveness, I next constructed component-based scales. First, using exclusion criteria similar to those employed in the initial cut described above (e.g., absence of univocality, superfluousness, etc.), 14 additional items were eliminated. I next constructed component-based scales comprising 52 items that most reliably and discriminantly marked their appropriate components. They are presented in Table 6.  Insert Table 6 about here  I note, however, that 15 items so selected did have double loadings (range = .1 to .28 difference in magnitude, M = .171); this was deemed acceptable owing to their provisional inclusion status and the nature of the (expected) oblique relations between component-based scales. The resulting  97  scales, coefficient alphas, mean inter-item correlations, normative statistics, intercorrelations and age and gender correlations are given in Table 7.  Insert Table 7 about here  A s can be seen in Table 7, the five component-based scales show good reliability; each has acceptable internal consistency, and all of the mean inter-item correlations are well within the "optimal level of homogeneity," i.e., the .20 - .40 range suggested by Briggs and Cheek (1986, p. 115).  Table 7 also shows the intercorrelations among the component-based scales, ranging  from .04 (Assured Leadership and Manipulativeness) to .48 (Assured Leadership and Vain Exhibition). T o the extent that the subscales overlap, a Principal Components Analysis (PCA) of the interscale correlation coefficients should yield a single component interpretable as the core of P O N D - P narcissism. The analysis (details available from the author) did in fact yield a strong first component (EV= 1.98) that accounted for 4 0 % of observed P O N D variance. Similar analyses in  Studies 3 and 4 (described therein) replicated the comparable finding, but there, the  emerging first (unrotated) component accounted for additional observed variance. A s shown in Table 7, all five subscales load upon this component. It is marked most strongly by 1) Competitive Ambition and 2) Vain Exhibition; Assured Leadership loads third, followed by Public Recognition then Manipulativeness. Given that this same component emerges in  Studies 3 and 4 (as the reader shall see), and represents the common variance, it can be  98  interpreted the essence or core of POND-assessed normal narcissism. Henceforth I will use the term POND core to represent this first unrotated principal component. Age and gender correlations are, for the most part, negligible, and where they are suggestive (e.g., female subjects score below males on Competitive Ambition), the relations intimated are consistent with past empirical demonstrations (see for example, Wiggins & Broughton (1985).  Initial correlates of the  POND  Having thus constructed reliable and conceptually coherent component-based scales, I next examined their structural location(s) with reference to the interpersonal circumplex operationalized here as the IAS-R. "Projecting" an external variable (in this case, the P O N D scales and their F U P C ) onto the interpersonal space defined by the orthogonal dimensions of Dominance and Nurturance involves computation, over subjects, of the angular location and vector length of a given variable. Completing these calculations produced the angular locations shown in Figure 1 (note that these computation procedures may be found in Wiggins and Broughton (1991)).  Insert Figure 1 about here  As is appropriate for scales purporting to measure facets of narcissism, most—the exception is Public Recognition (discussed below)—fell within the theoretically relevant upper left quadrant. Considering the location of each scale in turn, Assured Leadership falls, as was anticipated, at the virtual center of the Assured Dominant (PA) sector. Assured Leadership is, of  99  course, substantively similar to several other ascendance-related scales that have been shown to mark this sector of the circumplex, for example A C L Dominance, the Hogan Personality Inventory (HPI; Hogan &Hogan, 1986) Ambition scale, I A S - R P A , the Campbell (1959) Murray Need Scale Achievement and Stern's (1970) Ego Achievement measure (Wiggins & Broughton, 1991). The aggressive hostility implicit in certain Manipulative scale items (e.g., "If people let you manipulate them, it's their own damn fault") serves to deflect the scale to the Cold-hearted (DE) sector. The hostile cynicism content is echoed in the substance of the other "life is a jungle"— type measures that have previously been shown to locate here: Christie & Geis' (1970) Machiavellianism, Wiggins' (1966) Authority Conflict (rule infraction) and Alden, Wiggins and Pincus' (1990) IIP-C Vindictiveness measure (see Wiggins and Broughton, 1991). The Public Recognition scale is orthogonal to both Dominance and Nurturance; that is, the wholly insubstantial (i.e., .02) vector length of the scale (see the  General Discussion for a  discussion of vector length) suggests that it points off the circumplex. This result seems reasonable insofar as the items constituting the scale are not really interpersonal in the dyadic sense, instead invoking a generalized other only with respect to their hypothetical response (or lack thereof) to the self as a stimulus object. Because such sentiments are only peripherally related to interpersonal interaction, the underlying substance of Public Recognition may be better elucidated by considering its relations to the other three, less interpersonally-oriented dimensions of personality structure, that is, neuroticism, conscientiousness and openness to experience; this is examined below (see also Wiggins & Pincus, 1989).  100  Although its substance yields a slight deflection toward Gregarious-Extraverted, V a i n Exhibition falls almost precisely in the middle of the P A sector; this result seems consistent with the self-confident, somewhat sexual exhibitionism that is tapped by the scale. Conceptually related scales that have also been demonstrated to locate here (see Wiggins & Broughton, 1991), such as Campbell's (1959) Exhibition and Sexuality (Murray Need) scales, as well as Bern's (1974) Masculinity measure, similarly tap the sort of confidently assertive "if you've got it, flaunt it" mentality that is explicit in Vain Exhibition.  Finally, the Competitive Ambition scale—comprising mainly purely competitive-type items (e.g., " M y friends would probably describe me as competitive")—falls almost precisely where it should, that is, right in the mid-range of the Arrogant-Calculating (BC) sector. The slight displacement of the scale toward P A is owing to the ambitious power striving evoked by other scale items (e.g., "It's my ambition to somehow 'leave my mark' upon the world") that themselves are similarly characteristic of assured dominance. The location of the scale reflects its relative purity as a measure of the competitive narcissism described by Leary (1957).  Lanyon's (1973)  Social Non-Conformity scale and Hogan and Hogan's (1986) Ambition scale also share this approximate structural location.  The 'other' structural model of personality is, of course, the familiar B i g Five. While the B i g Five's correlations with the individual scales are described below, how the P O N D ' s core relates to these firmly empirically-grounded and well-understood dimensions of personality is important. A s the reader may recall, the dearth of literature on normal narcissism made specific predictions difficult to advance. Accordingly, I made no predictions with respect to Openness or  101  Conscientiousness. There was no relation with the former, and the latter correlated with the P O N D core at a relatively low (i.e, .21) level. Perhaps (as I speculated before) the narcissist is somewhat conscientious with respect to the achievement of certain personal goals (e.g., as regards 'successful' competition). Similarly, the interpersonal coldness attendant to certain P O N D items may give rise to the r=-.25 between the P O N D core and (dis)Agreeableness. The P O N D core, however, is negatively related to F F I Neuroticism (i.e., r=-.25). (Note that this information is not evident in the circumplex; it requires the more extensive B i g Five structure.) A n d finally, the correlation of .40 between Extraversion and the P O N D core (and its convergent correlation of .37 with Cheek Shyness) suggests the outgoing social presence thought to be characteristic of the narcissist.  Additional  Correlates  Having determined that the P O N D scales locate in the conceptually appropriate interpersonal space, I next examined their relations to a variety of theoretically-relevant external criteria. Zero order correlations between the full scale P O N D , its subscales and selected external variables are shown in Table 8.  Insert Table 8 about here  A s can be seen in Table 8, the full scale P O N D , the P O N D core and the measure's five component-based scales show meaningful and coherent patterns of relations with the external variables. First, in line with the projections on the circumplex, the full scale P O N D and all but  102  one of its component scales show at least weak to moderate relations to the I A S - R ArrogantCalculating (BC) scale. That Public Recognition should be all but uncorrelated with B C is to be expected, given the "non-interpersonal" nature of its content (discussed above). On the subscale level, the highest B C correlations are, not too surprisingly, with Manipulativeness and Competitive Ambition, scales that tap the cynically competitive B C interpersonal style; these scales' moderate negative relations toFFI Agreeableness similarly underscore the somewhat offensive interpersonal behaviors admitted by high P O N D scorers.  It is of some interest that among the other correlates of the P O N D (and three of its five component scales) is I A S - R Assured-Dominant (PA): While individuals who score high on the P O N D endorse items reflecting the aggressively competitive style characteristic of the ArrogantCalculating type, they appear almost equally at ease and presumably skilled in the deployment of more conventional means of exercising power in a social context. That persons who describe themselves as assertive, decisive and self-confident (PA) should similarly score highly on Assured Leadership, Vain Exhibition and even Competitive Ambition provides convergent evidence for the validity of these scales as markers of the strong, self-assured and interpersonally powerful style that I've argued is characteristic of the normal narcissist.  Assured Leadership. This pattern of self-assured social presence is seen most strikingly with respect to the correlates of Assured Leadership, including (as noted above) positive relations with I A S - R Assured Dominant (PA), I A S - R Gregarious Extraverted (NO), and F F I Extraversion; negative correlates are consistent and complementary, for example, I A S - R Unassured Submissive (HI), Aloof-Introverted (FG), Cheek (1983) Shyness, and IDI Dependency.  103  Wiggins and Broughton (1991) have noted that dominance-related constructs structurally locate in the P A sector—and as I have shown is true of Assured Leadership—typically reflect healthy or socially desirable aspects of interpersonal behavior; the moderate to strong negative correlation between the scale and F F I Neuroticism reinforces this interpretation of the positive and adjustive nature of Assured Leadership, as do the correlations between the scale and IDI Dependency—most strikingly with respect to the Unassured scale (r = -.71). Similarly suggestive of the effectiveness of the Assured Leader's interpersonal style is the lack of positive relations between it and the relevant scales of the IIP-C. While some might interpret this result as evidence for a perhaps overly vigorous denial of common interpersonal problems, one could just as easily argue that being socially skilled, extraverted and self-confident almost guarantees that one won't have too much difficulty getting along with others.  Manipulativeness.  The pattern of correlates of Manipulativeness, on the other hand, seems to  suggest a somewhat less socially desirable interpersonal style; the highest positive correlates, for example, being I A S - R Cold-hearted (DE) and Arrogant-Calculating (BC). Manipulativeness is comparably negatively related to F F I Agreeableness, reflecting the rather aggressive and cynical content in this scale. Self-reported interpersonal problems are consistent with this picture, with high Manipulativeness scorers reporting themselves as both vindictive (IIP-C B C , e.g., "It is hard for me to trust other people) and domineering (PA: e.g., "I try to control other people too much").  Public Recognition.  O f the five P O N D scales, that tapping Public Recognition shows the  most threadbare (as it were) nomological net. A s might be expected from the earlier structural location analysis, it is literally uncorrelated with any of the interpersonal variables. With the exception of the Unassured scale, the predicted positive relations between Public Recognition  104  and the IDI scales do obtain, albeit at a much lower level than expected. The few remaining correlations, while reliable, are also weak, averaging for the most part in the low to mid-twenties. At the risk of overinterpreting these few, perhaps merely suggestive relations, the scale does seem conceptually orthogonal to, for example, the self-assuredness that characterizes Assured Leadership, correlating only with variables that tap a variety of other-directedness (e.g., IDI Emotional Reliance on others, F F I Neuroticism, and IIP-C Intrusiveness, e.g., "I want to be noticed too much").  Vain Exhibition.  Inspection of the correlates of Vain Exhibition suggest some degree of  redundancy with those of Assured Leadership; in most cases, the pattern is virtually identical (e.g., I A S - R P A ) , albeit weaker. This result is perhaps not especially surprising given the substantive commonality between the scales, that is, elements of extraverted exhibitionism, and self-confidence in, and positive evaluation of one's attributes. However, given that this fourth scale shows an absence of differential relation to the external variables assessed here, its potential for incremental validity (vis-a-vis Assured Leadership) might appear limited. However, given the somewhat circumscribed range of external correlates considered in the present study, I decided to reserve judgement on its eventual inclusion pending evaluation of the scale's differential relation to an expanded spectrum of external variables (Studies 3 and 4, below).  Competitive-Ambition.  A s might be anticipated, the highest interpersonal correlate of  Competitive Ambition is I A S - R Arrogant-Calculating (BC), a scale that taps the kind of cynical exploitiveness considered by Leary (1957) as most diagnostic of the competitive narcissist. Probably owing to its ambition-oriented items, this scale is also moderately correlated with the I A S - R Assured-Dominant (PA) scale. A n d although four of the five P O N D scales relate  105  similarly, Competitive Ambition shows the highest negative correlation with I A S - R Unassuming-Ingenuous (JK), a scale tapping the "disposition to be deferent, obliging, modest and straightforward in social transactions" (Wiggins, 1992, p. 10), characteristics that are, of course, the antithesis of Competitive Ambition. Also of note is the moderate relation between this scale and Domineering problems, possibly suggesting that the intense drive of the Competitively Ambitious person to "get ahead" concurrently precludes their "getting along" (see also Hogan, 1983). High scorers on this scale also have a tendency to report Vindictiveness problems; perhaps if one is less than straightforward in one's own interpersonal behavior, distrust and suspicion of others' motives is a logical concomitant.  Discussion Taken together, these initial correlational results indicate promising—albeit occasionally circumscribed—evidence for the validity of the P O N D , its five component scales, and in particular, the substantive core. In general, the core shows similar correlates to the total P O N D score. Where they differ, however, the P O N D core should take precedence because it is unadulterated by content specific to one or two of the subscales. Overall, the pattern of correlations with the core indicate that the normal narcissist is dominant, manipulative, competitive, vain, and desires public recognition. Further, the normal narcissist is generally secure, but with a pattern of interpersonal problems similar to that of disagreeable, dominant individuals. Considering each subscale in turn, it is clear first, that the pattern and strength of the correlates of Assured Leadership provide strong convergent (e.g., Assured-Dominant,  106  Gregarious-Extraverted, F F I Extraversion) and discriminant (e.g., I A S - R Aloof-Introverted and Unassured-Submissive, Cheek Shyness, IDI Unassured and Dependency) evidence for its validity as an indicant of strong, self-assured ascendance/leadership and social self-confidence. Perhaps owing to the more limited range of relevant constructs considered, the pattern of correlates of Manipulativeness is less extensive. Correlations with substantively similar variables are, however, appropriate and consistent: Coldness toward, and rejection of others, for example, is seen in the positive relation to I A S - R Cold-hearted and the discriminant correlations between Manipulativeness and F F I Agreeableness, and I A S - R L M and Warm-Agreeable. Similarly, the convergence of Manipulativeness and I A S - R Arrogant-Calculating suggests its validity as a measure of the egotistical and self-aggrandizing manipulation of others that is intrinsic to that interpersonal style. The virtual lack of relations between Public Recognition and the other variables considered— while not wholly anticipated—is, upon reflection, not especially surprising. Consider the content of this scale, that is, primarily appearance-consciousness, and the need for attention to and/or recognition of one's stimulus value (as it were): i.e., wanting others to notice one's appearance, accomplishments, etc. Such content is, of course, evocative of a conceptually similar construct, that is, Fenigstein, Schier and Buss' (1975) description of Public Self-Consciousness, or the tendency of some individuals to be "especially cognizant of how they are being viewed by others in their social contexts and how those others are reacting to them" (Carver & Scheier, 1981, p. 193). Public Recognition items such as "I spend a lot of time on my appearance" or "It bothers me when people don't appreciate my accomplishments", seem, like the Fenigstein et al. measure, to similarly tap the "tendency to be aware of the publicly displayed aspects of the self, the self as  107  a social object that has an impact on others" (p. 193). A n d given this apparent awareness of the self as an object having impact on, but not in overt interaction with others, it seems consistent that the Fenigstein et al. (1975) Public Self-consciousness scale is, like Public Recognition, located  outside interpersonal space, or in other words, is orthogonal to the eight dimensions of  interpersonal behavior defined by the IAS (Wiggins & Broughton, 1985). This suggestion that Public Recognition may be substantively similar to the evaluation concerns of the Publicly Self-conscious individual is further bolstered by the few noninterpersonal variables that were reliably related to the scale, namely IDI Dependency and Emotional Reliance on Others. The small (though reliable) correlations between Public Recognition and the F F I and I A S - R B5 Neuroticism scales might similarly intimate a degree of social anxiety attendant to such evaluation concerns. The hypothesized relations between Public Recognition and the Fenigstein et al. Public and Private Self-consciousness (and Social Anxiety, 1975) scales are directly tested in the next study I report. A s noted earlier, the correlational pattern of Vain Exhibition showed substantial overlap with that of Assured Leadership, particularly with respect to dominance and social self-confidence related variables (e.g., I A S - R Assured-dominant, IDI Dependency (-), Cheek Shyness (-), etc.). In part, these findings are a function of double loadings (on Assured Leadership) of the items assessing confident self-evaluation (i.e., "I think that I'm pretty good looking, ...a special person, ...extraordinary person, ...a bit of a flirt, etc.), items that one might reasonably expect to be similarly characteristic of the self-assurance of high scorers on Assured Leadership. Other items on the scale, however, are more evocative of purely physical vanity, attention seeking and sexual  108  exhibitionism, yet are nonetheless similarly directionally related to the self-confidence (e.g. r= .40 with IDI Unassured) that marks Assured Leadership. This other "cluster" (if you will) comprises items that do not, in contrast, load additionally on the Assured Leadership factor. Instead, they seem to gauge two separate—and perhaps complementary—characteristics: a) an awareness of the self as a physical or sexual object (e.g., "I think that I have a pretty good body") and b) a gratification  from others' appreciation of the  self as a physical or sexual object (e.g., "I like it when members of the opposite sex 'check me out'"). In other words, while the first group of items reflects primarily physical se/f-evaluation, the second seems rather to gauge a self-assurance with respect to receiving others' positive evaluation or admiration of the physical self. I expect that external variables more directly assessing these characteristics (e.g., Exhibitionism, Sexuality) would produce relations more clearly differentiated from those of Assured Leadership. Finally, the pattern of intercorrelations obtained for Competitive Ambition is appropriately convergent (e.g., I A S - R Arrogant-Calculating; B C ) and discriminant (e.g., not I A S - R Unassuming-Ingenuous; JK). Again, while there is some degree of correlational overlap between this scale and Manipulativeness (e.g., with I A S - R B C and F F I Agreeableness), this result is reasonable given the substantive commonality between the two, namely an egotistical cynicism that pervades interpersonal behaviour. What appears to differentiate Manipulativeness from Competitive Ambition, however, is the latter's moderate relation to I A S - R Assured-Dominant (PA), a measure that reliably indexes an independent and self-assured (rather than merely aggressive) attitude toward achieving influence over others in a social context.  In this Discussion, I have devoted substantial time to considering the detailed correlates of each of the subscales; I believe that their contribution to a true profile of narcissistic dispositionshas been substantiated. This being said, the correlates of the P O N D  core may be  more important as the core represents the common variance, that is, substance that the subscales share. Therefore, for the subsequent studies, I will focus more on the core and less on the subscales.  110  STUDY 3: Extending the nomological net Given the factor analytic algorithm that one should have approximately ten times as many subjects as one has variables (e.g. Nunnally, 1978), a valid test of the structural stability (and scale reliability) of the P O N D and its external correlates requires crossvalidation on a fairly large sample. Because the P O N D has 52 items, a sample size of 583 was deemed suitable for this purpose. In addition to attempted replication of the pattern of POND-personality correlates observed in Study 2 (above), I also expanded upon and tried to provide further validation for the nomological net of the construct assessed by the P O N D and its scales. I attempted to illuminate the 'normality' or pathology issue by examining the relations between the P O N D and Weinberger's (1989) Adjustment Inventory (WAI), a measure that assesses anxiety, depression, repression and denial (among other variables), each of which might be expected to have predictable relations to what is assessed by the P O N D . Self-reported interpersonal problems were again measured by the IIP-C (see Study 2). Narcissists' evident concern for attention, admiration and their alleged insecurity was assessed with instruments such as the Public Self-Consciousness scale and a measure of Outer directedness. Finally, Block's newest (1989) measure of his Ego Resiliency construct was also included in the battery.  Based upon the findings of Study 2, some predictions were advanced. First, with respect to the adjustment-type measure (i.e., the WAI), I anticipate negative correlations with the P O N D core and several of the subscales, particularly those saturated with dominance-related content (i.e., Assured Leadership and Vain Exhibition). Given  Ill Emmons' (1984, 1987) and Gurtman, Fernandez, and Phillip's (1991) findings relating exploitativeness to maladjustment, it is likely that manipulative content is related to certain aspects of maladjustment (in that high Manipulativeness scorers indicate agreement with cynical or interpersonally aggressive items (e.g., "If people let you manipulate them, it's their own damn fault."). Similarly, Manipulativeness should also be negatively-related to Agreeableness. Desire and need for public recognition might be correlated with social anxiety. A n d as in Study 2,1 expect that relations to the IIP-C would obtain in this sample as well (e.g., Domineering problems correlating with surgent content, Vindictiveness with manipulative content, Intrusiveness with exhibition, etc.). I A S - R relations reported in the previous study should, of course, replicate. In terms of the other measures, I predict that the P O N D core (as well as Assured Leadership and Vain Exhibition) will be moderately related to Ego Resilience and W A I Well-being and Self-esteem scales because the latter tap general adjustment. The W A I Denial scale might be similarly correlated, given the oft-replicated association between measures of desirable responding and self-report indexes of psychological 'health' (e.g., Paulhus, 1991).  .  ;  I anticipate that the Public Recognition scale will be positively related to the Public Self-Consciousness scale; the excessive self-awareness associated with that disposition might also be picked up by the social anxiety measures in the P P S C and W A I . The concept of rumination is also substantively similar, and thus the R R Q Rumination should be related to the need for public recognition.  112  Method Subjects 583 lower-level undergraduates (201 men and 382 women) participated in exchange for course credit. While their ages ranged from 18 to 44 years (M=21.5, S D 3.3), by far, the lion's share were in their early twenties.  Instruments Aspects of Identity ( A l ; Cheek, 1989). The A l is a 17-item index measuring aspects of social and personal identity, i.e., inner and outer orientations. The measure has acceptable reliability. (I note that for brevity's sake, complete results are not described below, but are, of course, available from the author).  Ego Resilience (ER; Block, 1989). Block's most recent measure of his Ego Resilience construct (generally corresponding to good interpersonal function and adjustment) is a 14 item measure upon which respondents rate the self-applicability of statements reflecting openness, surgency, resilience, etc. Sample items include "I quickly get over and recover from being startled" and "I usually succeed in making a favorable impression on people". The E R was embedded within the P O N D and each item was rated on a six-place Likerttype scale, ranging from (1) "strongly agree" to (6) "strongly disagree". Reliability of the measure is acceptable.  FFI See Instruments, Study 2. IAS-R See Instruments, Study 2. Please note that the B i g Five (B5) version of the I A S - R was used in some studies; in the interests of economy in reporting, complete I A S - R B 5  113  results are not described; while full results are available from the author, certain B 5 relations are noted below. Complete results pertaining to another reliable B5 measure (i.e., the FFI; see above) are reported; in any case, I A S - R B5 relations were consistent with those obtained with the FFI. Profile of Narcissistic  Dispositions ( P O N D ; Taylor, 1991) O n the basis of the itemmetric  data gathered in Study 2, the final version of the P O N D comprises 52 self-descriptive statements that are rated on a Likert-type scale ranging from (1) "strongly agree" through (6) "strongly disagree". Coefficients Alpha for the scales range from .77 to .88; the full scale value exceeds .89. Public and Private Self-consciousness  (PPSC; Fenigstein, Schier & Buss, 1975). The  P P S C is a 23-item measure of two minimally related (i.e., r=.23) modes of selfawareness: Public, or outer-directed self-consciousness and Private, or inner-directed selfconsciousness. The first taps "an awareness and concern for the self as a social stimulus", the second a "cognitive, private mulling over of the self" (Fenigstein et al, 1975, p. 524). A third scale, Social Anxiety, is also included. Items are rated on a 5 place Likert-type scale ranging from 0 (extremely uncharacteristic) to 4 (extremely characteristic). The scales are internally consistent and show good temporal stability. Sample items include "I'm always trying to figure myself out" (Private), "I'm self-conscious about the way I look" (Public), and "I get embarrassed very easily" (Social Anxiety).  Weinberger Adjustment Inventory (WAI; Weinberger, 1989). The W A I is a 85-item measure of generalized psychological distress. In addition to ten facet scale scores (e.g., Anxiety, Depression, Self-esteem, Impulse Control, Denial, etc.), two global component  114  scores (Distress and Restraint) may be calculated. The measure and its component scales have acceptable levels of reliability.  Procedure With the exception of the additions to the battery noted above, the procedure was identical to those of Studies 1 and 2.  Results and Discussion The first analysis was a Principal Components Analysis (PCA) of the 52 x 52 interitem product-moment correlation coefficient matrix; five components were extracted and rotated to an oblique (oblimin) solution that accounted for 43.1% of the observed variance. While the observed structure showed some slight degree of variation from that identified in Study 2), the original 5 P C solution appears to be quite stable. The loadings are presented in Table 9.  Insert Table 9 about here  A s expected with a large replication sample, the order of emergence is the same as that observed in Study 2, i.e., I  Assured Leadership.  (Eigenvalue [EV]= 9.95; Variance Accounted For [VAF]=  19.1%). After two replications, I am confident describing this component as a good reflection of self-ascribed assured, effective leadership-relevant qualities (e.g., "I have a natural talent for influencing people"; loading =.79). 12/12 of the original items load  115  here, in almost the exact same order as before. Three additional items, however, do indeed drift to this factor, two from Vain Exhibition ("I'm an extraordinary person"; .60; and "I think that I'm a special person", .59) and one Competitive Ambition item ("It's my ambition to somehow leave my mark upon the world"; .46). This item drift is not terribly surprising, given the 'positive manifold' nature of this factor's evident substance, i.e., the drive toward ascendance/achievement, strong self-confidence, and robust (albeit perhaps over-inflated) self-esteem.  II  ( E V = 4.85, V A F = 9.3%), This is clearly a reliable index of  Manipulativeness.  approval and admission of tendencies toward interpersonal manipulation ("I sometimes take advantage of other people"; .71). This component retains 9/9 Manipulativeness items but also picks up 4 Competitive Ambition items (e.g., "I'll do anything within reason to get ahead"; .52; and "I usually accept defeat gracefully" (R); .49). I do not see this drift as especially problematic given these items' substantive consonance with the conceptual core of Manipulativeness, i.e., the evident need to 'achieve at any cost' intrinsic to the 'getting ahead' rather than 'getting along' orientation (e.g., Hogan, 1986) gauged by the component.  HI Public Recognition.  ( E V = 3.34; V A F = 6.4%). Public self-consciousness and need  for others' favorable acknowledgment (e.g., "I get upset when people don't notice how I look when I go out in public"; .66) are tapped directly by this component. These data provide, once again, a straight replication. rv" Vain Exhibition.  ( E V = 2.2; V A F = 4.2%). Once again, this component seems to  provide a good gauge of the tendency to perceive oneself as a physically pleasing  116  stimulus whose approval is both expected and not unwelcome (e.g., "I get uncomfortable when members of the opposite sex 'check me out'" ; (R); .70). In this analysis, the component retains 9/11 items; 2 grandiosity-type items drift to Assured Leadership (see P C I, above), but still retain significant loadings on this component. V  Competitive Ambition. ( E V =2.10; V A F =4.0%). This component shows some degree of instability as 5/11 items drift (as noted above). Its fractionation—while somewhat problematic in psychometric terms (I note however, that items originally designated here retain reliably strong albeit lesser loadings than on the components they drift to)—is, again, less so, substantively, insofar as Competitive Ambition seems to have two facets: 'Flat out' ambition/ascendance desires (e.g., "It's my ambition to somehow 'leave my mark' upon the world", loading on Competitive Ambition =.35; Assured Leadership = .46) as well as a lack of principles that implies rejection of socially conventional/approved means in favor of 'achieve at any cost'-type power strivings (e.g., "I'll do anything within reason to get ahead" drifting to Manipulativeness, loading = .52).  While I shall later describe correlational evidence for the partial substantive differentiation of Competitive Ambition from Manipulativeness, I should like to remark here upon two arguments for the scales' (albeit possibly altered but ultimate) retention. The first has to do with the method of rotation: Although I selected an oblique rotational method—reasoning that 'allowing' components/items to be correlated might more accurately reflect the 'true' state of narcissistic nature—a concomitant of this choice is that the resultant psychometric solution is somewhat less clear insofar as itemmetric vagaries  117  are somewhat highlighted by this methodology. Consider, for example, the item "I usually accept defeat gracefully" (R): While the oblique rotation renders this item more properly attributable to the Manipulativeness component (i.e., a loading of .49 vs. .32 for Competitive Ambition), Varimax rotation leads to a loading of .32 on that component vs. the .42 more substantively consistent with the rejection of interpersonal defeat intrinsic to Competitive Ambition (and obtained in this analysis). Similarly, in other cases, as in the item "I can't stand losing at games" (Oblique loading = .50 for Manipulativeness, .33 for Competitive Ambition), a Varimax rotation renders the difference between componential loadings a mere .01 in correlation terms. While these data could suggest that a certain proportion of the Competitive Ambition items are not ideally discriminantly valid, they're equally suggestive of a substantive commonality that is difficult, if not impossible, to tease apart psychometrically.  The second argument invokes a simple fact, i.e., that the component-based scales (with their original item complenents) remain quite reliable. In a subsequent section (below), I discuss the correlational evidence bearing upon not only the replication of Competitive Ambition's nomological net but also its differentiation from other substantively-related attributes.  I next reconstructed the P O N D scales: scale means, standard deviations, scale intercorrelations and alpha reliabilities are given in Table 10.  Insert Table 10 about here  118  Inspecting this table reveals that the appropriate values are clearly replicated, providing evidence that yields confidence in the stability of the P O N D and its component-based scales. A n d as in Study 2,1 performed a P C A of the 5X5 matrix of inter-scale correlation coefficients. A strong first (unrotated) component, upon which all variables loaded highly, once again emerged, this time accounting for 44.8% of observed P O N D variance. This result is also shown in Table 10.  Self-Report  Correlates  The self-report correlates are provided in Table 11. A s in Study 2, the P O N D core  Insert Table 11 about here  once again correlates with I A S - R Assured Dominant (PA) and but more strongly with Arrogant Calculating ( B C ; r=.55), or, in Leary's (1957) original I C L terms, Competitively Narcissistic. A n d recall that in contrast to other indices, the P O N D is meant to assess 'normal', non-pathological, non-neurotic narcissism. This it evidently does, insofar as the F U P C is wncorrelated with the Basic Symptom Inventory—a reliable index of somatic and psychological distress (additional details available form the author). A n d it is negatively related to indices of neuroticism, the I A S - R B5 N scale, and F F I Neuroticism (as of course are the relevant individual scales).  119  A s for the WAI—again a reliable measure of generalized psychological distress—there appear to be some interesting complementary relations. For example, while one sees a positive relation between the P O N D core and W A I Aggressiveness (consistent, of course, with its negative correlation with F F I Agreeableness i.e., r=-.45), this interpersonally offensive sequela of competitive narcissism seems somewhat 'balanced' by the P O N D core's negative correlations with Anxiety (converging with the P U S C Social Anxiety scale), Depression, and Global Distress (the latter correlation not being shown). A n d irrespective of the P O N D narcissist's evident 'non-niceness', one still sees a robust relation between their scores and high self-esteem (i.e., r=-.43, P O N D F U P C and Weinberger L o w Self-Esteem). Thus having replicated the structure and reliability of the P O N D and its derivative scales, and having replicated the P O N D core's correlates, I next sought to a) Replicate the scales' relations to the theoretically-relevant criteria identified in my previous study; and b) Expand upon and further clarify the nature of constructs tapped by the individual P O N D scales (most notably, Public Recognition and Vanity). Examination of Table 11 reveals that once more, both the full scale P O N D and its five component-based scales demonstrate a pattern of relation to external variables that is both meaningful and coherent—and is not serendipitous, having similarly emerged in the (multiple study)  1  programme of research described herein.  In Study 2, I demonstrated that the POND, its core and most component scales project onto conceptually appropriate interpersonal space. Examining their correlations with reliable markers of this space (e.g., IAS-R scales) was felt to suffice as adequate replication of this result. 1  120  Again, as should theoretically obtain for a measure gauging (normal or otherwise) narcissism, the P O N D subscales and their core fall in the upper-left quadrant of the circumplex. While each relates to I A S - R Arrogant-Calculating scale (i.e., P O N D = .60; M scale r = .40), once again, the highest subscale correlations are with Manipulativeness and Competitive Ambition, underscoring the reliability of these P O N D scales' assessment of the cynical, virtually scruple-less motive to set oneself apart from the interpersonal pack by, it seems, whatever means are necessary. Similarly, these scales' and the P O N D ' s core moderate (negative) and reliable relations to I A S - R L M (i.e., Warm-Agreeable) and FFI Agreeableness suggest that high scorers in these realms are probably not too terribly concerned about whose interpersonal/affective toes are stepped upon. Additionally, a significant interpersonal correlate of the P O N D , its core and (all but one of) its subscales is I A S - R Assured Dominant (PA). Although high P O N D scorers (in repeated samplings) self-ascribe items that index the aggressive competitiveness associated with the Arrogant-Calculating (BC) type, they also endorse those tapping ease and (presumable) skill in more conventional means of exercising and achieving power in a social context (as we shall see, there are moderate relations between high Assured Leadership scores and peer ratings of variables related to those "conventional means", specifically, verbal skills and social competence). Once again, the self-ascription of items gauging assertion, decisiveness and self-confidence (PA) is similarly related to high scores on Assured Leadership, Competitive Ambition and Vain Exhibition, providing converging evidence for the validity of these scales as indicants of the ascendant, self-  121  assured interpersonally powerful social demeanor that I think is characteristic of the normal narcissist.  Assured Leadership The self-confident social presence so striking in the emergent portrait of the P O N D narcissist is especially illuminated by considering the correlates of Assured Leadership ( A L ) . I noted earlier how Wiggins and Broughton (1991) showed how a dominancerelated construct (as Assured Leadership surely is) typically represents more healthy or socially desirable aspects of interpersonal behavior: The substantively consonant correlates of Assured Leadership ( A L ) consistently typify this. Correlations with the Weinberger (1989) Anxiety Inventory (WAI) Anxiety and Depression scales are uniformly negative (as are those with the F U P C ) , indicating either (depending upon one's interpretation) an absence of the distressing variable, or some low to moderate level of the implied "positive" pole of the symptom. Additionally illustrative of the adjustive construct tapped by P O N D A L are other correlations with the W A I , commonly construed as a marker of psychosocial health. A n d such health seems strongly indicative of the state reported by high A L scorers: A s well as W A I profiles characterized by the absence of anxiety and depression, high A L scorers also report an absence of psychological Distress (i.e., -.61; results available from author). A n d as I found in my previous research, there is a moderate relation (r=.45) between Denial (of psychosocial distress) and A L ; this is perhaps not an especially surprising result, irrespective of one's view upon the substance of denial (i.e., denial of common distressors and/or socially desirable impression management or actual veracity in  122  reporting the absence of psychological distress). In any case, I prefer to view this particular glass as half full; those who would argue that it's half empty might do well to consider the moderate—and negative—relations between high A L scorers and two separate indices of Neuroticism (i.e., I A S - R B 5 and FFI) as well as the moderate positive (i.e., r=.51) relation between A L and W A I Wellbeing. Finally, note the correlation between E g o Resilience—an index of effective interpersonal functioning and adjustment—and high A L scores, i.e., a not insubstantial .51.  In general, the self-reported characteristics of the Assured Leader are those that bespeak a non-neurotic, strong level of psychological health. A s the values in Table 11 indicate, the high A L scorer endorses items (on the P O N D ) and a host of other reliable measures reflecting an effective dominance, strong social presence, extraversion, and remarkably healthy self-confidence/esteem.  Manipulativeness The correlates of Manipulativeness, on the other hand, suggest a far less socially desirable—if not downright disagreeable—interpersonal style. Consider the correlation with F F I Agreeableness, for example: High Manipulativeness (M) scorers tend to be anything but agreeable; insofar as r= -.65 between that scale and Agreeableness [or, in this case, Antagonism, as some (e.g., Costa and McCrae, 1987) would have it]. A correlation of -.49 with I A S - R L M (Warm Nurturant) similarly underscores the interpersonal 'coolness' of high M scorers. The strong correlation (r=.60) between M scores and I A S - R Arrogant-Calculating (BC) levels is also theoretically conceptually  123  consistent with ray notion of the self-ascribed interpersonal sequellae o f — i n this c a s e — the manipulativeness intrinsic to narcissism. Despite this portrait of the high M scorer as (to say the least) interpersonally insensitive, other data suggest that he/she probably has no overt symptomatology. Insofar as the BSI is a reliable index of psychological distress, the correlation between the full scale BSI (see note, above) and M — a mere (albeit significant) .18—suggests that high scorers on this measure might not be especially nice to others, but neither are they particularly discomforted by it. The absence of correlation between M scores and two separatee measures of Neuroticism (i.e., I A S - R B 5 ; see note, above; and the FFI) contributes to what I view as a solid basis for asserting that M taps a non-abnormal, nonpsychiatric, non-neurotic—albeit non-warm and fuzzy—aspect of narcissism. Ruling out overt symptomatology then, one sees that high M scorers are not, evidently, especially burdened by psychosocial distress. Actually, they may deny psychosocial stress, but that is not to say that may not be instrumental in causing it in others: Witness, for example, the moderate positive relations between the self-ascription of Manipulativeness and W A I Aggression, Irresponsibility and Impulsiveness, as well as the moderate negative relations that obtain between high M scores and variables such as W A I Considerateness (I note that the global W A I Social Restraint value correlates -.59 with high M scores). A n d while highly Manipulative persons may, one infers, cause considerable psychosocial distress in others, my data indicate that they are relatively "distress-free" themselves: There are no, or at best, only negligible correlations between  • •  124  M and W A I Depression and Anxiety, and the M - G l o b a l W A I Distress score correlation is significant but weak.  Public  Recognotion  Those scoring high on Public Recognition (PR), appear, on the other hand, to be slightly more distressed by their inferred impact upon others. While—as in my previous study—high P R scores appear to have virtually no interpersonal profile (as indexed by, e.g., the I A S - R ) they are, evidently, somewhat discomforted by the nature of their social 2  interaction. While not strikingly so, high PR scores are reliably (albeit weakly) associated with F F I Neuroticism. The W A I scales said to measure psychosocial 'health' (or lack thereof) are similarly related, correlating at fairly low but reliable levels with high PR scores; for example, high PR scorers tend to report some Anxiety, Depression, low Selfesteem and (global) Distress (I note that these variables correlate moderately to strongly with Assured Leadership and Vain Exhibition—but in the opposite direction); Aggression and Irresponsibility are positively and significantly related to high P R scores while Repression and Denial are, interestingly, negatively related to P R scores. I say "interestingly" as conventional psychological wisdom has these dimensions as positively related to good psychosocial functioning.  More striking, however, is how PR's nomological net has become distinctly less threadbare than it was in my previous study. A s suggested earlier, however, the intricacy  As I described earlier, Public Recognition appears to fall "outside" of interpersonal space, invoking as it does a generalized other who exists only in terms of his or her hypothetical response/reaction to the self as stimulus object.  2  125  of the mesh is no doubt increased by expanding the conceptual relevance of the theoretical constructs examined. Earlier I described the content of P R as "primarily appearance-consciousness, and the need for attention to and/or recognition of one's stimulus value ... wanting others to notice one's appearance, accomplishments, etc." (e.g., "It bothers me when people don't appreciate my accomplishments"). I also noted its conceptual commonality with Fenigstein et al.s' (1975) notion of Public Self-Consciousness, or the tendency of some individuals to be "especially cognizant of how they are being viewed by others in their social contexts and how those others are reacting to them" (Schier and Carver, 1981, p. 193). A s I had predicted, Public Self-Consciousness does indeed appear to be the conceptual cousin to PR, correlating a strong .57 with Public Recognition. Consistent with this is the r= .46 between PR and Cheek's (1989) A l Outer Orientation index. Fenigstein et al's (1975) Private Self-Consciousness scale -tapping the tendency to introspect about one's self- also correlates at a fairly low (.18) but reliable level with PR. At first glance, this result appears somewhat surprising given the social evaluation apprehension evidently tapped by PR: It makes further sense, however, when one considers the (albeit modest) correlation between P R and the Rumination scale of Trapnell's (1991) Reflection and Rumination Questionnaire (i.e., r=.32; details available from author). The close-to-obsessive nature of that scale (tapping a psychological/mental re-evaluation of how one's behaved) is probably conceptually (and behaviorally) akin to the kind of social/physical evaluation apprehension that seems characteristic of the high  126  PR scorer. The suggestion that a narcissist might be somewhat concerned b y — a n d highly aware of (viz., the strong P U S C r)—his or her social and physical impact is perfectly consonant with the present thesis, namely that part of being a 'normal' narcissist is the concern and/or awareness as to whether one's social and physical being will have some impact upon, or at least be positively evaluated by, others.  Vain  Exhibition  The V a i n Exhibition ( V E ) scale, once again, has a pattern of correlation virtually identical to (albeit typically at a more moderate level than) that evinced by Assured Leadership, i.e., one indicating a dominant, assured, strong and skilled social presence. A n d as with high A L scorers, individuals endorsing V E items show high self-confidence and healthy self-esteem (as indexed by two separate measures). Also as with A L , high V E scorers similarly indicate a freedom from psychosocial distress (as indexed by negative FFI Neuroticism and W A I relations) and exhibit a comparable interpersonal profile. In part, Vain Exhibition's strong overlap with the correlational pattern of A L — especially as regards dominance and social self-confidence—related constructs—is due to the shared conceptual similarity (and double loadings) of the related constructs assessed by the two scales, i.e., confident self-evaluation (e.g., "Ithink that I'm pretty good looking" or "I'm not always sure of myself when it comes to making decisions" (R)). The shared variance evident—both psychometrically and conceptually—has, once again, been underscored in this study. Despite my appeal to a "positive manifold" - type explanation, it appears that I have been unable to fully tease out the discriminative variance. At least on the scale level: Their shared variance is evident in their strong loadings on the  127  F U P C / P O N D core (and their similar correlates). But while it's evident that Assured Leadership and Vain Exhibition share common conceptual substance, there is evidently some additional variance that they do not share. I note, for example, that the Vain Exhibition items evidently gauging gratification from self-exhibition (e.g. "Compliments embarrass me"(R)) do not load upon the Assured Leadership component.  3  Competitive Ambition Finally, the intercorrelations between Competitive Ambition (CA) and the external variables assessed are, for the most part, a replication of the previous study. Once again, there is similarity between CA's correlates and those of Manipulativeness. A s before, they tend to correlate in the same direction, albeit at far lower levels for C A . For example, the interpersonal profile of the high C A scorer is roughly comparable to that of the high M scorer on certain variables, e.g., moderate to strong scores on Arrogant Calculating (BC), virtually the same on W A I (interpersonal) Aggression and as evinced by moderately strong negative correlations with FFI Agreeableness,  4  both styles seem to be, in part,  characterized by a kind of cynicism and (occasionally) flagrant disregard for the niceties of social interaction.  5  There the similarities end, however: A s indexed by CA's moderately strong relation (i.e., r= .48) with I A S - R Assured Dominant (PA) scores, this variable is far more  I do not, however, know of any extant scale that taps this variable (i.e., gratification from positive evaluation of self as physically pleasing stimulus). ..Again, however, CA's correlations (—) with Agreeableness and IAS-R LM are substantially lower than between those variables and M scores (as is the relation between the CA scale and WAI Considerateness levels). ..I note, however, that CA endorsements seem to reflect less interpersonal 'toe stepping' than do high M scores, viz., a correlation of .52 between M and WAI Irresponsible vs. a mere .26 with CA. 3  4  5  128  saturated with the independence and self-confidence that characterize, for example, high Assured Leadership (and the P O N D F U P C ) : Similarly, the overall C A - B S I correlation is a mere -.08 (as opposed to a positive albeit low relation between the latter and M ) . In terms of overall self-ascribed psychosocial distress (as indexed by W A I values) the high C A scorer reports little (r Global Distress, C A scores = -.18). In general, however, the pattern of intercorrelation between C A scores and the external variables assessed is comparable more to that observed for Assured Leadership (albeit at a lower level) than it is for that of Manipulativeness. Granted, C A scores seem more associated with the interpersonal aggression, lack of restraint and the consideratelessness that the W A I measures than do those on A L , but they, as opposed to Manipulativeness items, seem also to gauge a dominance-related, confident social presence. N o doubt, there is overlap between both A L ' s and M's correlational patterns and that obtained for Competitive Ambition, but where the substance of the latter varies, I believe, is typified by the C A item "It's my ambition to somehow 'leave my mark' upon the world": 'leave my mark' it seems, by means either fair (as with Assured Leadership) or foul (via Manipulativeness). The commonality is, of course, achieving influence over others in a social context: The differentiator appears to be the strategy by which it's attained.  Discussion  .  .  Taken together, Studies 2 and 3 suggest that the Profile of Narcissistic Dispositions is a suitable instrument for the assessment of the syndrome of traits that together, appear to give rise to 'normal' narcissism. Indeed, it does so in a reliable and coherent manner.  129  The replication of the Study 2 factor structure was more than satisfactory.  Although  the observed componential structure showed some slight degree of variation from that identified in the derivation sample, its relatively minimal nature is neither especially telling nor problematic. More to the point is the simple fact that the component-based scales remain strongly reliable (i.e., mean coefficient alpha =.82, range=.77-.88, full scale alpha =.90). A n d they all load substantially on the common substantive core.  Thus it appears that the conception of normal narcissism is a measurable entity: From Leary's (1957) and related notions, I assembled reliable scales to measure the hypothesized facets of normal (i.e., non-pathological) narcissism. A s it turns out, not only is the core of P O N D narcissism non-pathological, it evidently assesses healthy psychological functioning. In any case, high P O N D scorers certainly describe themselves as dominant, assured and relatively free of psychosocial distress. The question remains—do others share this view? In the final study, I assess whether peer ratings confirm this self-image.  130  STUDY 4: Additional Correlates and Peer Ratings Although Studies 2 a n d 3 provide promising evidence for the convergent and discriminant validity of the P O N D , those data are, of course, circumscribed to the self-report realm. Study 4 represents an extension of this evidence to include observer perceptions. Among other things, this methodology permits evaluation of the P O N D ' s validity in the prediction of ordinary language conceptions of narcissism. Given my focus on normal narcissism, asking raters to judge a 'folk concept' (e.g., Gough, 1957) rather than a more clinically-derived concept was considered preferable. A n d for the same reason, I used peers instead of clinicians to render the judgment. Accordingly, the criterion ratings (described in detail in Method, below) asked raters to focus on a reasonably observable behavior, namely verbal self-presentation, specifically, speaking about oneself in "an overly positive way". This target behavior was selected as a reasonable representation of the adjectives (e.g., egoistic, conceited, arrogant, etc.) that mark conceptions of normal narcissism and indeed, might be expected to jibe with ordinary language peer descriptions (and also suggest substantive consonance with certain P O N D scales).  T w o 'kinds' of narcissists were described, those whose self-enhancement is grounded in the bulwark of a genuine self confidence in their positive attributes, i.e.,"egotistical" narcissists (who "really believe that they're great") vs. the insecure narcissist whose comparable style stems from insecurity ("deep down, they have a negative view of themselves"). In a very fundamental sense, these descriptions correspond to certain conceptions advanced earlier: The "egotistical" to the normal narcissism assessed by the P O N D , the "insecure" to Millon's (1981) notion of what underlies the superficially-confident  131  self-presentation of the narcissist.  Accordingly, I expect that high P O N D scorers will be  rated as egotistically, but not insecurely narcissistic. In addition to the direct ratings of narcissism, a variety of other peer-ratings were collected, including ratings on the B i g Five personality traits. A s demonstrated in  Studies 2  and 3, the common variance in the P O N D appears to assess an interpersonal style characterized by a self-confident, competitive and domineering social presence. I predict that, because high scorers on these scales present themselves in such a forceful fashion, they will evince an interpersonal style consistent with that self-perception, i.e., one rated by peers as extraverted and non-neurotic. Peers should also, however, find those self confident narcissists somewhat arrogant or irritating. The peer-ratings of agreeableness (or lack thereof) should indicate this. T o pursue this notion further, I included six other ratings addressing possible interpersonal reactions to narcissists: namely, adjustment, overestimation, supportiveness, defensiveness, bragging and sympathy. These ratings could illustrate the interpersonal sequelae of this style: If P O N D narcissists are rated as skilled and competent, do they still over-rate themselves and brag? These behaviors would presumably be seen as rather irritating and offensive. Despite this, I predict that P O N D narcissists will nonetheless be seen as well-adjusted. Finally, I considered the potential reasons why P O N D narcissists have a such a positive evaluation of themselves. Perhaps others, too, see them as gifted persons. Accordingly, I obtained peer ratings on a range of characteristics that I thought might give rise to impressions of assured social presence. These included: perceptions of verbal and social  132  skills, intellectual abilities, general competence/performance and of course, self-esteem. Thus while P O N D narcissists feel pretty good about themselves, it remains to be seen whether peers have a similarly positive perception. Study 4 also affords the opportunity for another (albeit relatively small-scale) replication of some of the empirical correlates of the P O N D (and its core) examined in Studies 2 a n d 3, for example, I A S - R and the FFI. Relations between the P O N D and three additional measures will also be assessed, including Block's (1989) Ego Resilience (ER), Paulhus (1991) Balanced Inventory of Desirable Responding (BIDR-6) and Rosenberg's (1967) SelfEsteem scale. T o this day, there is a hoary controversy concerning correlations with socially desirable responding. Some argue that personality scales correlate with such measures for substantive reasons (Block, 1971), others view such correlations as contamination (e.g., Edwards, 1970; Jackson & Messick, 1962). Nonetheless, it is an empirical fact that measures of desirable responding are reliably and positively correlated with indices of psychological 'health' and adjustment and conversely, negatively with standard measures of maladjustment or psychopathology (see Paulhus 1984, 1986, & 1991 for able summaries of these issues). Whether stemming from positively biased self-perception (and collateral tendencies to discount minor failures, ignore criticism, etc.; see Paulhus, 1986) or veridically-grounded adjustment and psychological well-being, there are substantial empirical bases from which to argue that such measures are valid indexes of 'healthy' interpersonal functioning. The B I D R Self-Deception (SDE) scale (Paulhus, 1991) is one such measure and is interpreted as simultaneously gauging adjustment, auto illusion (defensive biases in the processing of self-  133  relevant information; Paulhus, 1986, p. 154) and self-deception, (see also Taylor & Brown, 1988). Thus given my previous demonstration that certain P O N D scales tap assured selfconfidence, I expect that Vain Exhibition, Assured Leadership, and the P O N D core, for example, will be at least moderately related to this scale. B I D R Impression Management (IM), is claimed to tap a conscious attempt at positive self-presentation thus representing a genuine confounding variable (but, see also e.g., Hogan,. 1983, Jones & Pittman, 1982 on substantive interpretation of this general dimension). Given that the administration conditions involve minimal demand for impression management, the I M scale should be essentially orthogonal to most P O N D scales, except those for which an intrinsic lack of concern for positive self-presentation is a factor (i.e., Manipulativeness and Competitive Ambition) where correlations should be negative. The inclusion of Rosenberg's (1965) Self-Esteem scale is primarily to test convergence with other indicants of a strong self-confidence in one's personal attributes. After all, the Rosenberg scale is the single most widely-used measurement of personal adjustment. Block's (1989) Ego Resilence scale, also a respected measure of.general adjustment, is included for the same reason. A n d finally, this study affords the opportunity for another (again, relatively-) small-scale assessment of the factorial stability of the P O N D . Given, however, that the size of the present sample (i.e., approximately 80 Ss, see Method, below) falls considerably short of the 'Nunnally rule's' suggested size (i.e., 10 Ss per variable), one might well anticipate some degree of componential instability (e.g., item drift) relative to the P O N D derivation sample.  134  Method Subjects In partial fulfillment of course requirements, an intact third year undergraduate personality class served as participants. While the class comprised 89 students (55 females and 34 males), incomplete data and absences during relevant rating sessions(s) (see Procedure,  below) reduced the available ./V to 82 This reduced sample included females and 1  males whose ages ranged from 19 to 45 (Mean = 23.5; SD =3.5); the majority, however, fell in the age range between 20 and 24.  Instruments Although an extensive battery of self-report measures was administered as part of a larger, (unrelated) study (Paulhus & Bruce, 1992) only those instruments relevant to the present rating and replication issues are described. The Balanced Inventory of Desirable Responding - 6th Version (BIDR-6; Paulhus, 1991). The B I D R - 6 is a 40-item 7 place Likert-rated measure of self- ( S D E ; 20 items) and other(IM; 20 items) -deception. In line with the reasoning that these styles represent systematic exaggeration of the denial or ascription of threatening material, only extreme (i.e., 6 or 7) responses are scored. Reliabilities are acceptable and Paulhus (1991) presents information attesting to the validity of the scales in measuring individual differences in characteristic self-presentational styles.  Ego Resiliency (ER; Block, 1989). See Instruments, Study 3.  ' Please note that while self report zero order rs were available for 84 Ss, peer rating data were, in some cases, obtained from a slightly lesser number (although the N was never less that 77.  135  Five Factor Inventory (FFI; Costa & McCrae,1989). This, again, is a shortened (60-item), reliable and valid version of the N E O PI; see Instruments, Study 2. (Revised) Interpersonal Adjective Scales - B i g Five Version (IAS-R B 5 ; Trapnell & Wiggins, 1990). See Instruments, Study 2 and the note in Instruments, Study 3. Narcissistic  Personality Inventory (NPI; Raskin & Terry, 1988). See Instruments, Study 1.  Profile of Narcissistic  Dispositions  ( P O N D ; Taylor, 1991). O n the basis of the itemmetric  data gathered in Study 2, the final version of the P O N D comprises 52 self-descriptive statements rated on a Likert scale ranging from "strongly disagree" (1) through "strongly agree" (6). The measure and its component scales are reliable (e.g., Full Scale Alpha = .89). Rosenberg Self-Esteem Inventory (Rosenberg, 1965). Rosenberg's self-esteem measure is a 10-item index of positive self-evaluation; its reliabilty has been demonstrated repeatedly. Narcissism  Peer Ratings  Discussion group members (see Procedure, below) rated each other on two separate variants of what can be construed as 'common sense' or "folk" (Gough, 1957) conceptions of narcissism. These were described as follows:  A narcissist is a person who talks about (him or herself) in an overly positive way. There are two types: The egotistical [narcissist] really believes that they are great, while the insecure [narcissist], deep down, has a negative view of themselves. Subjects made their ratings of each other on separate 10 place scales ranging from "very low" (1) through "very high" (15) on the appropriate dimension. The single numerical rating on each scale served as the criterion of peer-rated narcissism.  136  Additional Peer Ratings Subjects rated each other on single-item B i g Five scales each bolstered with two related facets: Extraverted (talkative, forceful), Agreeable (warm, nurturant), Conscientious (hardworking, goal-oriented), Neurotic (anxious, negativistic), and Iimaginative (clever, philosophical). Each rating was made on a 10-point scale. Peer ratings on a range of variables related to social and intellectual skills were also assessed. This included peer ratings of the following dimensions: Self-esteem, wisdom, creativity, intelligence, verbal skills, social competence, problem solving and general group performance. Subjects made their ratings of each—with reference to the target's standing on the relevant dimension—on separate 10 place scales ranging from "very low" (1) through "very high" (10). Wisdom, creativity and intelligence ratings comprised a composite index of Mental Abilities. Ratings designed to gauge the interpersonal sequellae of the P O N D narcissist's style (e.g., is self-accepting, tends to brag, peer-rated adjustment; see Table 15b) were also obtained; these are described below.  Procedure Throughout the first week of class but prior to the initial group discussion, subjects were administered the self-report battery; the P O N D and the NPI were completed on a take-home basis and submitted to the instructor prior to the relevant peer-rating session. Participants were naive as to the constructs assessed. During the second week of class, subjects were randomly assigned to one of sixteen discussion groups. Participants in each of the groups (5 to 7 individuals) were previously  137  unacquainted and group composition was relatively heterogeneous with respect to characteristics such as age, gender, race and students' academic interests. Procedure for the relevant rating period (and each of the other 6 group meetings) was as follows: (1) A discussion topic, paralleling course lectures and designed to facilitate group discussion and allow for the emergence of rating-relevant, non-test behavior was assigned (e.g. Sternberg's conceptions of intelligence; Allportian constructs of adjustment, social issues, etc.); 2) A n approximately half-hour discussion period took place in a preassigned, relatively private environment (and students were advised that all exchanges were to be held in strictest confidence); 3) Assigned rating tasks were sealed in envelopes and submitted to the instructor prior to the following group meeting; and 4) Students were advised to refrain from extra-group socializing. Following a group discussion on "creative and absorbing experience," in week 7, participants provided peer ratings on the dimensions of "egotistical" and "insecure" narcissism.  Results and Discussion The first analysis was a P C A of the 52 x 52 inter-item product-moment correlation coefficient matrix; five components were extracting and rotated to an oblique (oblimin) solution that accounted for 46.3% of total P O N D variance. This figure is impressive given the virtually inevitable structural instability owing to the significantly decreased sample size (relative to the derivation N of 660) in the present study (i.e., N=S2). Although these facts undoubtedly militate against the likelihood of preserving structural stability, inspection of the current solution reveals striking similarity to that observed in Study 2. Although the order of emergence differed from Study 2 (i.e., Vain Exhibition emerged first, followed by  138  Competitive Ambition, Public Recognition, Assured Leadership and Manipulativeness; accounting for 17, 10.2, 8.1, 5.9 and 5.1% of observed variance, in turn), each component retained a minimum of 75% of its original item complement (Range = 75 - 82%, M=79%). It is reasonable to assert that on average, a mere 2 0 % item drift (per component) in a small replication sample is good evidence for the structural stability of the P O N D (especially given that the stability of the solution has already been demonstrated).  I next reconstructed the P O N D scales: Scale means, standard deviations, scale intercorrelations and alpha reliabilities are shown in Table 12.  Insert Table 12 about here  I should note, however, that a P C A of the 5X5 matrix of interscale correlation coefficients once again replicated the comparable finding of  Studies 2 and 3, i.e., a strong  first (unrotated) principal component (core) emerged; this time it accounted for 4 9 % of observed P O N D variance. Given the reliable emergence and stable structure of this component, I feel confident in asserting that it likely represents the substantive 'essence' or core of POND-assessed narcissism.  A n d once again, that core appears to be non-neurotic, competitive and vain selfassurance. While the core's correlates are shown in Table 13, the essence of their coherence bears highlighting. For example, while the core is, once again, correlated with I A S - R Assured Dominant, it is more strongly related (i.e., r=.47) to Arrogant Calculating (BC) or,  139  as Leary (1957) originally conceived of this dimension (in the IAS's progenitor, i.e., the ICL), Competitively Narcissistic. High P O N D core scorers self-describe as non-neurotic on the F F I and their peers concur in this assesssment (i.e., r=-.27). These respondents are not seen by their peers as especially modest (r core, "tends to brag" =.36), perhaps because they evidently feel quite good about themselves (r core, Rosenberg Self-Esteem = .36). A n d consistent with the subscale selfreports and peer ratings, high P O N D core scores are related to F F I Extraversion, (dis)Agreeableness, and (as noted above) non-Neuroticism. Correlations among the P O N D , its subscales, and the core scores and the battery of personality measures (self-reported and peer-rated) are reported in Tables 13, 14 and 15a and b, and described below. Implications of these results are considered further in the  General  Discussion. Inspection of Table 12 shows that reliabilty and internal consistency values are virtually identical to those observed in Studies  2 and 3, indicating good stability of the P O N D and its  component-based scales. Additional evidence for the stability of the P O N D , its core and the subscales is the clear replication of the previously observed pattern of empirical correlates, shown in Table 13  Insert Table 13 about here  140  While several of the substantively more important correlations are identical, it is true that some comparable values diverge somewhat. These fluctuations, do, however, appear to be confined to the lower-level, less reliable relations. For example, between F F I Openness and Assured Leadership, the correlation is .37 in the present sample, the comparable value is r=.19 in the derivation sample. In any case, all correlations remain in the appropriate direction, and the overwhelming percentage diverge by substantially less than the maximum observed discrepancy of .18 points of correlation. It is encouraging that even in this replication group—comprising less that 13% of the size of the derivation sample—the interpersonal and B i g Five nomological net of the P O N D remains reliably intact.  New Self-Report  variables  Consideration of the 'new' correlates reveals some interesting relations. In general, the P O N D core correlates theoretically and empirically consistently and coherently with the individual difference measures. For example, the correlation between the P O N D core and Rosenberg Self-esteem—acknowledged  as a good measure of the confidence associated with  positive self-evaluation—is .36. Similarly, the assured social presence gauged by the P O N D is echoed in the core's correlation with I A S - R Assured-Dominant (PA), i.e., .38; their competitiveness is reflected in the B C (Arrogant-Calculating) value, i.e., r=.47. A n d the 'disagreeableness' related to those same characteristics may be reflected in the r=-25 between the P O N D core and I A S - R Warm Agreeable ( L M ) . Relations with the remaining individual difference measures and the individual P O N D scales are theoretically appropriate and, for the most part, as predicted. Assured Leadership ( A L ) , for example, is moderately and reliably related to each of the adjustment measures,  141  i.e., B I D R - 6 S D E , Ego Resilience and Rosenberg Self-Esteem (i.e., rs = .47, .32 and .55, respectively); Vain Exhibition (VE) is similarly directionally related, albeit at somewhat lower (however reliable) levels. The exception here is the correlation of .56 with the Rosenberg measure, an expectable result when one considers the frankly (and perhaps overly) positive self-evaluation that saturates the V E scale (the Ego Resilience r is slightly higher, as well). Competitive Ambition (CA) was also related to B I D R - 6 S D E and, as predicted, negatively correlated with B I D R - 6 I M (although not, however, to Ego Resilience and Rosenberg Self-Esteem).  A s expected, Manipulativeness was negatively correlated with Rosenberg Self-Esteem and B I D R - 6 I M , a scale, you will recall, said to gauge a conscious attempt at "making a good impression"; (Public Recognition was similarly related to IM), again, underlining the specific lack of concern with positive self-presentation that that scale, in part, gauges. The negative relations among the Manipulativeness scale, Public Recognition and B I D R - 6 I M and the Rosenberg suggests that perhaps the specific variance on these scales may include some degree of insecure narcissism . 2  A s regards the B i g Five scales, relations observed in Studies 2 and 3 were replicated. For example, the P O N D core, A L and V E were all correlated with F F I Extraversion. Those same indices were all negatively related to F F I Neuroticism, and the peer ratings (below) were convergent. F F I (dis)Agreeableness was related to the above indices (with the exception of V E ) , M and C A scores and is considered further in the General Discussion.  Relations with  Note that correlations with Impression Management scale are said to measure content rather than style when collected under conditions of low demand for impression management.  2  142  the remaining individual difference measures are theoretically appropriate and for the most part, as predicted.  Insert Table 14 about here  Peer-ratings and the POND A s predicted, those subjects rated by their peers as egotistically  Ratings of Narcissism.  narcissistic show high levels on the P O N D and PONDcore—for a single item, this correlation (.35) is impressive. In contrast, and as expected, there was no significant relation between the P O N D and ratings of insecure narcissism (note that correlations between these ratings, other variables and the NPI are included for comparative purposes).  A s for the specific P O N D subscales, "egotistical" narcissism is significantly associated with a Assured Leadership, and Competitive Ambition; and marginally associated with Manipulativeness and Vain Exhibition; only Public Recognition is clearly unrelated to peer ratings on this dimension.  3  Collateral stepwise regression analyses were also conducted. The best prediction of peerrelated egotistical narcissism was from Assured Leadership and no interaction terms were reliable; nor did any other scales yield a significant increment in R. Additional analyses indicated that no higher order (i.e., 2- 3- 4- or 5-way) interaction terms were reliable and similarly, an alternative (i.e., forced entry) regression analysis indicated no reliable relations (more detailed results are available from the author).  3  Note that two-tailed tests were used for the subscales because no predictions were made.  143  With respect to insecure narcissism, the POND's core was absolutely unrelated to ratings on that dimension. The primary P O N D subscale to relate to this variable was V a i n Exhibition, correlating r=-.24 with insecure narcissism. Other hypotheses regarding insecure narcissism were partially supported, insofar as elevated Vain Exhibition scores were negatively correlated with Insecure Narcissism. There was the predicted (positive) directional relation of Public Recognition scores to those on insecure narcissism, but the effect (i.e., p<.10) did not achieve a conventional level of significance. Competitive Ambition was unrelated to peer-rated insecure narcissism.  The results of the stepwise regression analyses indicated that Vain Exhibition (-) produced the highest multiple R with peer ratings on this dimension. A s with the previous analyses, no higher order interactions were significant. Finally, it is important to note that peer-ratings of narcissism are not the only possible criterion. Criteria such as overestimation (Paulhus & Y i k , 1995) and hindsight bias (Campbell & Tesser, 1983) might be used. Therefore, I cannot use poor validity on this criterion as a reason for dropping subscales from the P O N D .  Big Five Ratings.  Although not reported in the tables, the observed relations with the  B i g Five (peer-rated) indices 'fleshed out' the portrait of the P O N D narcissist: A s with the self-reports, high P O N D core scorers were rated as somewhat disagreeable (r=-.28), Extraverted (r=.28), and non-Neurotic (r=-.27). Ability and Interpersonal Ratings.  I noted earlier that several additional peer ratings  were collected as part of the larger study; these are shown in Tables 15a and b. Certain relations obtained are particularly telling with respect to an understanding of the P O N D .  144  Insert Tables 15a & b about here  The ability correlates add further richness to the image by assessing a possible source for narcissists' positive self-evaluations. The results lend support to the notion that P O N D narcissists do indeed have the wherewithal to back up their positive self-views; this is strongly suggested by the pattern of positive correlations between the P O N D scales and mental abilities, verbal skills and general social competence. Also striking are the correlations of the P O N D scales with peer-rated adjustment. The fact that the P O N D and its core as well as A L and C A all correlate strongly indicate good convergent validation for the for the self-reports of adjustment.  4  The peer ratings also show that, even though P O N D narcissists  have the 'right stuff to  support their positive self-evaluations, peers seem to find this (or perhaps the narcissist's flaunting of same) rather irritating. They are perceived as (dis)Agreeable on the F F I (r=.25), as "overestimating" their abilities, as "defensive", and as "tend[ing] to brag". A l l this and yet, P O N D narcissists are seen by their peers as as non- neurotic and well-adjusted. A n d while not all P O N D scales had especially high peer rating correlates, Tables 15a & b indicate first, that several did, and secondly, that those observed—even where not achieving a conventional level of significance—are all in a substantively predictable direction.  West and F i n c h (1990), for example, cite one of Campbell & Fiske's (1959) criteria for evaluation of the multitrait—multimethod matrix, namely, that the correlations of different methods of assessing the same trait must be both statistically significant and large enough to warrant deeper consideration. The values obtained for peer-rated/self-reported selfesteem and the POND scales that i n part, measure that variable easily meet that twopronged criterion.  4  145  Similarly, as might be expected, the 'positive' scales (i.e., Assured Leadership, V a i n Exhibition and to a lesser extent, Competitive Ambition) correlated with 'good' peer-rated characteristics (e.g., self-esteem, social competence and verbal skills) while 'less positive' scales (i.e., Manipulativeness and Public Recognition) are negatively (or un) related to those same variables. Taken together, these results suggest that the P O N D subscales—particularly Assured Leadership, Manipulativeness and Vain Exhibition and, especially, the P O N D c o r e — provide useful information in predicting which group members are most likely to be judged as narcissistic by their peers. While the actual magnitude of some of the zero order P O N D — peer-rating correlations might most generously be described as moderate, consider for a moment the factors that militate against even these modest relations from emerging, not the least of which is the relatively small sample size . For example, while group members were 5  reasonably well acquainted by virtue of mutual participation in seven one - half hour discussion periods, the restrictions inherent in their relationships as mere classmates suggest that they hardly comprise a subset of best-known peers. Thus for the targets to be judged as even slightly narcissistic, the relevant interpersonal behaviors must have been relatively striking.  A n d consider also that certain behaviors one might guess would be especially indicative of narcissism (e.g., self-assurance and -esteem, even manipulative and cynical attitudes) require considerable  inference on the part of observers. Ozer and Buss (1991) have argued  It may be recalled that the POND data were collected i n a take-home situation; such data are occasionally construed as somewhat compromised thus for data so gathered to be correlated at all with in-class peer ratings is somewhat impressive.  5  146  that such inference requirements exert a moderating influence on peer ratings, one serving to suppress self-other agreement. Given this, the relative dearth of Manipulativeness correlates, for example, may be in part due to the operation of this moderating influence. Especially, I would argue, as regards acts related to being 'calculating', a category Ozer and Buss (1991) described as virtually "unobservable".  Structural Issues A s to the evidence pertaining to the factorial stability of the P O N D , I note again that flouting (even unavoidably) the 'Nunnally rule' (i.e., 10 times as many subjects as variables) renders some degree of item drift and attendant structural instability virtually inevitable. A n d even while I predicted/anticipated some "possibly substantial" level of componential restructuring (if you will), the degree observed was not especially striking. A s I noted earlier, the order of emergence varied, and there was some slight degree of item drift, most notably between the items gauging the substantively similar aspects of Assured Leadership and Competitive Ambition. Again, however, a mean 2 0 % item drift is not, in a relatively small sample size (such as this one), especially telling evidence of factorial instability. T o the contrary (and despite the exceptions noted above), examination of the data reveals that the solution replicated quite clearly.  A n d similarly, the P O N D core and the component-based P O N D scales were demonstrated to be highly reliable with the alphas meeting (and in some cases beating) those obtained in the much larger derivation sample. Additionally, the pattern of the P O N D ' s observed external correlates was, for the most part, strongly replicated. Correlation values,  147  when they did vary, did so, for the most part, by only a few points; the few exceptions can probably be attributed to small sample-based instability. In light of the results of Study 2,1 alluded to a potential dearth of evidentiary-based discriminant validity between Assured Leadership and Vain Exhibition; again, the present data indicate an intercorrelation of .42. A s in Study 2, the network of external correlates is, once again, similar; the level of Vain Exhibition's correlations with external measures is, to be sure, somewhat weaker, albeit directionally identical. I alluded earlier to (at least) two separate arguments for the scales' retention (despite the correlational similarity): First, the near inevitability of this occurrence given their substantive similarity (and their accordingly shared variance on the P O N D core), and second, the "positive manifold" nature of these constructs. B y positive manifold, I refer, of course, to the well known finding that 'good' attributes are typically co-related: In the present sample, for example, both scales correlate moderately and positively with three separate measures of self-reported adjustment (i.e., the B I D R - 6 S D E scale, the E R , and the Rosenberg Self-Esteem measure), with peer-rated adjustment and negatively with measures of self- and other-rated maladjustment (e.g., the F F I Neuroticism scale). In addition to the scales' shared extraversion/dominance type content, and insofar as variables such as ego resiliency and self-esteem are 'good' things (and I submit that it is fairly easy to argue that they are), I note once more that the commonalities in Assured Leadership and Vain Exhibition's external correlates should, substantively, be anticipated. Indeed, were they not to show similarities in their external correlates, such a finding would be greater cause for concern.  148  Additional data lend credence to the arguments in favor of retention of Vain Exhibition (and its') discriminativeness from Assured Leadership. First, there was a modest (albeit significant) negative relation to the I M scale of the B I D R - 6 (no correlation existed with respect to Assured Leadership); note again, that the B I D R - 6 I M scale is said to tap a conscious attempt at positive self-presentation. Second, there was also a negative (and significant) relation to peer ratings of insecure narcissism. Such findings indicate a number of interpretive possibilities; the most obvious being that Vain Exhibitionists are (to use subcultural parlance) "in your face". In other words, by virtue of their 'showboating' and/or wanting their physicality to be admired and/or appreciated, respondents might wish to appear 'subtle' in that respect by not endorsing items obviously reflecting attempts at positive self-presentation (BIDR-6 IM) but perhaps, by virtue of their wishing to be noticed (e.g., "I get uncomfortable when checked out"; R) they behave in such a fashion as to render their peers' surmisal that even though they "talk about themselves in an overly positive way", "deep down," they in fact have a positive view of themselves—hence the negative correlation between V E and insecure narcissism.  In any case, it is true that neither of these correlations—while significant—was overwhelmingly large. They do, however, intimate that they add substantive variance that is not tapped by only examining that which Assured Leadership gauges. A n d in addition, they suggest that further expansion of the nomological net of the Vain Exhibition s c a l e especially with respect to 'separating' its' unique elements from the self-confidence variance evidently shared with Assured Leadership—may be warranted. Given their strong loadings  149  on the core component, however, they obviously share common substantive variance that corresponds to the essence of P O N D narcissism.  Conclusion It appears that the P O N D core does indeed gauge a normal form of narcissism—one with minimal signs of pathology. The results of Study 4 confirm this conception but also raise some complex questions. For example, it seems puzzling that peers perceived normal narcissists to be disagreeable, yet well-adjusted.  How these and the other results bear upon  our understanding of non-pathological narcissism is considered in the General  Discussion.  150  GENERAL DISCUSSION The corpus of this dissertation is my endeavor to establish the existence of a nonpathological, that is, normal narcissism.  A s I discovered, the literature on narcissism was  delimited largely to 'pathological' or clinical characterizations, that is, abnormal or insecure narcissism. From the few treatments of its 'opposite number', that is, normal narcissism variously construed, I derived the commonalities and assembled a set of subscales to assess them in the Profile of Narcissistic Dispositions (POND). The facets tapped by the P O N D subscales were shown to cohere to a degree sufficient to suggest the existence of a normal narcissism. Rather than a disconnected assortment of facets, they appear to constitute a syndrome. Their common variance, as reflected in the first unrotated principal component, represents the essence or core of normal narcissism. The subscales were shown to represent distinctive, yet overlapping facets. Thus it was deemed useful to retain the subscales to allow consideration of subjects' specific patterns of subscale scores as well as their core standings. Isolation of these facets also accords the theoretical benefit of clarifying the constituent features of normal narcissism. The P O N D was shown to reliably assess a normal narcissism that is not only selfascribed, but is also confirmed by persons with whom respondents interact. The form of narcissism tapped by the P O N D and its core seems better described as 'egotistical' than insecure: This conception is substantiated both by self and peer ratings. The nature of this construct is, of course, multifaceted, but it indicates a consistent pattern of features that justifies the term "narcissism": There was evidence for an overwhelmingly positive selfpresentation, i.e., high self-esteem, ego resilience, and self-assured confidence. T o some  151  degree, this confidence seemed justified by an array of peer-assessed positive social assets, e.g., intelligence and social skills. A n d as with most interesting configurations, P O N D identified narcissists have negative as well as positive characteristics. They are perceived as arrogant, perhaps because they flaunt and exaggerate their positive assets, and their competitive style appears to render them less than likeable. In sum, although the normal narcissists' self-presentation is based on a kernel of truth, their apparent braggadoccio is seen as rather irritating. I expand on these thoughts further below.  The Literature in Retrospect How do the qualities assessed by the Profile of Narcissistic Dispositions ( P O N D ) correspond with those gleaned from the literature? Consider, first, the structural location of normal narcissism: Does it accord with the circumplex vector that Leary (1957) termed "competitively narcissistic"? Insofar as the circumplex is adequately operationalized by Wiggins et al.'s (1988) Interpersonal Adjective Scales-Revised, then the answer appears to be a resounding "yes". Recall that Leary specified the Arrogant-Calculating vector of the circumplex as the locus for normal narcissism; when projected on the circumplex, the core of the P O N D lies almost precisely along that same vector. Moreover, each of the subscales falls within the theoretically-relevant sector of the interpersonal circle, primarily, the upper-left quadrant. The slight angular displacements that do occur (e.g., Assured Leadership slightly past the Assured-Dominant axis) are perfectly theoretically consistent. Because it was designed to assess competitiveness, it was reassuring to see that the Competitive Ambition subscale virtually coincides with the relevant Leary vector. A n d this subscale is also closest to the core's vector on the circumplex projection. Nonetheless, in three separate samples,  152  there is clear evidence that the P O N D scales tap additional substantive variance beyond that which is gauged by the Arrogant-Calculating vector. A s well as having a theoretically-consistent structural location, the content of the P O N D largely cohered with the elements identified in the literature. In factor analyses of the items developed to reflect those characteristics, four of the seven identified features of the normal narcissist emerged as distinct factors and were replicated. O n the other hand, some of the features, for example, ambitious and competitive, merged into a single factor. A clear finding was the evidently good psychological adjustment of high P O N D scorers. Instead of psychopathology, I repeatedly obtained reliable relations with measures assessing healthy psychosocial function, i.e., self-esteem, self-assurance, ego resilience and social competence. O f course, this finding is not unrelated to my general selection of only non-pathological features; I did not, however, avoid negative qualities. The Manipulativeness subscale, for example, can be indicative of what is generally construed as 'non-healthy' interpersonal functioning (witness, for example, the correlation between this scale and W A I Aggressiveness). In any case, I did not select features bearing directly on adjustment. Moreover, the positivity of the association was not predetermined—it could have been null—or even negative. In developing the extant measure of normal narcissism, the NPI constructors attempted to gauge how college populations might reflect characteristics of a clinical syndrome in its "normal" complexion (whatever  that might be); in stark contrast, the P O N D was derived (and  appears) to reliably assess a specifically non-clinically-derived conception of normal narcissm.  153  At this point, rather than recapitulating my findings individually, my discussion will focus on select critical issues that speak to their general nature.  What the POND reflects: The Essence of secure narcissism Given the apparent measurement anomalies and 'pathology error' intrinsic to the NPI, I saw the market for a better mousetrap. Nonetheless, my alternative construction process began with its ostensible conceptual forebear: A s demonstrated in the first three analyses of Study 1, what emerged singularly across methods and formats was that content clusters related to assured dominance and vanity seemed to be the most stable in the NPI—and accounted for the most componential variance. Thus (not coincidentally) it happens that among the P O N D items are characteristics that reflect self-assured dominance—nor, however, due to serendipity (a la the NPI), but because dominance was among the few commonalities that I gleaned from the literature on non-pathological narcissism. O f course, the literature suggested that normal narcissism comprises not just dominance and vanity but a total of seven facets. O f these, a total of four facets appeared as distinct factors (two others cohered as one factor) that were substantiated. Articulating the connections among those facets, however, was not, evidently, a priority in the literature. Freud, for example, never directly addressed what might make a narcissist judged as a "personality", one upon whom others are "specially likely to lean" (Freud, 1911/1914, p.45); he refers instead to their effects upon those same others. Even Leary merely alluded to what might comprise "competitive, self-confident narcissism" (1957, p.332). M y attempt was to render those allusions more tangible, i.e., specify what might give rise to that normal narcissism (and of course, measure it reliably). If not a fundamental  154  insecurity, as is evidently the case in pathological narcissism, then what might be the underlying dynamic? According to Leary, narcissistic expression (be it of an attribute, via exerting interpersonal influence, etc.) can be viewed as a coping behavior in times of temporary stress. In other words, it is a  security operation, in which the narcissist flexes interpersonal muscle  in the aid of comfort from effectance and its concomitant of obtaining "considerable admiration and social approval" (Leary, 1957, p.334). Thus, although the narcissist may be no more insecure than others, he/she uses such behavior as a coping style. O f course, Leary also emphasized the term "competitive" in describing the normal narcissist. A strong competitive drive might lead narcissists toward ambition, but an ambition—according to Leary at least—that emphasizes superiority over others, rather than accomplishment for its own sake. Naturally, this behavior could offend others and be perceived as interpersonally insensitive (i.e., disagreeable). Thus the facets of P O N D narcissism may be linked through a strong competitive motivation. A second possible dynamic linking all other features could be the normal narcissists' strong social presence. A variety of characteristics may underlie that behavior: The desire to be ascendant and the need to be valued or recognized as such; the drive toward, and recognition that others can be manipulated toward developing an impression of skill and/or power; some degree of wherewithal to do so (and get its attendant pleasure, viz., Leary's notion of competitive self-enhancement) and finally, display of, arid gratification from (self and others') acknowledgemt of personally-valued attributes, i.e., "admiring [one's] own strength or beauty or wisdom". (Leary, 1957, p.333). O f course, such a strong social presence  155  would likely create negative reactions in others, therefore giving rise to ratings of disagreeableness, etc. A third possibility is that the normal narcissist is quite justified in his/her positive selfappraisals. The evidence accumulated in Study 4 indicates that not only does the normal narcissist have positive self-sentiments, but also the behavioral repertoire to ensure some degree of their corroboration in the eyes of others. If narcissists indeed possess such socially desirable attributes, then their self-descriptions may be-quite accurate. It could be that others respond jealously by derogating them and rating them as disagreeable. However, my data also show that normal narcissists appear to overestimate their assets. In Study 4, peers rated them significantly higher than non-narcissists on this overestimation dimension. Previous research with the NPI similarly demonstrates that narcissists over-rate themselves even when performance is scored objectively (John & Robins, 1992; Raskin et al., 1991; Paulhus & Y i k , 1994). Thus the normal narcissists' claim that they are admirable persons may indeed contain a 'kernel of truth'; one that may be, however, exaggerated. Yet narcissists do not self-enhance in all respects. They do not see themselves as especially warm and nurturant. T o the contrary, their self-reports are repeatedly correlated with measures of interpersonal aggression, low agreeableness and in general, cynically manipulative attitudes that appear to render them somewhat unlikeable. (I am put in mind of actor Robert Morley's advice to fall in love with one's self at an early age; that way [he argues] if you're not a good mixer, you can always fall back on your own company). The fact that narcissists self-ascribe certain negative traits confirms that they are aware of them. This is important because others' reactions to them may not be just jealousy.  156  Rather, narcissists appear to engage in a conscious trade off: 'getting along' is subserviated to the goal of 'getting ahead'. Summary. This combination of five facets makes the narcisssist an intriguing character: None of the three dynamic theories fully accounts for all five in combination. What does seem certain is that the high P O N D scorer is better described as "egotistically"—as opposed .to "insecurely"—narcissistic.  The POND's representation in the Big Five In evaluating how this portrait of P O N D narcissism blends into the broader tableau of personality in general, I assessed its configuration within one current consensual framework— the familiar B i g Five, a structural model of personality that is robust across methods of analysis, assessment and informants (i.e., self and other; see, for example, Wiggins and Trapnell, in press). T o ensure broad replication, two self-report inventories were included: (1) The Five Factor Inventory in three of the four studies, and (2) I A S - R B 5 scales in two studies. Finally, peer-ratings of the B i g Five were collected in Study 4. In general, these three sources of information corroborated each other. Taking each factor in turn, then,  Extraversion.  The full-scale P O N D and its core are reliably positively related  to this dimension in both self and peer ratings. Nonetheless, it is clearly the social dominance aspect of Assured Leadership (and to a lesser extent, V a i n Exhibition) that correlates most strongly (i.e., mid-40s to 50s) and seems to carry the relation to the core.  157  Agreeableness.  Given that the P O N D core and especially Manipulativeness  and Competitive Ambition scores are negatively related to self and other perceptions of agreeableness, it could be that such items reflect an unabashed 'achieve-at-any-cost'-type striving—and its attendant negative interpersonal sequellae. In Hogan's (1982) terms, such scorers are not especially likeable, and with respect to his trade off of "getting along" and "getting ahead", the egotistical narcissist shows a clear preference for the latter.  Conscientiousness.  In general, the P O N D core and the subscales are unrelated  to this trait. The primary exception are the relatively low (yet reliable) correlations of Assured Leadership with self- and other-reports of conscientiousness. This is understandable in that a forceful group presence would magnify any contribution that an individual had to make toward group goals. This conjecture is supported by the Study 4 correlation between A L and peer-rated "group contribution".  Neuroticism.  This dimension represents a susceptibility to negative affect and  the disposition to be anxious or insecure: It is considered by some to be the general factor of psychopathology.  In three separate samples, the P O N D and  its core have reliably negative relations with Neuroticism in both self- and peer ratings. Although two subscales have low positive associations with  158  some aspects of neuroticism, relations with Assured Leadership, V a i n Exhibition, and Competitive Ambition are moderate and negative.  Openness to Experience. A s indicated in two large sets of self-report data, openness seems largely unrelated to P O N D scores. Possibly owing to the discussion group format of the peer rating sample however, Assured Leadership was somewhat related to self- and other-rated openness, as well as peer ratings of "creativity". It could be that the strong social presence of high P O N D scorers makes them seem unusually curious or creative.  M u c h of the information from the B i g Five evidence overlaps, and is consistent with, that obtained in the circumplex data. This is not suprising, given that narcissism is largely interpersonal (i.e., it projects onto the circumplex) and the circumplex represents a rotated version of the first two dimensions of the B i g Five (Wiggins & Trapnell, in press). Unlike the circumplex, however, the B i g Five does contain a direct measure of adjustment, namely, the Neuroticism factor. Given the importance to this thesis of the finding of positive correlations with adjustment, the B i g Five analyses were especially instructive.  Structure and Reliability Issues Further to the issue of psychometric quality and what might be the sine qua non of measurement, i.e., reliability, I have described how analysis of the matrix of P O N D interscale coefficients yields a strong, replicable first component that in all cases, accounts for 40-49%  159  of observed P O N D variance (I note that a similar NPI analysis accounts for only 35% of its observed variance). A n d in those three analyses, what consistently marks this component are, interestingly, Competitive Ambition and Vain Exhibition, with Assured Leadership as a the strong third marker. There is mixed evidence about which facet is most important. Although the strength of Competitive Ambition and Vain Exhibition's marker status consistently surpasses that of Assured Leadership, the latter typically showed the strongest correlations with standard personality variables. It is also true, however, that both Competitive Ambition and Vain Exhibition reliably correlated in the same direction and most often with the same variables as did Assured Leadership, thus again suggesting their shared variance. Thus the claim that the Competitive-Ambition component represents the 'essence' of normal narcissm has at least two points of support. First, the marking strength of Competitive Ambition underscores its validity as an index of the core of normal narcissism. Second, that component appears to have substantial shared variance with the next 'strongest' markers of the core of normal narcissism, i.e., Vain Exhibition and Assured Leadership. T o summarize, the componential interscale analyses suggest that while Assured Leadership is an important marker of what is, evidently, the 'essence' of normal narcissism, that scale's importance seems overshadowed by the even greater marking strength of Competitive Ambition and Vain Exhibition on the strong first component. It could be that dominance, although linked closely to major dimensions of personality, is important in normal narcissism only to the extent that it shares some substantive variance with competitive, vain exhibition.  160  Origins In reviewing the literature described in the Introduction,  I observed that (etiologically  speaking) there seemed to be theoretical concurrence on the idea that disorderered parentchild interactions are what initially gave rise to a fundamental insecurity, one marked in adult behavior by the attempt to get formerly parent-denied approval and/or admiration. Most speculated that the insecurity assumed to underlie adult, clinical narcissism stemmed from early parental overvalaution (e.g., Freud, Horney, Millon, Murray), parental  undervaluation  (also Horney), or coldness or indifference (e.g., Kernberg, Kohut). In writing on narcissistic patients for example, Leary (1957) noted that they were "consciously disidentified" with their parents, whom they perceived as "relatively sadistic" (p.334). I pointed out that data so obtained were, of course, more than a tad equivocal, insofar as they might be beset by, for example, memory-based or self presentational biases; Bell and Byrne (1978), for example, found that the veracity of adults' recollections of their early relationships was demonstrably ^suspect. In any case, 'obsessed' as I was by isolating any commonality in the narcissism literature, I tried to provide a miniature (albeit distal) test of the theory that the child is indeed father to the man. T o do so, I examined the properties of four relevant items in the original pool, i.e., items such as "When I was growing up, my parents were cool and indifferent toward me." A s it turned out, those items (dropped from the ultimate P O N D ) cohered tightly, forming their own component that, unfortunately, didn't correlate with anything substantively  161  meaningful. For example, one of the highest relations obtained was a correlation of .25 between this component ("parental aloofness") and I A S - R Gregarious-Extraverted. Thus while speculation as to the etiology of narcissism might be compelling, these data do not really substantiate the "insecurity hypothesis"~one way or the other. M y other data do, of course, suggest that insecurity is not (consciously at least) an issue with high P O N D scorers. However, given the component's threadbare nomological net, added to the fact that it wasn't related to any other component, suggested two possible interpretations. The first is that I may not have written adequately evocative items. The second is that the "insecurity hypothesis" is irrelevant to the evidently secure narcissism assessed by the P O N D . I shall offer one last point with respect to narcissism and insecurity. I noted above how most theorists view insecurity as underlying (i.e., 'motivating') narcissistic behavior, and how narcissistic, ostensibly confident self-display is rather merely a blustering front belying the insecurity (and implied dependence on others' positive evaluation) that lurks behind it. Birtchnell (1984), for example, has argued that one component of dependence is what he calls ontological  insecurity (i.e., lack of identity, low self-esteem, etc.), which  substantively corresponds to the circumplex location labeled "Unassured Submission" (Wiggins and Pincus, 1989). For normal narcissism, my data show a different location: I obtained negative relations between the P O N D and Hirshfeld Dependency, moderate to strong negative correlations between the P O N D and IAS R Unassured Submission (in three separate samples) and self-reports and peer ratings of high self-esteem in high P O N D scorers.  162  Thus it seems clear that the normal narcissism assessed by the P O N D is not underlaid by insecurity, but is instead characterized by a self-assured social presence whose impact is a front only in terms of bolstering the strong and secure self-esteem that lies behind it.  From normal to abnormal narcissism: A continuum? A s regards what might be described as the "brambly problem" of normality, a great deal has been written with respect to the continuum notion, i.e., the idea that personality traits— extraversion, self-regard, any number of traits—form a continuum, ranging from their normal expression (flexible and adaptive) to levels at either extreme that are considered abnormal (inflexible and ma/adaptive). A n d it is this latter 'intensity' that is thought to partially underlie personality (and other) disorders. For example, in the Introduction to the Axis II (personality) disorders, the committee of D S M III contributors (prominent among them, Millon, an advocate of the 'narcissism continuum' notion) declares that it is when "personality traits are inflexible and maladaptive and cause either significant functional impairment or subjective distress that they constitute personality disorders ( A P A , 1987, p.335; D S M I V similarly asserts that). Unfortunately, this view is founded primarily upon clinical judgment and experience, and as Wiggins and Pincus (1989; and others, e.g., Livesley, 1987) have observed, a paucity of empirical evidencesupportive or otherwise. A n d with specific reference to narcissism, Carson (1969) and Millon (1981), as well as Leary (1957), are all very much explicit about a continuum from normal to pathological and, as I noted in the Introduction, these theorists (and the D S M ) view the difference between normal and abnormal narcissism as one of degree, not kind. But given the rather disordered  163  state of the relevant literature, assessing the tenability of this proposition begins with footing that is rather slippery, at best. In response to this general issue, (at least) two questions arise: 1(a) Is there, in fact, a continuity between normal and abnormal expression of traits (with respect to the (in)flexibility question) and (b) if so, is there some way to empirically quantify this? and (2) should the POND—derived specifically to be a measure of normal narcissism—be able to assess abnormal narcissism? The answer to 1(a) is "I don't know", primarily as there is only a small and conflicting empirical basis upon which to assess the tenability of this proposition, and to 1(b), "Well, it depends upon whom you ask". A s far as I can determine, there are relatively few ways to measure trait/behavioral flexibility by self-report. The two that I am aware of are Paulhus and Martin's (1988) Functional  Flexibility,  and Wiggins' notion of IAS octant score-based "diagnostic category",  specifically, one's vector length as an index of behavioral inflexibility and implied maladaptiveness. With respect to the first, Paulhus and Martin's (1988) measure was, of course, derived, normed and validated on a 'normal' college sample, so the characteristics of that group are probably comparable to those of the samples described herein. However, the measure's relation to flexibility in clinical populations (whom, presumably, comprise the 'abnormal' comparative pole), or its applicability to the realm of normal narcissism is unknown and remains to be demonstrated. A s regards the Wiggins, Phillips and Trapnell (1988) procedure for determining 'interpersonal type category', the reader may recall that I did, in fact, examine this in Study 2,  164  finding, for example, that the P O N D core showed weak to moderate relations with the relevant I A S - R 'diagnostic category' i.e., (BC) Arrogant-Calculating. A number of qualifiers are important here, however;: The B C scale is, of course, relevant to the P O N D insofar as its conceptual forebear was Leary's (1957) notion of competitive narcissism, and my oft-repeated evidence that B C levels are reliably and moderately related to the P O N D (and certain subscale scores.) However, throughout this dissertation I have argued that normal narcissism is considerably more than 'just' B C , and have convergently demonstrated that. Stated another way, relevance—however critical—is not equivalence. Additionally relevant is Wiggins et al.'s (1989) research specific to vector length within the B C diagnostic category in which the authors examined that variable's relation to the Lanyon (1973) Psychological Screening Inventory. The only scale (of four) to correlate with Arrogant-Calculating vector length (in a sample of 20 subjects) was Social Non-conformity; I note that this variable did not correlate with the Discomfort scale—one described as a measure of maladjustment, anxiety or neurosis. The .41 correlation with Social Non-conformity  was explained by the first author's  previous demonstration (with Broughton, 1985) that Arrogant-Calculating scale scores correlate with Christie and Geis' (1970) Machiavellianism index and the Buss, Higgins, Gomes and Lauterbach (1987) Tactics of Manipulation.  With respect to the latter finding, I  note first that the Buss et a/.(1987) research that Wiggins et al. (1989) allude to examined the relation of the Tactics to Arrogant-Calculating scale scores, not vector length, and second, I  165  examined the relation between the P O N D and the Buss et al. measure, finding reliable but only low (i.e., low 20s) correlations between the Tactics and only certain of my subscales. In any case, one must first 'buy' the assumption that extreme vector length is equivalent to behavioral rigidity or inflexibility, and second, accept the premise that one Lanyon (1973)related finding—not even with the scale said to index maladjustment—is adequate evidence for assuming related rigidity. Accepting that assumption strikes me as tantamount to accepting a theoretical position, while the evidentiary basis for its premise (i.e, a low to moderate correlation with only one of four psychological screening scales) shouldn't be the only grounds on which to assume unequivocal abnormality or even category-determined 'rigidity'. Especially  as this Wiggins et al. (1989) finding is 'guilt by association' in that my measure  happens to correlate with a scale found to correlate with one of four separate scales of maladjustment in an unrelated sample. So, with respect to the continuity issue, I think that my various data sets bear directly upon it insofar as extraversion and neuroticism—two reliable P O N D correlates (positive and negative, respectively)—are in fact dimensions of personality and (as Wiggins and Pincus, 1989, have shown) its disordered state. One relevant demonstration is that analysis of the Personality Adjective Checklist ( P A C L ; described earlier) and the MMPI-based personality disorder (PD) scales yield components that correspond to the 'Big Five' (also discussed earlier, and below). Wiggins and Pincus (1989) found that a Principal Components Analysis of the combined Neuroticism, Extraversion and Openness to Experience Personality Inventory ( N E O PI), P A C L and M M P I P D scales' item pools revealed a B i g Fiveinterpretable structure upon which the latter two's Narcissism scales loaded low to  166  moderately on components of extraversion (+) and neuroticism (-), and, to a lesser degree, on antagonism, or low agreeableness. In short, I would not disagree that there is a continuity between normal and abnormal expressions of being outgoing and getting along with onesself or others; I would, however dispute that the P O N D can or indeed should measure rigidity or inflexibility, and further, and that the latter is adequately operationalized by any current measure. So in an admittedly somewhat roundabout way, I think that I've at least partially addressed the question of should the POND—designed specifically to gauge normal, nonpathlogical narcissism—be able to assess the 'submerged' abnormal pole of the construct and its implied behavioral rigidity? Begging that question, however, is another, viz., "Does it?" Judging from the (considerable) data I have obtained, the answer is "No, apparently not." The P O N D consistently shows negative correlations with 'maladjustment' and reliable, moderate positive relations with a number of self-report indexes of self-assurance, ego resilience, selfconfidence and, more generally, self-esteem. Moreover, peers convergently validate the strong self-esteem of high P O N D scorers. O f course, moderate negative correlations between the P O N D and measures of pathology may obscure the fact that extremely high scores show some psychopathology. In other words, the relation between the P O N D and adjustment may be curvilinear. T o address this possibility, I conducted additional analyses with my Study 4 data (details available from the author). Insofar as extremely high or extremely low P O N D scores might indicate maladjustment, an examination of a scatter plot of adjustment against P O N D scores should reveal an inverted-U pattern. A total of four plots were examined: Self-reported adjustment  167  (as indexed by ego resilience and self-esteem) and peer ratings of that variable (assessed via ratings of self-esteem and neuroticism) against P O N D scores invariably revealed a monotonic relation between the two. In no case was there evidence of curvilinearity: Rather, the relation between P O N D scores and adjustment is best described as a continuous, positive trend. T o conclude, I cannot argue that no trait shows a continuum between normality and abnormality. I can conclude that the P O N D likely does not. Nor should it be incumbent upon the POND—again, a measure of normal narcissism—to have a range of convenience that extends beyond its focus. In sum, I argue that in one sense there is some degree of continuity between the abnormal and normal expression of narcissism; the possible commonality seems, again, that they share the same underlying personality traits as represented by the 'Leary vector' of the circumplex, i.e., Arrogant-Calculating. I do not believe, however, that the pathological narcissism is simply an extreme version of normal narcissism. Rather, they differ in kind with respect to one fundamental differentiator, i.e., distress or insecurity.  Definitions of Adjustment: Personal or Interpersonal? I have noted that as indexed by standard measures of psychological distress, i.e., the Basic Symptom Inventory and the Weinberger Adjustment Inventory (not to mention the Neuroticism scale of the FFI), high P O N D scores seem, for the most part, to be uncontaminated by pathology. M y evidentiary basis for this claim is that in at least two studies, P O N D scores are un- or negatively correlated with instruments said to gauge psychologicaly disordered behavior. It is true, however (as I have also noted), that while  168  P O N D narcissists describe themselves as relatively free of psychosocial distress, this is not to say that they can not be instrumental in causing it in others. For example, the P O N D core is associated with several negative interpersonal behaviors on the Weinberger Adjustment Inventory (WAI). For example, Aggression, Irresponsibility and Considerateness (negatively related). Interestingly, this pattern of correlations resembles a typology that Weinberger and Schwartz (1990) chararacterize as "undersocialized", that is, scoring low on both global factors of the W A I : Distress and Restraint. They obtained correlations with this type that are consistent with those obtained for the P O N D , i.e., negative relations with neuroticism and agreeableness, and positive correlations with assertiveness and self-esteem. In fact, Weinberger and Schwartz (1990) suggest that individuals with low global distress and restraint scores—such as high P O N D scorers—tend to have "subclinical" characteristics of the narcissistic personality, they are nonetheless high in self-esteem and well-being. A s the authors phrase it, such scorers are probably not "particularly concerned about others but [are] basically confident about their ability to meet their own needs" (p. 391). Thus in this case (and as suggested by other correlates), high P O N D scorers seem to choose 'getting ahead' with the attendant interpersonal sequelae of not 'getting along'; their IIP self-reported problems (e.g., Domineering, Vindictive) are consistent with this. T o the extent that interpersonal offensiveness is considered to be a form of maladjustment, one could argue, therefore, that high P O N D scorers are maladjusted. A n d apparently, Weinberger seems to imply this view of maladjustment by including such behaviors as aggression, irresponsibility, and inconsiderateness.in his adjustment inventory.  169  Nonetheless, it appears that the narcissist's problems lie more on Weinberger's Restraint factor than on the Distress factor. In short, one's conclusion about maladjustment in narcissists may ultimately rest on one's fundamental view of psychological adjustment.  What about the NPI? Based on the evidence adduced here, the P O N D seems to surpass its nearest competitor, the Narcissistic Personality Inventory (NPI)—conceptually, psychometrically, and practically, i f not, predictively. The tripartite series of analyses detailed in Study 1 provided a critical evaluation of the structural and psychometric characteristics of the heretofore bestvalidated measure of normal narcissism, namely, the NPI. Recall, however, that its construction was directed by the "original conceptual rational [sic] of developing an item pool that would fully reflect the D S M III criteria for the narcissistic personality" (Raskin and Terry, 1988, p.892.). The appropriateness of that presumption aside, because the NPI is the only multi-scale inventory purporting to measure normal narcissm, I felt that it was incumbent upon me to examine it empirically. A s the reader may recall, I collected a large-scale sample where subjects completed both the original forced-choice NPI and a Likert-rated form of the narcissitic alternatives of the instrument. From analysis of the resulting matrices, rather ill-defined structures emerged. Thus I concluded, given Raskin and Terry's (1988) own retention criteria, as well as Nunnally's (1978) considerations of what comprises "strong structure" (e.g., each component with some essentially univocal items, at least four items loading strongly on only one factor, etc.) that the NPI was only relatively structurally stable.  170  The question arises then, why, with a large-scale sample, was I unable to fully replicate the structure reported by Raskin and Terry (1988)? The first possible answer involves considering differing sample characteristics. For example, were age and gender distibutions markedly different, one could argue that the responses of a non-comparable sample (e.g., more or less heterogeneous) could yield a different structure. A s Raskin and Terry's (1988) sample composition information provides age and gender breakdowns, one can at least partially determine that the samples described in Studies 1 through 4 are, in terms of those variables, very much comparable to the college undergraduate groups typically reported in the relevant literature (e.g., JPSP)—and indeed, to those of Raskin and Terry (1988). While my sample composition reflected the fact of greater female college enrollment, this is as equally true in California as in British Columbia. More importantly, although Raskin and Terry (1988) failed to report age and gender correlations, my own are both unremarkable and perfectly consistent with past literature . One final possibility invokes the potential influence of different ethnic composition (e.g., more Latin subjects in the Southern California samples). Again, though, Raskin and Terry (1988) do not report the relevant figures; one fairly recent study (Smith, 1990) found that Latinas did not score significantly different on the NPI than did Caucasian women. With respect to the potential influence of a greater Asian proportion in the current samples, again, I can not directly assess this given that Raskin and Terry (1988) do not provide ethnicity data. While Smith (1990) found that his female Asian subjects had significantly lower NPI scores than did Caucasian females (males were not considered), he also found that in his sample, female Caucasians scored significantly  higher than did those Raskin (1988) reported in  171  unpublished normative data. None of those studies, however, compared the structure of narcissism in Asians and Caucasians. Recently, however, Paulhus and Y i k (1994) found that although their Asian subjects scored lower on the NPI, their overall patterns of personality variable interrelation were directly comparable to that observed in non-Asian subjects. In any case, I believe that the above concerns are rendered moot by the fact that the sub- and full scale NPI normative statistics—in the data reported by those authors as well as that from my own sample—are practically identical. Where they do vary, the amount is trivial, e.g., a mere one third point difference on a six-item scale. Having ruled out that possibilty, I considered alternative explanations for the absence of a complete replication. Chief among these was the issue of (un)reliability of measurement in the NPI. I earlier described a number of such problems plaguing the instrument, among them, relatively low internal consistency (e.g., four subscale Lambda 3 values <_ .54). I argued that evincing measurement error on so fundamental a level could easily render structural solutions that were less than stable. Similarly, I held that the NPI's dichotomous, forced-choice format could introduce method variance, i.e., skewed endorsement frequencies arising from variations in item content social desirability; as Comrey (1973) has observed, results from the analysis of dichotomous items might be especially unstable owing to the possibility of extreme item endorsement splits. I noted Edwards' (1970) admonition that such dyads must be of comparable if not fully equated social desirability; I also noted that such evidence is not proffered for the NPI. T o best address these potential confounds, I tested how adopting a less constrained, multiple-response Likert-type item format might influence reliability, and concomitantly,  172  underlying componential structure. Again, the analysis of the Likert-rated narcissistic alternatives yielded a structure very similar to my first two: A s before, the first four item clusters evinced a composition fairly true to that of Raskin and Terry (1988), while the latter three were fractionated and unstable. Given the results of these three analyses—consistent, I note, across different correlational methods and item formats—I can only conclude that a seven component structural solution for the NPI is not wholly appropriate. A n d finally, despite the relative non-replicability of their reported structure, it is telling that as I had suspected (and as the literature suggested), opting for a Likert-type over a forced-choice item format gave rise to substantial increases in reliability (e.g., a mean subscale alpha of .72 vs. their Lambda 3 of .58). So while Emmons (1987) argues that the NPI "possesses desirable psychometric properties" ( p . l l ) , my data suggest otherwise.  A Head to Head  Comparison.  Although comparative validity studies can be embarrassing, I ventured such in Study 4 by comparing the P O N D and the NPI head to head in predicting a variety of criteria, among them, peer ratings. The results were amazingly similar given the dramatically different routes to scale construction and the different facets that emerged. Sometimes the P O N D outperformed the NPI and sometimes the reverse, but the differences were largely trivial. Thus while it might be true that the NPI packed the same degree of predictive punch, it is abundantly unclear how or why. In the POND's derivation, I started with the ideas of the very few writers who even addressed 'normal' narcissm, delineated their commonalities, and tried to measure them reliably and well. In contrast, Raskin and Terry (1988) set out from a  173  rather conceptually bereft position; while it is evident that the NPI measures  something that  may be related to normal narcissm, it is unclear what, substantively, that is. Beyond this theoretical advantage, there are other good reasons for preferring the P O N D to the NPI. First is administration time. The NPI requires that subjects read 80 statements, that is, both portions of 40 forced-choice items. In contrast, the present version of the P O N D requires reading only 52 statements (and see the note below on 'economy' of measurement). I noted in Study 4 that I have re-analyzed much of the data reported here using a 44item version of the P O N D ; as it turns out, reliabilty and validity of that instrument is not radically different from that of the P O N D 52. Specifically, the P O N D 44's full scale alpha is a respectable .89 (compared with .90 and .89 in  Studies 2 and 3, respectively). This alpha still  surpasses that of the NPI and thus researchers concerned with, for example, economy of administration time, may elect to use the short(er) form. The second reason for preferring the P O N D is that its scoring focuses on the core, that is, its first unrotated principal component, rather than the total instrument score. When the total score is used, a large degree of variance specific to the various subscales is lumped together with the common variance. Hence, the total P O N D score would predict a wide variety of criteria extending across an entire quadrant of the circumplex as well as into a third dimension. A l l analyses conducted with the NPI total score will similarly be too broad. Unfortunately, there is no recommendation from the NPI constructors to follow this procedure. Indeed, given that the NPI first component accounted for only 30 percent of the variance, its ability to represent the NPI core may be questionable.  This research might be considered indirect praise for the NPI. Despite the more careful instrument development, the P O N D did not improve on the predictive power of NPI. Thus the original instrument remains useful for predicting narcissistic behavior.  Future Directions One obvious future direction is determinining the fate of the Public Recognition (PR) scale. The PR scale is clearly the weakest of the five subscales; although perfectly adequate psychometrically (i.e., alphas ranging from the high 70s to low 80s), its nomological net is, as I have noted elsewhere, rather threadbare, and its order of loading on the core is typically the 4th or 5th of five. O n the other hand, the fact that its loading is reliable and strong suggests that P R does have some significant commonality with the other P O N D scales. What lies  beyond that commonality remains problematic. A s opposed to the other  P O N D scales, the P R shows minimal projection on the circumplex, suggesting that the scale is not especially interpersonal. It does show some low (i.e., 'teens and low 20s) correlations with anxiety and dependence and without question, its strongest correlate is Fenigstein et  al's  (1975) Public Self-Consciousness measure (i.e., r=.57; a close second is the r= .46 between PR and Cheek's (1989) Aspects of Identity Outer Orientation). Overall, the results suggest two points: First, that P R appears to fall 'outside' of interpersonal space, invoking as it does a generalized other who exists only in reference to his/her hypothetical reponse or reaction to the self as a stimulus object. Second is the related point that that P R appears to be the 'conceptual cousin' to Public Self-Consciousness (which also lies outside interpersonal space, Wiggins and Broughton, 1985), and accordingly,  175  appears to tap an awareness of the self as having an impact upon, but not in overt interaction with, others. So because of the delimited nature of the additional correlational information that it imparts—beyond consistency with Public Self-Consiousness—it may be that the Public Recognition aspect of the Profile of Narcissistic Dispositions is less 'important' substantively. Alternatively, it could be that an evident need/desire for public recognition is more  affective,  and thus not especially germane to personality assessment, strictly speaking. T o clarify its meaning, it would be useful to block subjects on PR, and applying a relevant experimental manipulation (e.g., where actor-focused attention could be supplied or withdrawn). In any case, its reliable marker status on the core suggests that it is relevant to narcissism; its evident paucity of correlates may be alleviated by a more judicious selection of potentially relevant instruments. There are a number of other directions in which POND-related research could move. Given that the P O N D is more substantively grounded than the NPI, and that its scoring is focused on the common variance (the P O N D core), then there is a need for replicating or at least reconsidering the research done with the NPI. In particular, use of the full-scale NPI, rather than just the core, may have yielded too broad a band of correlates. Determining the causal relations among the various facets of normal narcissism, as measured by the five subscales, would require different methodology from those utilized here, specifically, methodologies that permit causal analyses. Experimental manipulations in the laboratory and causal analyses of longitudinal data are two possibilities here.  ...  ,  176  Further research on the nature of the narcissist's impact upon the group—social or otherwise (e.g., at work)—is both theoretically and practically important. Reactions to the narcissist may be even more important than the narcissists' immediate behavior. Hogan's recent research on narcissism takes us in that direction, partly because Hogan's (1982) whole theory of personality is predicated upon predicting performance in work and social settings. The recent work by John and Robins (1992) also provides a model for studying narcissism in the group context. Given my findings suggesting that high P O N D scorers are low on agreeableness, collecting more detailed peer ratings including straight likeability (or lack thereof), as well as how they are perceived as influencing group cohesion would be useful. One could venture arguments that, in group settings—and depending upon group goals or 'atmosphere' (e.g., competitive or cooperative)—narcissists could be either facilitative or actively disruptive! Another remaining issue is the question of whether extremely high scorers show a narcissism that is pathologically rigid. Available tests of such rigidity could be examined vis a vis profiles of the top five (or so) percent of P O N D scorers. Should such inflexibility be observed, it is possible that despite its origin in a conception of secure narcissism, extremely high (or indeed, very low) P O N D scores could be associated with some degree of maladjustment, or, at the very least, a tendency to behave independently of situational demands or characteristics. This issue was, of course, partially addressed in the analyses of adjustment and extremity alluded to earlier. However, replication with a variety of dependent variables (laboratory criteria, clinical ratings) in a variety of contexts could only increase one's confidence in the conclusion that extreme scores do not yield pathology.  177  Another question to pursue is when is narcissism relevant and when is it not? Delineating a taxonomy of situational 'requirements', or general themes that are especially evocative of narcissistically-oriented tendencies could constitute another research direction. A n attempt to taxonomize actual situations predicted to 'pull' narcissistic behavior (e.g., being seated at the head of the table or with a prominent mirror placement) is likely to meet failure, if only as actor's perceptions of those situations could well differ from those of their creators. The latter would be setting themselves a near-Herculean task if only as the process of taxonomizing 'relevant' situations would be (as we have seen) hard-pressed to be theoretically informed. It's also possible that narcissism might be promoted in situations where needs for attention or approval are unlikely to be met or are perceived as threatened. Alternatively, the normal narcissist might even be  deferent in a situation where his/her own skills are not  perceived as relevant, required—or likely to be successful. Another potentially fruitful avenue, however, would emulate Christie and Geis' (1970) enumeration of situational variables facilitating the success of Machiavellians. What variables would influence the emergence of narcissistically-oriented behavior? Short face-toface encounters where social skills are critical are likely to benefit narcissists. A n y necessity for public display of one's attributes (acting, politics, etc) might also be facilitative. A n d situations where one is constantly subjected to criticism might provoke the narcissist's disagreeable side, as could situations in which evident needs for attention and/or approval are unlikely to be met. The likelihood and form of resultant interpersonal aggression could then be assessed. A n especially telling situation might be one invoking the necessity of  178  cooperation (as opposed to competition) as requisite to success. I suspect that the average high P O N D scorer (particularly one scoring rather highly on Manipulativeness or Competitive Ambition) may not be especially skilled at this task. A n d given the high P O N D scorer's evident arrogance and 'unlikeability', at least two paradigms could be instructive here. The first would entail trying to identify—perhaps via some kind of observer rating 'content analysis'—what precisely is it that gives others the impression of arrogance or low agreeableness, and therefore (one might assume) low likeability. In other words, to what degree is narcissism in the eye of the beholder? A second useful paradigm is that used by Godfrey, Jones, and Lord (1986) in their work on ingratiation. They found that when instructed to do so, their subject/confederates could get other subjects to respect them, but they could not get those same others to like them. Future research might attempt to instruct previously-identified high P O N D scorers (typically viewed as unlikeable) to "make your partner like you". Could narcissists, with their attendant skills—perhaps bolstered by relevant instructions—get others to like them, if they so desired? Coda.  O n the nature of romance, I observed earlier that "the poets said it first".  With regard to narcissism, it appears that Freud (1911/1914) said it first when he alluded to "a primary and normal narcissism in everyone" (p.45, italics added). The P O N D , a psychometrically-grounded attempt to measure that narcissism, clearly confirms its existence, but indicates that it is reflected more in some individuals than others.  179  REFERENCES Akhtar, S. & Thomson, J.A. (1982). Overview: Narcissistic personality disorder. American Journal of Psychiatry,  139, 12-20.  Alden, L . E . , Wiggins, J.S., & Pincus, A . L . (1990). Construction of circumplex scales for the Inventory of Interpersonal Problems. Journal of Personality Assessment, 55, 521-536. Allport, G . W . (1937). Personality: A Psychological  Interpretation.  New York: Holt.  American Psychiatric Association (1980). Diagnostic and statistical manual of mental disorders. (3rd ed.). Washington, D.C.: Author. American Psychiatric Association (1987). Diagnostic and statistical manual of mental disorders.  (3rd ed., rev.). Washington, D.C.: Author.  American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders. (4th ed.). Washington, D.C.: Author. Ashby, H . U . , Lee, R.R., & Duke, E . H . (1979, August). A narcissistic personality MMPI  disorder  scale. Paper presented at the meetings of the American Psychological Association,  New York. Bern, S.L. (1974). The Measurement of psychological androgyny. Journal of Consulting Clinical  Psychology,  42, 155-162.  Benjamin, L . (1987). Use of the S A S B in treatment plans: Narcissism. Journal of Disorders,  and  Personality  1, 43-70.  Birtchnell, J. (1984). Dependence and its relationship to depression. British Journal of Medical Psychology,  57, 215-225.  Block, J. (1971). Lives through time. Berkeley, C A : Bancroft Books. Block, J. (1989). Revised Ego-Resiliency  scale. Unpublished manuscript, University of  California, Berkeley, C A . Blum, G.S. (1950). Manual for the Blacky Pictures Test. New York: The Psychological Corporation. Briggs, S.R., & Cheek, J . M . (1986). The role of factor analysis in the development and evaluation of personality scales. Journal of Personality,  54, 106-148.  Buss, D . M . (1988).Evolution of human intrasexual competition: Tactics of mate attraction. Journal of Personality  and Social Psychology,  54, 616-628.  Buss, D . M . , & Chiodo, L . M . (1991). Narcissistic acts in everyday life. Journal of Personality,  59,179-215.  180  Buss, D . M . , Gomes, M . , Higgins, D.S. & Lauterbach, K. (1987). Tactics of manipulation. Journal of Personality  and Social Psychology,  52, 1219-1229.  Campbell, M . M . (1959). The Primary dimensions of item ratings on scales designed to measure 24 of Murray's Manifest Needs. Unpublished manuscript, University of Washington, Seattle. Carroll, J.B. (1961). The nature of the data, or how to choose a correlation coefficient. 26, 347-372.  Psychometrika,  Carson, R . C . (1969). Interaction  concepts of personality.  Chicago, IL: Aldine.  Carver, C.S. (1989). How should multifaceted personality constructs be tested? Issues illustrated by self monitoring, attributional style, and hardiness. Journal of and Social Psychology,  Personality  56, 577-585.  Carver, C.S. & Schier, M.I. (1981). Attention and self-regulation:  A Construct theory  approach to human behavior. New York: Springer Verlag. Cattell, R.B'. (1957). Personality and motivation structure and measurement. New York: World Book. Cattell, R . B . (1966). The scree test for the number of factors. Multivariate Research,  Behavioral  7,245-276.  Cheek, J. (1983). The revised Shyness scale. Unpublished manuscript. Wellesley College, Wellesley, M A . Cheek, J. (1989). Identity orientations and self-interpretation. In D . M . Buss & N . Cantor (Eds.) Personality psychology: Recent trends and emerging directions,  (pp. 275-285)  New York: Springer-Verlag. Christie, R. & Geis, F . L . (1970). Studies in Machiavellianism.  New York: Academic Press.  Comrey, A . L . (1973). A First course in factor, analysis. New York:Academic Press. Costa, P.T., Jr. & McCrae, R.R. (1989). The NEO Personality Inventory Manual. Odessa, F L : Psychological Assesssment Resources. Cronbach, L . J . (1950). Further evidence on response sets and test design. Educational Psychological  and  Measurement, 10, 3-31.  Cronbach, L . J . (1951). Coefficient alpha and the internal structure of tests.  Psychometrika,  16, 297-334. Edwards, A . L . (1970). The Measurement of personality York: Holt, Rinehart and Winston.  traits by scales and inventories.  New  181  Ellenberger, M . F . (1970). The Discovery of the unconscious:  The History and evolution of  New York: Basic Books.  dynamic psychiatry.  Ellis, H . (1898). Auto-eroticism: A psychological study. Alienist and Neurologist,  19, 260-  299. Emmons, R . A . (1981). Relationship between narcissism and sensation seeking. Psychological  Reports, 48, 247-250.  Emmons, R . A . (1984). Factor analysis and construct validity of the Narcissistic Personality Inventory. Journal of Personality Assessment, 48, 291-300. Emmons, R . A . (1987). Narcissism: Theory and measurement. Journal of Personality Social Psychology,  and  52, 11-17.  Eysenck, H.J., & Eysenck, S.B.G. (1969). Personality structure and measurement. San Diego, C A : Knapp. Fenigstein, A . , Schier, M . F . , & Buss, A . H . (1975). Public and private self-consciousness: Assessment and theory. Journal of Consulting and Clinical Psychology,  43, 522-527.  Ferguson, L . W . (1941). A study of the Likert technique of attitude scale construction. Journal of Social Psychology,  13, 51-57.  Fischer, C H . (1984, April). Correlates of subclinical  narcissism  in college males and  females. Paper presented at the meeting of the Southern Society for Philosophy and Psychology, Columbia, S C . Franken, R . E . , Gibson, K., & Rowland, G . L . (1989). Sensation seeking and feelings about the forced-choice format. Personality and Individual Differences,  10, 337-339.  Freud, S. (1911/1914). O n Narcissism: an Introduction. In J. Riviere (Trans.) Collected papers of Sigmund Freud (Vol. IV, pps. 30-59). London: Hogarth Press. Freud, S. (1922). Two encyclopaedia articles: (B) The Libido theory. In J. Strachey (Ed.) Collected papers of Sigmund Freud (Vol. V, pp. 131-135). London: Hogarth Press. Freud, S. (1931). Libidinal types. In J. Strachey (Ed.). Collected papers of Sigmund Freud (Vol V, pp. 247-251) London: Hogarth Press. Geis, F . L . & Christie, R. (1970). Overview of experimental research. In R. Christie & F . L . Geis. Studies in Machiavellianism.  New York: Academic Press.  Gifford, R. (1991). Mapping nonverbal behavior on the Interpersonal Circle. Journal of Personality  and Social Psychology,  61, 279-288.  182  Gifford, R., & O'Connor, B . (1987). The interpersonal circumplex as a behavior map. Journal of Personality and Social Psychology, 52, 1019-1026. Glass, G . V . , & Stanley, J.C. (1970). Statistical methods in psychology  and  education.  Englewood Cliffs, N J : Prentice-Hall. Gough, H . G . (1987). California  Psychological Inventory: Administrator's  guide. Palo Alto,  C A : Consulting Psychologists Press. Gough, H . G . & Heilbrun, A . B . , Jr. (1965/1980). The Adjective Check List Manual. Palo Alto, C A : Consulting Psychologists Press. Gough, H . G . (1957/1987). Manual, California  Psychological  Inventory. Palo Alto, C A :  Consulting Psychologists Press. Gough, H . G . (1965). Conceptual analysis of psychological test scores and other diagnostic variables. Journal of Abnormal Psychology,  70, 294-302.  Guilford, J.P. (1954). Psychometric methods. New York: McGraw-Hill. Gullikson, M . (1950). Theory of mental tests. New York: Wiley. Gurtman, M . B . , Fernandez, S.J., & Phillips, L . A . (1991, August). Narcissism from the viewpoint of the interpersonal problem circumplex. In L . Alden (chair). Assessment of interpersonal problems: Implications for treatment and research. Symposium conducted at the meeting of the American Psychological Association, San Francisco, C A . Guttman, L . (1954). Some necessary conditions for common factor analysis.  Psychometrika,  19, 149-161. Hakstian, A . R . & Bay, K.S. (1972). User's manual to accompany the Alberta General Analysis Program (AGFAP).  Factor  Research and Independent Reports, Division of Educational  Research, University of Alberta, Edmonton, Alberta. Hakstian, A . R . , Rogers, W . T . & Cattell, R.B. (1982). The Behavior of number-of-factors rules with simulated data. Multivariate Behavior Research, 17, 193-219. Hirschfeld, R . M . A . , Klerman, G . L . , Gough, H . G . , Barrett, J., Korchin,S.J. & Chodoff, P. (1977). A Measure of interpersonal dependency. Journal of Personality Assessment, 41, 610-621. Hogan, R.T. (1983). A socioanalytic theory of personality. In M . M . Page (Ed.) Nebraska Symposium on Motivation 1982 (pp. 55-89). Lincoln: University of Nebraska Press.  183  Hogan, R.T., Carpenter, B . N . , Briggs, S.R. & Hansson, R.O. (1984). Personnel assessment in personnel selection. In J.J. Bernardin & D . A . Bownas (Eds.) Personality  assessment in  (pps. 221-252). New York: Praeger.  organizations,  Hogan, J.C. & Hogan, R.T. (1986). Hogan Personality Inventory manual. Hogan personnel selection series, Minneapolis, M N : National Computer Systems. Horney, K. (1939). New ways in psychoanalysis.  New York: Norton.  Jackson, D. & Messick, S. (1967) (Eds.) Problems in human assessment. New York: M c G r a w Hill. Jahoda, M . , Deutsch, M . , & Cook, S.W. (1951) (Eds.) Research methods in social  relations.  New York: Dryden Press. John, O.P. & Robins, R.W. (1993). Accuracy and bias in self-perception: The role of individual differences in self-enhancement and narcissism. Journal of Personality Social Psychology,  and  66, .206-219.  Jones, E . E . & Pittman, T.S. (1982). Toward a general theory of straategic self-presentation. In J. Suls (Ed.), Psychological  perspectives on the self (Vol.1).  Hillsdale, N J : Erlbaum.  Jones, R.R. (1968). Differences in response consistency and subjects' preferences for three personality inventory response formats. Proceedings the American Psychological  Association,  of the 76th Annual Convention  3, 247-248.  Kaiser, H . F . (1974). A n index of factorial simplicity. Psychometrika, Kelly, G . (1955). The Psychology  of  39, 31-36.  of personal constructs. (Vols. 1 & 2). New York: Norton.  Kernberg, O. (1975). Borderline conditions and pathological  narcissism.  New York: Jason  Aronson. Kiesler, D.J. (1983). The 1982 Interpersonal circle: A Taxonomy for complementarity in human transactions. Psychological  Review, 90, 185-214.  Kohut, H . (1968). The Psychoanalytic treatment of narcissistic personality disorders. Psychoanalytic  Study of the Child, 23, 86-113.  Kohut, H . (1971). The analysis of the self. New York: International Universities Press. Komorita, S.S., & Graham, W . K . (1965). Number of scale points and the reliability of scales. Educational  and Psychological  Measurement, 4, 987'-995.  LaForge, R. (1977). Using the ICE (Unpublished; available from R. LaForge, M i l l Valley, CA).  184  LaForge, R., & Suczek (1955). The Interpersonal dimension of personality: III, an Interpersonal check list. Journal of Personality, Lanyon, R.I. (1973). Psychological  24, 94-112.  Screening Inventory manual. Goshen, New York:  Research Psychologists Press. Lasch, C L . (1979). The culture of narcissism. New York: Norton. LaVopa, L . A . (1981). Relationship  between narcissism and Machiavellianism.  Unpublished  paper, University of Southern Maine, Portland. Leary, T. (1957). Interpersonal  New York: Ronald Press.  diagnosis of personality.  Lissitz, R.W., & Green, S.B. (1975). Effect of the number of scale points on reliability: A Monte Carlo approach. Journal of Applied Psychology,  60, 10-13.  Livesley, W . J . (1987). Theoretical and empirical issues in the selection of criteria to diagnose Personality Disorders. Journal of Personality,  I, 88-94.  Loevinger, J. (1957). Objective tests as instruments of psychological theory.  Psychological  Reports, 3, 635-694. (monograph No. 9). Lord, F . M . , & Novick, M . R . (1968). Statistical theories of mental test scores. Menlo Park, C A : Addison-Wesley. Machover, S. (1955). A Compulsive personality with psychosomatic reactions. In A . Burton and R . E . Harris (Eds.). Clinical studies of personality:  Vol. II of case histories in clinical  and abnormal psychology (pp. 21-48). New York: Harper & Brothers. Margolis, H . D . , & Thomas, V . (1980). The measurement of narcissism and without behavioral and emotional disabilities.  in adolescents with  Unpublished Master's thesis, United  States International University, San Diego, C A . Mattell, M . S . , & Jacoby, J. (1971). Is there an optimal number of alternatives for Likert scale items? Study 1: Reliability and Validity. Educational  and Psychological  Measurement,  31, 651-61 A. Mayman, M . (1955). A Narcissistic character. In A . Burton and R . E . Harris (Eds.). studies of personality:  Clinical  Vol. II of case histories in clinical and abnormal psychology (pp.  307-321). New York: Harper & Brothers. McCrae, R.R. & Costa, P.T.Jr. (1987). Validation of the five factor model of personality across instruments and observers. Journal of Personality and Social Psychology,  52, 81-  90. McNemar, Q. (1969; 4th Ed.). Psychological  Statistics. New York: John Wiley and Sons.  185  Millon, T. (1969). Modern psychopathology.  Philadelphia, P A : Saunders.  Millon, T. (1981). Disorders of personality. Millon, T . (1982). Millon  New York: Wiley.  Clinical Multiaxial  Inventory manual (3rd ed.). Minneapolis, M N :  National Computer Systems. Morey, L . , Waugh, M . H . , & Blashfield, R . K (1985). M M P I scales for DSM-III personality disorders: Their derivation and correlates. Journal of Personality Assessment, 49, 245251. Mullahy, P. (1952, Ed.). The contributions  of Harry Stack Sullivan. New York: Hermitage  House. Mullins, L.S., & Kopelman, R . E . (1988). Toward an assessment of the construct validity of four measures of narcissism. Journal of Personality Assessment, 52, 610-625. Murray, H . A . (1938). Explorations  in personality.  New York: Oxford University Press.  Murray, H . A . (1955). American Icarus. In A . Burton and R . E . Harris (Eds.), Clinical of personality:  studies  Volume II of case histories in clinical and abnormal psychology (pp. 615-  641). New York: Harper & Brothers. Nacke, P. (1899). Die sexuellen perversitaten in der irrenanstalt. Psychiatrische Neurologische  en  Bladen, 3, 20-30.  Nunnally,J. (1978). Psychometric  theory. New York: McGraw-Hill.  Osgood,C.E., Suci, G.J., & Tannenbaum, P . H . (1957). The Measurement of meaning. Urbana, IL: University of Illinois Press. Oswald, W . T . & Velicer, W . F . (1980). Item format and the structure of the Eysenck Personality Inventory: A Replication. Journal of Personality Assessment, 44, 283-288. Ozer, D.J., & Buss, D . M . (1991). Two views of behavior: Agreement and disagreement among marital partners. Paddock, J.R. & Nowicki, S. (1986). A n examination of the Leary circumplex through the Interpersonal Check List. Journal of Research in Personality,  20, 107-144.  Paulhus, D . L . (1984). Two-component models of socially desirable responding. Journal of Personality  and Social Psychology,  46, 598-609.  Paulhus, D . L . (1986). Self-deception and impression management in test responses. In A . Angleitner & J.S. Wiggins (Eds.), Personality assessment via questionnaire 165). New York: Springer-Verlag.  (pp. 143-  186  Paulhus, D . L . (1991). Measurement and control of response bias. In J.P. Robinson, P.R. Shaver, & L.S. Wrightsman (Eds.), Measures of personality attitudes.(pp-17-59).  and social  psychological  New York: Academic Press.  Paulhus, D . L . , & Bruce, M . N . (1992). The effect of acquaintanceship on the validity of personality impressions. Journal of Personality and Social Psychology,  63, 816-824.  Paulhus, D . L . , Fridhandler, B., & Hayes, S. (in press). Psychological defense: Contemporary theory and research. In J.Johnson, R. Hogan & S.R. Briggs (Eds.). Handbook Personality  Psychology.  of  New York: Academic Press.  Paulhus, D . L . & Martin, C L . (1988). Functional flexibility: A new approach to interpersonal flexibility. Journal of Personality and Social Psychology,  55, 88-101.  Peabody, D. (1962). Two components in bipolar scales: Direction and extremeness. Psychological  Review, 69, 65-73.  Raskin, R . N . (1980). Narcissism and creativity: Are they related? Psychological  Reports, 46,  55-60. Raskin, R . N . (1988). Narcissistic Personality Inventory: Psychometric information. (Unpublished data; cited in Smith (1990); see below). Raskin, R . N . , & Hall, C S . (1979). A narcissistic personality inventory.  Psychological  Reports, 45, 590-. Raskin, R . N . , & Hall, C S . (1981). The narcissistic personality inventory: Alternative form reliability and further evidence of construct validity. Journal of Personality  Assessment,  45, 159-16. Raskin, R . N . , Novacek, J. & Hogan, R.T. (1991a). Narcissism, self-esteem and defensive self-enhancement. Journal of Personality,  59, 19-38.  Raskin, R . N . , Novacek, J. & Hogan, R.T. (1991b). Narcissistic self-esteem management. Journal of Personality  and Social Psychology,  54, 911-918.  Raskin, R . N . , & Shaw, R. (1988). Narcissism and the use of personal pronouns. Journal of Personality,  56, 393-404.  Raskin, R . N . , & Terry, H . (1988). A Principal-components analysis of the Narcissistic Personality Inventory and further evidence of its construct validity. Journal of and Social Psychology,  Personality  54, 980-902.  Reich, W . (1938). Some circumscribed character forms. Character analysis. New York: Farrar, Straus and Giroux.  187  Rosenberg, M . (1965). Society and the adolescent self-image. Princeton, N J : Princeton University Press. Schafer, R. (1948). The Clinical application of psychological  tests: Diagnostic  summaries  and case studies. New York: International Universities Press. Serkownek, K. (1975). Subscales for scales 5 and _ of the Minnesota Personality  Multiphasic  Inventory, (unpublished; cited in Raskin and Terry, 1988).  Smith, B . M . (1990). The measurement of narcissism in Asian, Caucasian & Hispanic American women. Psychological  Reports, 67, 779-785.  Smith, M . B . , & Anderson, J.W. (1989). Obituary: Henry A . Murray. American  Psychologist,  44, 1153-1154. Solomon, R.S. (1982). Validity of the M M P I narcissistic personality disorder scales. Psychological  Reports, 50, 463-466.  Spielberger, C D . , Gorsuch, R . L . , & Luschene, R . E . (1970). Manual for the  State-Trait  Anxiety Inventory. Palo Alto, C A : Consulting Psychologists Press. Spielberger, C D . (1979). Preliminary Manual for the State-Trait Personality  Inventory  (STPI). Tampa: University of South Florida Human Resources Institute. Stern, G . G . (1970). People in context: Measuring person-environment  congruence  in  education and industry. New York: Wiley. Strack, S. (1987). Development and validation of an adjective checklist to assess the M i l l o n personality types in a normal population. Journal of Personality Assessment, 51, 572-587. Strong, E . K . , Jr.(1959). Strong Vocational Interest Blank. Palo Alto, C A : Consulting Psychologists Press. Sullivan, H.S. (1949). Psychiatry: Introduction Sullivan, H.S. (1953). The Interpersonal  to the study of interpersonal  theory of psychiatry.  relations.  New York: Norton.  Taylor, S.E. & Brown, J.D. (1988). Illusion and well-being: A Social psychological perspective on mental health. Psychological  Bulletin, 103, 193-210.  Trapnell, P.D. (1991). The Reflection and Rumination  Questionnaire.  Unpublished  manuscript, The University of British Columbia, Vancouver, B . C . Trapnell, P.D. & Wiggins, J.S. (1989). Extension of the Interpersonal Adjective scales to include the B i g Five dimensions of personality. Journal of Personality Psychology,  59, 781-790.  and Social  188  Velicer, W . F . , DiClementi, C . & Corriveau, D.P. (1979). Item format and the structure of the Personal Orientation Inventory. Paper presented at the meeting of the Eastern Psychological Association, Philadelphia, P A , April, 1979. Velicer, W . F . & Stevenson, J.F. (1978). The relation between item format and the structure of the Eysenck Personality Inventory. Applied Psychological  Measurement, 2, 293-304.  Velicer, W . F . (1976). Determining the number of components from the matrix of partial correlations. Psychometrika,  41, 321-327'.  Wallach, M . A . & Wallach,L. (1983). Psychology's  sanction for selfishness: The Error of  egoism in theory and therapy. San Francisco: Freeman. Weinberger, D . A . (1989). Social-emotional adjustment in older children and adults: I. Psychometric properties of the Weinberger Adjustment Inventory. Weinberger, D . A . , & Schwartz, G . E . (1990). Distress and restraint as superordinate dimensions of self-reported adjustment: A typological perspective. Journal of 58, 381-417.  Personality,  West, S.G. & Finch, J.F. (in press). Measurement and analysis issues in the investigation of personality structure. In R.T. Hogan, J. Johnson and S. Briggs (Eds.) Handbook Personality  Psychology.  of  New York: Academic Press.  Wetzler, S. (1990). The Millon Clinical Multiaxial Inventory ( M C M I ) : A Review. Journal of Personality Assessment, 55, 445-464. Wiggins, J.S. (1965) Interpersonal diagnosis of personality. In O . K . Buros (Ed.) The Sixth mental measurements yearbook, (pp.451-453). Highland Park, N J : Gryphon Press. Wiggins, J.S. (1966). Substantive dimensions of self-report in the M M P I item pool. Psychological  Monographs,  80, No.22, Whole No.630.  Wiggins, J.S. (1973). Personality and prediction.  Reading, M A : Addison Wesley.  Wiggins, J.S. (1982). Circumplex models of interpersonal behavior in clinical psychology. In P.S. Kendall & J.N. Butcher (Eds.), Handbook of research methods in clinical psychology, (pp. 183-221). New York: Wiley. Wiggins, J.S. (1985). Interpersonal circumplex models: 1948-1983 (commentary). Journal of Personality Assessment, 49, 626-631. Wiggins, J.S. (1995). The Interpersonal  Adjective Scales (IAS) professional  F L : Psychological Assessment Resources.  manual. Lutz,  189  Wiggins, J.S., & Broughton, R. (1985). The Interpersonal Circle: a structural model for the integration of personality research. In R.T. Hogan & W . H . Jones (Eds.) Perspectives in personality.  (Vol. 1, pp. 1-47). Greewich, C N : JAI Press, Inc.  Wiggins, J.S. & Broughton, R. (1991) A Geometric taxonomy of personality scales. European Journal of Personality,  5, 343-365.  Wiggins, J.S. & Trapnell, P.D. (in press) Personality structure: The Return of the B i g Five. In (Eds.) R.T. Hogan, J.A. Johnson & S.R.Briggs. Handbook of personality  assessment.  Orlando, F L : Academic Press. Wiggins, J.S. & Pincus, A . L . (1989). Conceptions of personality disorders and dimensions of personality. Psychological  Assessment: A Journal of Consulting and Clinical  Psychology,  1, 305-316. Wiggins, J.S., Trapnell, P. & Phillips, N . (1988). Psychometric and geometric characteristics of the revised Interpersonal Adjective Scales (IAS-R) Multivariate  Behavioral  Research,  25,517-530. Wink, P. (1992). Three types of narcissism in women from college to mid-life. Journal of Personality,  60,1-30.  Wink, P. & Gough, H . (1990). New narcissism scales for the CPI and the M M P I . Journal of Personality Assessment, 54, 446-462.  190  List of Tables  Study 1. Table l a : Principal components analysis of the matrix of NPI inter-item tetrachoric correlations. Table l b : Intercorrelations among the NPI (tetrachoric-based) components. Table 2a: Principal components analysis of the matrix of NPI inter-item phi coefficients. Table 2b: Intercorrelations among the NPI (phi-based) components. Table 3:  Principal components analysis of the 40 Likert-rated narcissistic alternatives of the NPI dyads  Table 4a:  Normative statistics and reliabilities.  Table 4b:  Increments in alpha reliabilities over those obtained by Raskin and Terry (1988).  Study 2. Table 5:  Principal components analysis of the 65 P O N D - P items: Oblique rotation.  Table 6:  Final 52 P O N D scale items.  Table 7:  Intercorrelations, means, standard deviations, age and gender correlations and coefficients alpha.  Table 8:  Correlations of the P O N D scales with various self-report measures.  Study 3 . Table 9:  Principal components analysis of the P O N D items: Oblique rotation.  Table 10: Descriptive statistics and principal components analysis of P O N D subscales: Unrotated. Table 11:  Correlations of the P O N D scales with various self-report measures.  Study 4. Table 12:  P O N D scale intercorrelations, means, standard deviations, and coefficients alpha.  Table 13:  Correlations of P O N D scales with selected self-report measures.  Table 14:  Correlations of P O N D scales with peer ratings of narcissism.  Table 15a: Correlations of P O N D scales with performance-related ratings. Table 15b: Correlations of P O N D scales with various interpersonal ratings.  191  Table la: Study 1: Principal components analysis of the matrix of NPI inter-item tetrachoric correlations.  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40  Item I would prefer to be a leader. I see myself as a good leader. People always seem to recognize my authority. I like having authority over other people. I am a born leader. I have a strong will to power. I will be a success. I am assertive. I have a natural talent for influencing people. If I ruled the world, it would be a much better place. I want to amount to something in the eyes of the world. I like to look at my body. I like to display my body. I like to look at myself in the mirror. I'll never be satisfied until I get all that I deserve. I expect a great deal from other people. I insist upon getting the respect that's due to me. I am going to be a great person. I am an extraordinary person. I think that I'm a special person. I always know what I'm doing. I can live my life any way I want to. I can read people like a book. I'm more capable than others. I rarely depend on anyone else to get things done. I like to take responsibility for making decisions. I can talk my way out of anything. I find it easy to manipulate people. I would do almost anything on a dare. Modesty doesn't become me. I can make anyone believe anything. I know that I am good-looking because everyone keeps telling me so. I like to be complimented. I like to be the center of attention. I really like to be the center of attention. I get upset when people don't notice how I look when I go out in public. I like to start new fads and fashions. I'm apt to show off if I get the chance. Everybody likes to hear my stories. I wish someone would someday write my biography.  R&T AUT AUT AUT AUT  I .86 .86 .83 .77  II  III  IV  V  .41  .57 .57 .45 .58 .42  .41 .41 .40 .46  AUT ENT AUT AUT AUT ENT ENT  .67 .65 .59 .57 .47  .56 .50 .42  VAN VAN  .50 .45 .40  .47 .57  .41  .47  .76 .63 .59  ENT ENT ENT SFS  .46 .46  SUP  .79 .75 .63  .45 .44  .52 .44  SFS SFS EXP SFS  .44  SFS SFS EXP  .62 .61 .55 .53 .48 .48  .47  .76 .66 .57 .54 .44  EXP EXH EXH EXP SUP  EXH  .55  .46 .45  .49  EXH EXH  .45 .78 .60  SUP EXH EXH  VII  .86 .81 .79  VAN  SUP  VI  .42  .45  EXP SUP  Note. R & Tcolumn indicates original Raskin & Terry subscale labels. They are as follows: AUT = Authority; EXH = Exhibitionism; ENT = Entitlement; EXP = Exploitativeness; SFS = Self-sufficiency; SUP = Superiority; VAN = Vanity. Loadings < .39 are not shown. Eigen values and their associated variances accounted for are as follows: I. 11.56, 5.71%; II. 2.95, 2.66%; III. 2.57, 2.96%; IV. 1.89, 3.88%; V. 1.76, 2.87%; VI. 1.48, 1.96%; VII. 1.32, 3.48%. N = 654.  .47 .85 .83 .69 .57 .54 .52 .51  192  Table lb: Study 1: Intercorrelations among the NPI (tetrachoric-based) components.  1.  2.  3.  4.  5.  6.  1.  1.0  2.  .14  1.0  3.  .12  .28  1.0  4.  .58  .14  .02  1.0  5.  .42  .06  .02  .44  1.0  6.  .08  .17  .14  .13  -.01  1.0  7.  .44  .39  .12  .47  .29  .04  Note. N = 654.  193  Table 2a: Study 1: Principal components analysis of the matrix of NPI inter-item phi coefficients.  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40  Item I would prefer to be a leader. I like having authority over other people. I see myself as a good leader. People always seem to recognize my authority. I have a strong will to power. I will be a success. If 1 ruled the world, it would be a much better place. I am a born leader. I am assertive. I want to amount to something in the eyes of the world. I like to look at my body. I like to display my body. I like to look at myself in the mirror. I'll never be satisfied until I get all that I deserve. I insist upon getting the respect that's due to me. I expect a great deal from other people. I am going to be a great person. I am an extraordinary person. I always know what I'm doing. 1 think that I'm a special person. 1 can live my life any way I want to. I can read people like a book. 1 rarely depend on anyone else to get things done. 1 like to take responsibility for making decisions. Modesty doesn't become me. I know that I am good-looking because everyone keeps telling me so. I'm apt to show off if I get the chance. I find it easy to manipulate people. I'm more capable than others. I would do almost anything on a dare. I can make anyone believe anything. 1 can talk my way out of anything. I have a natural talent for influencing people. I like to be the center of attention. I really like to be the center of attention. I get upset when people don't notice how I look when I go out in public. I like to start new fads and fashions. I wish someone would someday write my biography. I like to be complimented. Everybody likes to hear my stories.  R&T AUT AUT AUT AUT ENT AUT ENT AUT AUT ENT  I 77 70 69 69 61 53 50 47 47 42  II  III  VAN VAN  -35 37  ENT  33  36 68 63 54 53 54 42 42 38  SFS  36  SFS  37  SFS EXP  -33  SFS  63 48  EXH SUP EXH EXP SFS  32  EXH  40  47 47 43 40  EXP  48  EXP AUT  -33 41 37  52 51 39 76 75 55  EXH EXH EXH  38  EXH SUP SUP EXP  VII 36  36 39  63 50 48  ENT  SFS  VI  33  ENT  SUP  V  80 72 65  VAN  SUP  IV 24  -32 32  Note: R & T column indicates original Raskin & Terry item subscale. Scale abbreviations appear in note for Table la. Eigen values and their associated variances accounted for as follows: I. 7.25, 18.19%; II. 2.22, 5.5%; III. 1.95, 4.9%; IV. 1.59,4.0%; V. 1.49, 3.7%; VI. 1.27, 3.2%; VII. 1.22,3.1%. N=654.  47 43 40 35  194  Table 2 b : Study 1: Intercorrelations among the NPI (phi-based) components.  1.  2.  3.  4.  5.  6.  1.  1.00  2.  .12  1.00  3.  .07  .11  1.00  4.  .35  .10  -.04  1.00  5.  .21  .09  .05  .18  1.00  6.  .11  -.03  -.02  .08  .10  1.00  7.  .26  .24  .03  .24  .12  .09  Note. N = 654  7.  1.00  195  Table 3: Study 1: Principal components analysis of the matrix of the 40 Likert-rated narcissistic alternatives of the NPI dyads. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40  Item I would prefer to be a leader. I see myself as a good leader. I am a born leader. I really like to be the center of attention. I like having authority over other people. I like to be the center of attention. People always seem to recognize my authority. I am assertive. I like to take responsibility for making decisions. I have a strong will to power. I like to look at my body. I like to look at myself in the mirror. I like to display my body. I get upset when people don't notice how I look when I go out in public. I can talk my way out of anything. I find it easy to manipulate people. I have a natural talent for influencing people. I can make anyone believe anything. Everybody likes to hear my stories. I can read people like a book. Modesty doesn't become me. I always know what I'm doing. I rarely depend on anyone else to get things done. I know that I am good-looking because everyone keeps telling me so. I can live my life any way I want to. I would do almost anything on a dare. I like to start new fads and fashions. I wish someone would someday write my biography. I think that I'm a special person. I am going to be a great person. I am an extraordinary person. I will be a success. I'm more capable than others. If I ruled the world, it would be a much better place. I expect a great deal from other people. I insist upon getting the respect that's due to me. I'll never be satisfied until I get all that I deserve. I'm apt to show off if I get the chance. I want to amount to something in the eyes of the world. I like to be complimented.  R&T AUT AUT AUT EXH AUT EXH AUT  I .83 .82 • .76 .68 .65 .62 .60  AUT SFS  .59 .55  ENT VAN VAN VAN  .50  EXH EXP EXP AUT EXP EXP EXP EXH SFS SFS  II  III  IV  .37 .43  .35  .58 .59  .47  .47  .35  .33  .47  .41 .75 .74 .70 .59 .49 .48 .40  .45  .37 .33  .36  .65 .63 .51  SUP  .50  SFS EXH EXH  .46 .44  SUP SUP SFS SUP AUT SFS ENT  .35 .44 .34 .36 .37  .84 .80 .73  .35  .33 .37  .59 .56 .44  .43  .53 .41  .77 .75 .74 .71 .35 .42  .33 .39  .45 .44 .37  .67 .67  ENT ENT  .52  ENT EXH ENT  SUP  VII  .36  .51  .55  VI .37 .44 .47  .40 .34 .33  .35 .50  V  .36 .36  .34  .34 .37  .59 .41  .43 .41  .37  .41  Note: R & T's original component-scale designations are provided in the two previous tables.. Loadings < .33 are not shown. Eigen values and their associated variances accounted for are as follows: I. 10.6, 26%; II. 2.9, 7.2%; III. 2.1, 5.2%; IV. 1.7, 4.1%; V. 1.5, 3.8%; VI. 1.3, 3.3%; VII. 1.5, 3.2.9%. N = 654.  Table 4a. Study 1: Normative statistics and reliabilities. Raskin & Terry (1988)  Scale s  1 2 3 4 5 6 7 Full  Auth.  Exh.  Sup.  Ent.  Expl.  SSuf.  Van.  1  2  3  4  5  6  7  —  —  .42 .39 .34 .34 .39 .21 .58  —  .37 .34 .32 .19 .26 .51  .25 .20 .28 .31 .48  .29 .24 .14 .43  .25 .12 .41  .11 .40  .30  Exh. 2  Sup. 3  Ent. 4  Expl. 5  SSuf. 6  Van. 7  —  —  —  —  # items  M  5  2.17 1.74 1.36 1.40 1.69 1.30 1.08 6.66  Lambda D J  8 7 5 6 5 6 3 40  4.16 2.21 2.54 1.67 1.47 2.09 1.37 15.55  .73 .63 .54 .50 .52 .50 .64 .83  # items  M  8 7 5 6 5 6 3 40  3.97 2.34 2.04 2.00 1.613 2.165 1.138 15.27  2.42 1.87 1.42 1.58 1.32 1.58 1.05 7.66  .79 .68 .58 .57 .50 .55 .56 .87  # items  M  S.D.  Alpha  8 7 5 6 5 6 3 40  31.51 22.93 19.55 23.35 17.70 24.34 10.06 145.43  6.45 5.34 3.57 4.17 3.97 3.74 2.96 21.56  .88 .75 .65 .64 .72 .64 .80 .96  NPI-FC (JV=553)  Scale  Auth. 1  1 2 3 4 5 6 7 Full  .46 .41 .43 .47 .45 .19 .80  —  —  —  .48 .34 .41 .31 .34 .74  .35 .27 .41 .36 .69  .34 .28 .13 .63  .34 .13 .63  .14 .64  .43  Exh. 2  Sup. 3  Ent. 4  Expl. 5  SSuf. 6  Van. 7  —  —  —  —  Alpha  NPI-L (N = 586)  Scale  Auth. 1  1 2 3 4 5 6 7 Full  .50 .53 .58 .61 .60 .25 .79  —  —  .53 .53 .52 .30 .46 .77  —  .50 .41 .43 : .45 .76  —  .47 .45 .26 .73  —  .41 .27 .73  —•  .22 .64  —  .62  Note. Auth=authority; Exh=exhibition; Sup=superiority; Ent=entitlement; Expl=exploitativeness; Ssuf=selfsufficiency; Van=vanity.  197  Table 4b. Study 1: Increments in alpha reliabilities over those obtained by Raskin and Terry (1988).  NPI Forced-Choice Format R&T  Present Sample  Authority  .73  Exhibitionism  NPI Likert format Present Sample  .80  Increase over R&T +.07  .88  Increase over R&T +.15  .63  .69  +.06  .75  +.12  Superiority  .54  .57  +.03  .64  +.10  Entitlement  .50  .57  +.07  .64  +.14  Exploitativeness  .52  .50  -.02  .72  +.20  Self Sufficiency  .50  .55  +.05  .64  +.14  Vanity  .64  .66  +.02  .80  +.16  Full Scale  .83  .87  +.04  .92  +.09  Note. R & T=Raskin & Terry (1988).  198  Table 5:  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39  Study 2: Principal components analysis of the 65 POND-P items: Oblique rotation.  I probably wouldn't make a very good leader. (R) * I have a natural talent for influencing people. I am a born leader. I see myself as a good leader. I'm not sure that I'd make a very good leader. (R) * I am not good at influencing people. (R) * I am assertive. The thought of talking in front of a group terrifies me. (R) * I can usually talk my way out of anything. It's pretty easy for me to get other people to do what I want them to do. * I'm not always sure of myself when it comes to making decisions. (R) * I'm not very good at telling jokes or stories. (R) * I must admit, I'm a pretty fair talker. * I prefer to blend in with the crowd. (R) People instinctively trust my judgement. * People often tell me how great I look. * Sometimes I'm not always sure of what I'm doing. (R) I don't mind blending in with the crowd. (R) I always know what I'm doing. I sometimes give in, even if I know I'm right. (R) * I sometimes take advantage of other people. * Taking advantage of other people can never be justified. (R) * Honesty is always the best policy. (R) * I don't like it when I find myself manipulating people. (R) * If I have to, I'll lie to get out of a sticky situation. * If people let you manipulate them, it's their own damn fault. * I really believe it's a 'dog eat dog' world. * I don't like it, but sometimes I'm pretty selfish. * I really resent it if someone does better than I do. * I spend a lot of time on my appearance. * I get upset when people don't notice how I look when I go out in public. I like to dress in trendy fashions. * If I spend a lot of time getting dressed up to go out, it bugs me when people don't notice how I look. * I don't pay very much attention to new trends or fashions. (R) When I pass a mirror, I usually 'check myself out.' * It bothers me when my accomplishments aren't appreciated. * It's important to me to get the respect that I deserve. I work out mainly to look good. * I like to start new fads and fashions.  I .74 .74 .71 .69 .69 .66 .60 .59 .57 .56 .53 .52 .52 .50 .48 .46 .43 .37 .37 .36  II  III  V  .33  .40 .40  .34  .45  .36  .33 .33 .70 .62 .58 .56 .54 .52 .52 .49 .38 .68 .62 .60 .60  .34  IV  .58 .55 .52 .51 .47 .46  199 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65  I get uncomfortable when members of the opposite sex 'check me out.' (R) * I think I have a pretty good body. * I think that I am pretty good looking. * I like to look at my body. I like it when members of the opposite sex 'check me out.' * Compliments embarrass me. (R) I am an extraordinary person. I think that I'm a special person. When people compliment me. I sometimes get embarrassed. (R) I don't particularly like to show off my body. (R) I'm probably a bit of a flirt. * I feel superior to most people. I like to be complimented. I tend to look before I leap. (R) * My friends would probably describe me as competitive. * I'd like to be successful, but it's not a big deal to me. (R) * Being an authority doesn't mean that much to me. (R) * I can't stand losing at games. * I'm not particularly ambitious. (R) * Power for its own sake doesn't interest me. (R) * It's my ambition to somehow 'leave my mark' upon the world. * I'll do anything within reason to get ahead. * I believe in the survival of the fittest. * The thought of ruling the world frightens the hell out of me. (R) * I usually accept defeat gracefully. (R) * In an argument, I'll sometimes give in, even if I know that I'm right. (R) *  .35 39 .32  .35 .35  .33 .36 .38  .62 .61 60 .56 .53 .48 .48 .47 .43 .41 .41 .39 .38 .32  .34  .33 .39 .38 .38 .40 .34  Note. New POND-P items have an asterisk; others were selected from the NPI. Components were interpreted as follows: I. Assured Leadership; II. Manipulativeness; III. Public Self-consciousness/Need for Recognition; IV. Vain Exhibition; V. Competitive Ambition. ./V = 622.  .61 .58 .58 .56 .52 .52 .51 .49 .44 .41 .37 .37  200  Table 6:  Study 2: Final 52 P O N D scale items.  Assured Leadership 1  I probably wouldn't make a very good leader. (R)  2  I have a natural talent for influencing people.  3  I see myself as a good leader.  4  I am not good at influencing people. (R)  5  I am assertive.  6  The thought of talking in front of a group terrifies me. (R)  7  I can usually talk my way out of anything.  8  It's pretty easy for me to get other people to do what I want them to do.  9  I'm not always sure of myself when it comes to making decisions. (R)  10  I'm not very good at telling jokes or stories. (R)  11  I must admit, I'm a pretty fair talker.  12  People instinctively trust m y judgement. Manipulativeness  13  I sometimes take advantage of other people.  14  Taking advantage of other people can never be justified. (R)  15  I believe that honesty is always the best policy. (R)  16  I don't like it when I find myself manipulating people. (R)  17  If I have to, I'll lie to get out of a sticky situation.  18  If people let you manipulate them, it's their own damn fault.  19  I really believe it's a 'dog eat dog' world.  20  I don't like it, but sometimes I'm pretty selfish.  21  I really resent it if someone does better than I do. Public Recognition  22  I spend a lot of time on my appearance.  23  I get upset when people don't notice how I look when I go out in public.  24  I like to dress in trendy fashions.  25  If I spend a lot of time getting dressed up to go out, it bugs me when people don't notice how I look.  26  I don't pay very much attention to new trends or fashions. (R)  27  When I pass a mirror, I usually 'check myself out.'  28  It bothers me when people don't appreciate my accomplishments.  29  It's important to me to get the respect that I deserve.  30  I work out mainly to look good.  201 Vain Exhibition 31  I get uncomfortable when members of the opposite sex 'check me out.' (R)  32  I think I have a pretty good body.  33  I think that I'm pretty good looking.  34  I like to look at my body.  35  I like it when members of the opposite sex 'check me out.'  36  Compliments embarrass me. (R)  37  I am an extraordinary person.  38  I think that I'm a special person.  39  When people compliment me. I sometimes get embarrassed. (R)  40  I don't particularly like to show off my body. (R)  41  I'm probably a bit of a flirt. Competitive Ambition  42  M y friends would probably describe me as competitive.  43  I'd like to be successful, but it's not a big deal to me. (R)  44  Being an authority doesn't mean that much to me. (R)  45  I can't stand losing at games.  46  I'm not particularly ambitious. (R)  47  Power for its own sake doesn't interest me. (R)  48  It's my ambition to somehow 'leave my mark' upon the world.  49  I'll do anything within reason to get ahead.  50  I believe in the survival of the fittest.  51  The thought of ruling the world frightens the hell out of me. (R)  52  I usually accept defeat gracefully. (R)  Note. Approximately 80% of the POND items were newly-generated. N = 637  202  Table 7:  Study 2: POND scale intercorrelations, means, standard deviations, age and gender correlations and coefficients alpha.  2  1  1  1.00  2  .04  1.00  3  .08  .26  1.00  4  .48  .09  .28  1.00  5  .40  .43  .22  .29  Full Scale POND  .69  .52  .53  .68  First unrotated principal  3  4  Subscale  Mean  SD  # items  Age  Gender  Alpha  45.49  8.74  12  .03  -.06  .87  30.06  6.89  9  -.09  -.14  .76  33.63  6.41  9  -.18  .17  .79  40.87  7.45  11  .09  -.24  .81  1.00  40.74  7.90  11  -.14  -.20  .77  .74  190.7  23.6  52  -.07  -.15  .89  5  component of subscale  1. Assured Leadership  .72  2. Manipulativeness  .23  3. Public Recognition  .53  4. V a i n Exhibition  .75  5. Competitive Ambition  .75  correlations.  Variance accounted for by first component = 4 0 %  Note. 1 = Assured Leadership; 2 = Manipulativeness; 3 = Public Recognition; 4 = Vain Exhibition; 5 = Competitive Ambition. For Age and Gender correlations, N= 706. For all other correlations, N= 637.  203  Table 8.  Study 2: Correlations of the POND scales with various self-report measures.  POND  POND  Assured  Manipu-  Precog-  Core  Lead  lative  nition  Vain Ex  Compe titive  Circumplex Measures IAS-R Assured-Dominant  .43  .40  .67  .06  .04  .39  .39  IAS-R ArrogantCalculating IAS-R Warm-Agreeable  .50  .50  .28  .46  .13  .31  .44  -.32  -.33  .02  -.39  .11  .05  -.23  FFI Extraversion  .30  .40  .54  -.18  .10  .34  .12  FFI Agreeable  -.37  -.25  -.02  -.60  -.15  .00  -43  FFI Conscientious  .11  .21  .32  -.28  -.03  .17  .15  FFI Openness  -.13  -.11  .18  -.23  -.15  .13  -.11  FFI Neuroticism  -.22  -.25  -.53  .24  .26  -.34  -.06  Cheek Shyness  -.35  -.37  -.63  .14  .09  -.39  -.11  STPI Trait Anxiety  -.26  -.26  -.56  .21  .17  -.35  -.11  IIP Domineering  .42  .55  .14  .50  .26  .11  .39  IIP Vindictive  .18  .46  -.21  .52  .20  -.08  .28  IIP Cold  -.03  .30  -.31  .31  .10  -.22  .14  IIP Socially Avoidant  -.27  -.30  -.60  .18  .07  -.35  -.02  IIP Unassertive  -.41  -.44  -.65  .02  .09  -.30  -.31  IIP Exploitable  -.30  -.49  -.48  -.06  .16  -.17  -.30  IIP Overly-nurturant  -.14  -.20  -.26  -.08  .17  -.11  -.11  IIP Intrusive  .23  .36  -.01  .28  .34  .09  .11  Big Five Measures  Other Anxiety Measures  Interpersonal Problems  Note:  N = 637. POND Core is the first unrotated principal component of the subscale intercorrelations.FFI=Five Factor Inventory; IAS-R=Interpersonal Adjective Scales-Revised; STPI = State-Trait Personality Inventory; IIP = Inventory of Interpersonal Problems. All correlations > .14, p<.001, two-tailed.  204  Table 9.  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52  Study 3: Principal components analysis of the POND items: Oblique rotation.  I have a natural talent for influencing people. I see myself as a good leader. It's pretty easy for me to get other people to do what I want them to do. I'm not good at influencing people. (R) I'm probably not a very good leader. (R) I am assertive. I can talk my way out of anything. I must admit, I'm a pretty fair talker. I am an extraordinary.person. I'm a special person. People instinctively trust my judgement. I'm not very good at telling jokes or stories. (R) The thought of talking in front of a group terrifies me. (R) It's my ambition to somehow 'leave my mark' upon the world. I'm not always sure of myself when it comes to making decisions. (R) I sometimes take advantage of other people. Taking advantage of other people can never be justified. (R) If I have to, I'll lie to get out of a sticky situation. If people let you manipulate them, it's their own damn fault. I believe that honesty is always the best policy. (R) I really believe it's a'dog eat dog'world. I believe in the survival of the fittest. I'll do anything within reason to get ahead. I can't stand losing at games. I don't like it when I find myself manipulating people. (R) I usually accept defeat gracefully. (R) I really resent it if someone does better than I do. I don't like it, but sometimes I'm pretty selfish. I spend a lot of time on my appearance. I get upset when people don't notice how I look when I go out in public. If I spend a lot of time getting dressed up to go out, it bugs me when people don't notice how I look. When I pass a mirror, I usually 'check myself out.' I work out mainly to look good. I don't pay much attention to new trends or fashions. (R) It bothers me when my accomplishments aren't appreciated. I like to dress in trendy fashions. It's important to me to get the respect that I deserve. I get uncomfortable when members of the opposite sex 'check me out.' (R) When people compliment me. I sometimes get embarrassed. (R) Compliments embarrass me. (R) I don't particularly like to show off my body. (R) I think I have a pretty good body. I think that I am pretty good looking. I like it when members of the opposite sex 'check me out.' I like to look at my body. I must admit that I'm a bit of a flirt. Being an authority doesn't mean that much to me. (R) Power for its own sake doesn't interest me. (R) I'm not particularly ambitious. (R) My friends would probably describe me as competitive. The thought of ruling the world frightens the hell out of me. (R) I'd like to be successful, but it's not a big deal to me. (R)  I .79 .73 .73 .68 .68 .65 .62 .62 .60 .59 .54 .51 .48 .46 .41  II  III  IV  v  .39  .42  .37 .34  .46 .35  .34  .37 .40 .71 .63 .58 .58 .58 .57 .54 .52 .50 .50 .49 .47 .46  .35 .37  .33 .32 .42 .70 .66 .66 .60 .54 .49 .49 .44 .40  .41 .47 .33 .36 .33 .40 -.41 .45 .35 .40  .40  .38 .33 .70 .69 .69 .67 .63 .58 .57 .54 .50 .72 .68 .53 .47 .46 .41  205  Table 10. Study 3: Descriptive Statistics and Principal Components Analysis of POND subscales: Unrotated.  Sub  1  2  3  4  5  Mean  SD  No.  Alph  items  scales 1  1.00  2  .16  1.00  3  .04  .28  1.00  4  .49  .21  .31  1.00  5  .47  .52  .24  .34  Full  .70  .60  .49  .72  First Unrotated Principal  45.31  9.68  12  .88  29.88  6.70  9  .77  33.45  6.63  9  .77  40.11  8.48  11  .84  1.00  41.02  9.11  11  .82  .78  189.08  27.30  52  .90  Component of the 5 subscale  Assured Leadership  .67  Manipulativeness  .64  Public Recognition  .48  Vain Exhibition Competitive Ambition  intercorrelations  .71 .80  Variance accounted for by first component = 44.8% Note: I=Assured Leadership; II=Manipulativeness; III=Public Recognition; IV=Vain Exhibition; V=Competitive Ambition. N = 583.  206  Table 11: Study 3: Correlations of the P O N D scales with various self-report measures.  POND IAS-R Assured Dominant  .61  POND Core .48  A-Leader  Manip  Pub Recog  Vain Ex  .73  .23  .04  .44  Compete tive .48  IAS-R Arrogant-Calcul'g  .60  .55  .33  .60  .21  .33  .52  IAS-Warm-Agreeable  -.28  -.42  -.05  -.49  .03  -.07  -.35  FFI Extraversion  .37  .34  .53  -.09  .14  .39  .19  FFI Agreeableness  -.52  .-45  -.26  -.65  -.14  -.18  -.52  -.14  -.02  .23  .11  FFI Conscientious  .15  .09  .26  FFI Neuroticism  -.30  -.33  -.48  .07  .17  -.42  -.18  FFI Openness  .04  .11  .18  -.07  -.07  .10  -.05  Private Self-consciousness  .0.8  .08  .04  .11  .18  .02  -.01  Public Self-consciousness  .13  .16  -.15  .13  .57  -.01  .07  Social Anxiety  -.41  -.34  -.65  .02  .02  -.45  -.23  Ego Resilience  .31  .33  .51  -.12  .04  .37  .12  BSI total score  -.10  -.10  -.25  .18  .12  -.26  .04  Other Measures  Weinberger Adjustment Inventory Considerateness  -.35  -.35  -.01  -.55  -.16  -.12  -.37  Aggressiveness  .49  .44  .25  .48  .22  .23  .49  Irresponsibility  .35  .33  .04  .52  .23  .17  .26  Impulsivity  .26  .33  .11  .23  .16  .16  .22  Anxiety  -.21  -.28  -.43  .13  .24  -.35  -.12  Depression  -.25  -.31  -.48  .18  .16  -.43  -.11  Low Self-Esteem  -.45  -.43  -.63  .07  .09  -.59  -.25  Weil-Being  .23  .26  .51  -.25  -.04  .38  .03  Denial  .36  .36  .45  -.04  -.11  .44  .29  Repression  -.10  -.04  .15  -.38  -.26  .06  -.11  Note.  FFI=Five Factor Inventory; IAS-R=Interpersonal Adjective Scales-Revised; BSI = Basic Symptom Inventory. All correlations > .14, /?<.001. N=425.  207  Table 12: Study 4: P O N D scale intercorrelations, means, standard deviations, and coefficients alpha.  P O N D Subscales  1  2  3  4  5  Mean  SD  Alpha  46.17  7.64  .83  29.22  6.24  .76  33.91  6.92  .81  40.63  7.86  .84  1. Assured Leader  1.00  2. Manipulative  .01  1.00  3. Public Recognition  .01  .23  1.00  4. Vanity  .43  .11  .36  1.00  5. Competitive Ambition  .31  .52  .31  .35  1.00  36.25  8.40  .81  Full Scale P O N D  .57  .55  .58  .72  .79  '186.21  23.83  .89  Note.  N=S2  208 c  a *-> o  o  rt !/) (D _>  .2  l |  ro  O  O ro  VO  5  2  o  oo O  ON  CN  O  CN  .  ^° e  .  O CN  s 'S3 >  ro  m O  m  in CM  o  rn  o  m  VO  CN  ro CN  a> u co  83  O  A  u  j"  •5  C  II Pi ^* ob P  CO  3  S2co a  OO  3 o 3 <u  i-H  O  m —'  vo  CO  o  CO  O  O  oo  vo  —i  ON  CN  < X '"l ^ vb" of  c tn U  o  a S-  rt  *—i  VO  oo m  oo wo  m CN  ON  CN  o  CN  CN O  VO CO  u co  T3  <D T3 rt co CD  o  3  r-CN  ON  m  CN m  m  in m  •*  ON  O  c  O O  9  oo m  U  r-~ x  Q Z  S  u  *S«  co  Vi  o ^  rt^;  >—I  O  S>in .S o •a v U CN  1/3  "a*  if  in CN  (N  m m  in  CO  ON  m CN  o  vo m  m  in — •V  u  § ft.  ON  co  ro in  CO CN  o  O  vo  CN  CO  vo m  oo m  CN  °  fx.  « II « ^  rt ^ Q _o  «•§ "a! Ui i*  •St-  Tl-  VO CN  vo CN  o U  in  CO  ,—i T-H  m rN  CN  m m  vo ro  00  2 §  > c .S5 co  3  m  co CD  rt  |  CD  rt >  CO Ui  CD >  —  3 o  U  a  o  CM 00  U  w  CD  m  00  <  CO  3  b0  m  Ui  w I—I  Pi  fe  co co CD  3  CD  1 c  CD  a  <  H-1 fe fe  • rH Ul  CO  3 • o  U  i—i fe fe-  a  CO  '3 o Ul 3 D  fe fe  CD CD  3  CO  +->  CD  ul 3 <D CI,  O  i—i fe fe  CD  3  00  O  CD  e o  CO  3  ^  CO  CD O 00  CD  JO  3  <D co O  rt  CD  Q  CD H  00  2-bo g rt HH 3  Q m  PQ  r-H  CD  O  a) •  O  CO  co CD  CO C I  O  n  u C3  209  Table 14: Study 4: Correlations of POND scales with peer ratings of narcissism.  Type of Narcissism Sub Scale  Egotistical  Insecure  .26*  -.20  Manipulativeness  •21  .22  Public Recognition  .03  .15  V a i n Exhibition  .21  -.24*  .31*  .04  NPI  .35*  -.06  POND  .35*  -.09  P O N D Core  .35*  -.11  Assured Leadership  Competitive-Ambition  Note. N = 82. The core was tested with one-tailed significance, p < .01. The subscales were tested with two-tailed significance. The correlation between egotistical and insecure narcissism ratings was .17, n.s.  210  Table 15a: Study 4: Correlations of POND scales with performance-related ratings.  Narcissism Measure Rating Mental Ability Verbal Skills Social Competence Problem Solving Ability Talk-time Group Contribution  Note.  NPI  POND  POND Core  Assured Leadershi P  ManipPublic ulativenes Recognitio s n  Vain Exhibit ion  Compet Ambition  .30*  ..31*  .24*  .33*  .12  -.11  .18  .17  .20*  .24*  .22*  .38*  .16  -.03  .19*  .21*  .24*  .21*  .23*  .32*  .00  .23*  .16  .16  .12  .09  .25*  .11  -.25  .19*  .13  .30*  .35*  .28*  .20*  .18  -.12  .16  .14  .14  .11  .10  .35*  -.13  -.08  .16  .22*  .04  N = 11. Mental Ability=Mean(Intelligent, Wise, Creative). * indicates one-tailed significance for NPI, POND, and POND core correlations; and two-tailed significance for the subscales, p<.Q5  211  Table 15b. Study 4: Correlations of POND scales with various interpersonal ratings.  Rating  NPI  POND  POND  Assured Leadership  Manipulativenss  Public Recognition  Vain Exhibit.  Compet Ambition  Core .23*  .24*  .23*  20  -.09  -01  00  .25*  .33*  .30*  .46*  -.11  .09  .22  Is supportive  .00  -.14  -.12  .19  -.30*  -03  -.08  -.16  Is defensive  .06  .15  .19*  .20  .03  -11  .18  .10  Tends to brag  .28*  .35*  .36*  48*  .03  -.00  .30*  .33*  Sympathetic  -.03  -.13  -.11  .20  -.23  .04  -.17  Overestimate  .06  abilities Well-  .30*  Adjusted  Note: N - 11. * indicates one-tailed significance for NPI, POND, and POND core correlations; and two-tailed significance for the subscales, p<.05. Adjustment was calculated as = Mea«(SeIf-acceptance-Insecurity).  -.17  212  List of Figures Figure 1:  Projection of the P O N D scales onto the circumplex.  213  0)  Appendix I List of documents in the order that they were discussed in the text.  1.  Narcissistic Personality Disorder (NPI)  2.  Profile of Narcissistic Dispositions Preliminary ( P O N D - L )  3.  Five Factor Inventory (FFI)  4.  Interpersonal Adjective Scales (IAS-R-B5)  5.  Profile of Narcissistic Dispositions Final (POND)  6.  Fenigstein Public Private Self-Consciousness Scale (PPSC)  7.  Inventory of Interpersonal Problems (IIP)  8.  Cheek Revised Shyness scale  9.  Reflection-Rumination Questionnaire (RRQ)  10. State Trait Anxiety Inventory (Form S) 11. Aspects of Identity Questionnaire (Al) 12. Basic Symptom Inventory 13. Rosenberg Self-Esteem Scale 14. Balanced Inventory of Desirable Responding (BIDR-6)  Male  I.D.#  Female Age  NPI Instructions:  The  N P I consists  identify. "I  don't  own  mind  feelings  other  of a n u m b e r  Consider this  following  about  people",  with  A  orders."  yourself?  than  of pairs of statements  example:  "I like  more  " n o t minding following  which  authority  W h i c h of these  If you identify  with  having  y o u m a y or m a y not  over  two statements  with  people", is closer  "liking to have  orders", then  versus to your  authority  y o u would  B  choose  over option  "A". j You  m a y identify  which with  seems either  words, own item.  (01)  (02)  (03)  (04)  (05)  (06)  closer  Please  both  to y o u r  statement,  read e a c h  feelings.  with  select  " A " a n d " B " . In this personal  the one which  pair of statements  Indicate  your  do not skip  feelings  answer  about  case  yourself.  is least  and then  y o u should choose  objectionable  choose  or remote.  the one t h a t  b y circling t h e letter  the  statement  O r , i f y o u do not identify is closer  In other to your  " A " o r " B " to the left  a n y items.  A  I have a natural talent for influencing people.  B  I am not good at influencing people.  A  Modesty doesn't become me.  B  I am essentially a modest person.  A  I would do almost anything on a dare.  B  I tend to be a fairly cautious person.  A  When people compliment me I sometimes get embarrassed.  B  I know that I am good because everybody keeps telling me so.  A  The thought of ruling the world frightens the hell out of me.  B  If I ruled the world it would be a much better place.  A  I can usually talk my way out of anything.  B  I try to accept the consequences of my behavior.  o f the  (07)  (08)  (09)  216  A  I prefer to blend in with the crowd.  B  I like to be. the center of attention.  A  I will be a success.  B  I am not too concerned about success.  A  I am no better or no worse than most people.  B  I think I am a special person.  A  I am not sure if I would make a good leader.  B  I see myself as a good leader.  A B  I am assertive. I wish I were more assertive.  A  I like having authority over other people.  B  I don't mind following orders.  A  I find it easy to manipulate people.  B  I don't like it when I find myself manipulating people.  A  I insist upon getting the respect that is due me.  B  I usually get the respect that I deserve.  A  I don't particularly like to show off my body.  B  I like to display my body.  A  I can read people like a book.  B  People are sometimes hard to understand.  A  If I feel competent I am willing to take responsibility for making decisions.  B  I like to take responsibility for making decisions.  A  I just want to be reasonably happy.  B  I want to amount to something in the eyes of the world.  i-  (10)  (11)  (12)  (13)  (14)  (15)  (16)  (17)  (18)  217  (19)  (20)  (21)  A  My body is nothing special.  B  I like to look at my body.  A  I try not to be a show off.  B  I am apt to show off if I get the chance.  A  I always know what I am doing.  B  Sometimes I'm not sure of what I'm doing. f  (22)  (23)  (24)  (25)  (26)  (27) .  (28)  (29)  (30)  A  I sometimes depend on people to get things done.  B  I rarely depend on anyone else to get things done.  A  Sometimes I tell good stories.  B  Everybody likes to hear my stories.  A  I expect a great deal from other people.  B  I like to do things for other people.  A  I will never be satisfied until I get all that I deserve.  B  I take my satisfactions as they come.  A  Compliments embarrass me.  B  I like to be complimented.  A  I have a strong will to power.  B  Power for its own sake doesn't interest me.  A  I don't very much care about new fads and fashions.  B  I like to start new fads and fashions.  A  I like to look at myself in the mirror.  B  I am not particularly interested in looking at myself in the mirror.  A  I really like to be the center of attention.  B  It makes me uncomfortable to be the center of attention.  218  (31)  A  I can live my life in any way I want to.  B  People can't always live their lives in terms of what they want.  A  Being an authority doesn't mean that much to me.  B  People always seem to recognize my authority.  A  I would prefer to be a leader.  B  It makes little difference to me whether I am a leader or not.  (34)  A B  I am going to be a great person. I hope I am going to be successful.  (35)  A  People sometimes believe what I tell them.  B  I can make anybody believe anything I want them to.  (36)  A B  I am a born leader. Leadership is a quality that takes a long time to develop.  (37)  A  I wish somebody would someday write my biography.  B  I don't like people to pry into my life for any reason.  A  I get upset when people don't notice how I look when I go out in public.  B  I don't mind blending into the crowd when I go out in public.  A  I am more capable than other people.  B  There is a lot that I can learn from other people.  A  I am much like everybody else.  B  I am an extraordinary person.  (32)  (33)  (38)  (39)  (40)  I.D.#  Male  Female_ Age  POND (Form  L)  This questionnaire consists of 104 statements. For each of the statements indicate the response which best represents your opinion according to the scale shown below (and reprinted at the top of each page):  JL  JL  Strongly Disagree  JL  Disagree  Slightly Disagree  ±  Slightly Agree  For example, consider the statement "/ like to do things for slightly disagree with this statement, you would circle "3":  1  2 (|)  4  5  6  JL  Agree  JL  Strongly Agree  other people."  If you  I like to do things for other people.  If, on the other hand, you strongly agree with this statement, you would circle "6". If you strongly disagree, you would circle "1", and so on.  There are no "right" or "wrong" answers, and you need not be an "expert" to complete this questionnaire. The purpose of this questionnaire will be best served if you describe yourself and state your opinions as accurately as possible. Please read each item carefully and indicate the one response that best corresponds to your agreement or disagreement. Please be sure to answer (don't  skip any).  all items  220  1.  T h e thought of r u l i n g the w o r l d frightens the h e l l out of me.  1  2  3  4  5  6  2.  If I have to, I ' l l lie to get out of a sticky s i t u a t i o n .  1  2  3  4  5  6  3.  M o s t people are pretty easy to figure out.  1  2  3  4  5  6  4.  I sometimes take advantage o f other people.  1  2  3  4  5  6  5.  M y parents always gave me the r e c o g n i t i o n that I deserved.  1  2  3  4  5  6  6.  E v e n though I k n o w I'm right, I sometimes g i v e i n to others.  1  2  3  4  5  6  7.  I don't m i n d blending into the c r o w d w h e n I g o out i n public.  1  2  3  4  5  6  8.  I don't r e a l l y ask m u c h from  1  2  3  4  5  6  9.  I find i t easy to manipulate people.  1  2  3  4  5  6  10.  I don't p a r t i c u l a r l y l i k e to show off m y body.  1  2  3  4  5  6  11.  I l i k e i t w h e n members of the opposite s e x " c h e c k me out".  1  2  3  4  5  6  12.  r*m probably a bit o f a f l i r t  1  2  3  4  5  6  13.  I f I want to make sure that something's done r i g h t , I have to do it m y self.  1  2  3  4  5  6  14.  I ' m not what you'd c a l l a modest p e r s o n .  1  2  3  4  5  6  16.  I believe that honesty is always the best  1  2  3 ' 4  5  6  16.  I l i k e to dress i n trendy fashions.  1  2  3  4  5  6  17.  I feel uncomfortable i f I try to influence people.  1  2  3  4  5  6  18.  I get upset w h e n people don't notice h o w I l o o k w h e n I go out in public  1  2  3  4  5  6  19.  I must admit F m a pretty fair t a l k e r .  1  2  3  4  5  6  20.  P e o p l e often tell me how great I l o o k .  1  2  3  4  5  6  21.  F m not v e r y good at telling j o k e s a n d stories.  1  2  3  4  5  6  22.  I don't m i n d doing favors for other people, e v e n i f i t m i g h t inconvenience me.  1  2  3  4  5  6  23.  I a m apt to show off i f I get the c h a n c e .  1  2  3  4  5  6  24.  P o w e r for its o w n sake doesn't interest m e .  1  2  3  4  5  6  25.  I f somebody cuts i n front of me i n a line-up, I u s u a l l y w o n t say anything.  1  2  3  4  5  6  26.  B e i n g a n authority doesn't mean that m u c h to m e .  1  2  3  4  5  6  27.  I f 8 pretty easy for me to get other people to d o w h a t I want t h e m to do.  1  2  3  4  5  6  28.  I don't l i k e it, but sometimes F m pretty  1  2  3  4  5  6  29.  I c a n t stand losing at games.  1  2  3  4  5  6  life.  policy.  selfish.  221  30.  Sometimes  31.  I  think  32.  I  like  33.  rd  34.  Compliments  35.  I  36.  1*11 d o a n y t h i n g  37.  I  am  38.  I  believe  39.  Even if I mess my behavior.  40.  I  believe  41.  I  see myself  42.  F m  43.  Pd  like  44.  If s  my  46.  H  46.  I  47.  People  48.  When  49.  I  probably  60.  I  don't  51.  If  I'm  that  am  a  I'm  give  5  6  good-looking.  1  2  3  4  5  6  of attention.  1  2  3  4  5  6  orders  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  ,.2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  than  have  doing.  to take  them.  me.  leader. within  reason  i n the survival  that  easily  up, I  to get  ahead.  as  a  bored  always  good  with  ambition  people  show  probably  wouldn't  pay very  less  don't  up, m y make  a  really  believe  54.  I  pretty  much  55.  I  think  56.  F m  not really  what  57.  Vm  not above  flattering  58.  People  59.  I  am  not good  at influencing  60.  I  am  not sure  if I  that  seem  that  with  I  a  to m e .  n^ontnpl^kmfdtf  work,  you'd  really  average. appreciated,me.  leader.  n e w trends  or  to go  fashions. out, it  bags  look. "check  life  myself out".  call  to instinctively  throws  anyone  m y way.  can become  the independent  people  would  pretty  " d o g eat dog" world.  whatever  hard  to  as  never  good  usually  ifs  accept  myself.  on the world.  my  dressed-ap  how I  I  mirror,  parents  getting  notice  53.  a  than  my mark"  m y body  attention  When  pass  of  experience.  b i g deal  ApprorintA  very  52.  I  peoples'  intelligent  "leave  describe  much  don't  the consequences  off.  lot of time  a  other  b u t it's not a  pompta  was growing  spend  from  to somehow  wh«n  m«  if people  try to accept  leader.  to b e successful,  would I  o f t h e fittest.  y o u c a n profit  try not to be a  I  4  I'm  assertive. -  hnthAra  me  3  pretty  embarrass  born  2  of what  to b e the center  rather  1  not sure  in  trust  judgement.  people.  make  a  type.  authority. my  good  leader.  a  leader.  222  61.  If  people  62.  I'm  63.  I  64.  When  not  can  let y o u  manipulate  particularly  make I  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  body.  1  2  3  4  5  6  anything.  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  it's  their  own  darn  fault.  ambitions.  anybody  was  them,  believe  growing  up,  anything  my  I  parents  want  were  them  always  to. very  warm  and  supportive. 65.  Getting  along  I  like  67.  I  think  68.  I  can  usually  69.  I  like  to  70.  In  an  P m 71.  I  to  is  look  that  at  I  be  more  important  my  have talk  than  getting  ahead.  body. pretty  a my  way  good  out  of  complimented.  argument,  111  sometimes  give  in  even  if  I  know  that  right  try  to  avoid  having  to  depend  on  other  people  to  get  things  done. 72.  I  couldn't  73.  I  don't  like  74.  I  think  I  I  think  that  75.  take  tell  a  it  am  lie  if  when a  my  I  find  special  depended  myself  on  H.  manipulating  people.  person.  establishing  priority over  life  successful  a  marriage  and  career  should  take  family.  a  •  76.  P m  sometimes  77.  If s  important  78.  I  will  79.  I  always  80.  I  feel  81.  When  never  I  was  like  83.  I  have  84.  I  p r e f e r to  85.  When  86.  My  a  off  am  the  what to  blend  not  their  "stuck  up".  respect  that  I  deserve.  1  2  3  4  5  6  that  I  deserve.  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  I  get  all  doing.  people.  up,  my  parents  were  usually  cool  and  me. fads  talent in  and for  with  me  if  fashions.  influencing  the me  probably  people.  crowd. I  sometimes  describe  somebody  me  takes  get  as my  embarrassed.  competitive. ideas  and  passes  own.  particularly  mirror.  the  until  am  compliment  bother as  I  being  get  most  new  natural  people  doesn't  to  growing  start  of  satisfied  friends/ would  them I  to  me  toward  I  88.  be  superior  82.  It  to  know  indifferent  87.  accused  interested  in  looking  at  myself  in  223  89.  F m  not  always  sure  of  myself  when  it  comes  to  making  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  1  2  3  4  5  6  1  2  3  4  5  6  1  2  3  4  5  6  decisions. 90.  I  really  91.  I  get  resent  it  when  uncomfortable  "check  me an  someone  when  else  members  does  of  the  better  that  opposite  I  sex  do.  out".  92.  I  am  extraordinary  93.  I  have  a  strong  94.  I  tend  to  look  96.  I  usually  96.  I  don't  97.  The  98.  I  would  99.  I  don't  100.  I  am  101.  Taking  102.  I  spend  103.  I  work  will  it  thought  prefer  more  out  leap. gracefully.  people  to  be  having  capable  lot  I  talking  advantage a  power.  defeat  when  of  mind  to  before  accept  like  person.  of  leader.  to  depend  to  other  other  time  me  front  a  than  of  mainly  in  ask  on look  of  on  a  myself.  group  other  terrifies  me.  people.  people.  people my  about  can  never  appearance.  good.  be  justified.  1  2  3  4  5  5  6  6  224  NEO  - FFI  Using the s c a l e below as a guide, w r i t e a number b e s i d e each t o i n d i c a t e hov much you agree v i t h i t . •  not t r u e  •  1  ~  2  •  •  ~3 somewhat  +  +  4  5  statement  very t r u e  1.  I am  not a w o r r i e r .  2.  I l i k e t o have a l o t of people around  3.  I don't  4.  I t r y t o be courteous to everyone I meet.  5.  I keep my  6.  I o f t e n f e e l i n f e r i o r to others.  7.  I laugh  8.  Once I f i n d the r i g h t way  9.  I o f t e n get i n t o arguments with my f a m i l y and co-workers.  me.  l i k e t o waste my time daydreaming.  belongings c l e a n and  neat.  easily. t o do something. I s t i c k t o i t .  I' m p r e t t y good about pacing myself so as t o get t h i n g s 10. done on time. 11. When I'm under a great d e a l o f s t r e s s , sometimes I f e e l l i k e I'm g o i n g t o pieces. 12.  I don't  c o n s i d e r myself e s p e c i a l l y  13.  I am  14.  Some p e o p l e t h i n k I'm  15.  I am  16.  I r a r e l y f e e l l o n e l y or blue.  17.  I r e a l l y enjoy t a l k i n g to people.  "lighthearted".  i n t r i g u e d by the p a t t e r n s I f i n d i n a r t and nature. s e l f i s h and  not a very methodical  egotistical.  person.  18. I b e l i e v e l e t t i n g students hear c o n t r o v e r s i a l can o n l y c o n f u s e and mislead them.  speakers  19. I would r a t h e r cooperate with o t h e r s than compete with them. 20. I t r y t o perform a l l the t a s k s a s s i g n e d t o conscientiously.  me  225  21.  I o f t e n f e e l tense and j i t t e r y .  22.  I l i k e t o be where t h e a c t i o n i s .  23.  Poetry has l i t t l e or no e f f e c t on me.  24.  I tend t o be c y n i c a l and s k e p t i c a l of o t h e r ' s i n t e n t i o n s .  25. I have a c l e a r s e t o f g o a l s and work toward them i n an orderly fashion. 26.  Sometimes I f e e l completely worthless.  27.  I u s u a l l y p r e f e r t o do t h i n g s alone.  28.  I o f t e n t r y new and f o r e i g n foods.  29. I b e l i e v e t h a t most people w i l l take advantage you l e t them.  o f you i f  30.  I waste a l o t o f time b e f o r e s e t t l i n g down t o vork.  31.  I r a r e l y f e e l f e a r f u l o r anxious.  32.  I o f t e n f e e l as i f I am b u r s t i n g with energy.  33. I seldom n o t i c e t h e moods or f e e l i n g s that d i f f e r e n t environments produce. 34.  Most people I know l i k e me.  35.  I work hard t o a c c o m p l i s h my goals.  36.  I o f t e n g e t angry a t t h e way people t r e a t me.  37.  I am a c h e e r f u l , h i g h - s p i r i t e d person.  38. I b e l i e v e we s h o u l d l o o k t o our r e l i g i o u s a u t h o r i t i e s f o r d e c i s i o n s on moral i s s u e s . 39.  Some people t h i n k o f me as c o l d and c a l c u l a t i n g .  40. When I make a commitment, I can always be counted on t o f o l l o w through. 41. Too often, when t h i n g s go wrong, I get d i s c o u r a g e d and f e e l l i k e g i v i n g up. 42.  I am not a c h e e r f u l  optimist  43. Sometimes when I am r e a d i n g poetry or l o o k i n g a t a work o f a r t , I f e e l a c h i l l o r wave of excitement. 44.  I'm hard-headed  and tough-minded i n my a t t i t u d e s .  226  45. be.  Sometimes I'm not as dependable or r e l i a b l e as I should  46.  I am seldom s a d or depressed.  47.  My l i f e  i s fast-paced.  48. I have l i t t l e i n t e r e s t i n s p e c u l a t i n g on t h e nature o f the universe o r t h e human c o n d i t i o n . 49.  I g e n e r a l l y t r y t o be t h o u g h t f u l and c o n s i d e r a t e .  50.  I am a p r o d u c t i v e person vho always g e t s t h e j o b done.  51. I o f t e n f e e l h e l p l e s s and want someone e l s e t o s o l v e my problems. 52.  I am a very a c t i v e person.  53.  I have a l o t o f i n t e l l e c t u a l  54.  If  55.  I never seem t o be a b l e t o get organized.  56.  At times I have been so ashamed I j u s t wanted t o hide.  57.  I would r a t h e r go my own way than be a l e a d e r o f others.  58.  I o f t e n enjoy  p l a y i n g with t h e o r i e s o r a b s t r a c t ideas.  59.  If necessary,  I am w i l l i n g t o manipulate  60.  I s t r i v e f o r e x c e l l e n c e i n e v e r y t h i n g I do.  I don't l i k e people,  curiosity.  I l e t them know i t .  people.  227 i LD.#  '  '  Male  Female Age  I N T E R P E R S O N A L  A D J E C T I V E  S C A L E S  (Form 1ASR-B5)  On  the next  describe word you  t w o pages  people's  describes  as  you  on t h e following  2.  1 extremely inaccurate  For  description  a  consider  y o u as  a  person.  4 sligKtly inaccurate  the word  person?  o f y o u , write  you think  the next  number  this  word  " 4 " next  list  o f words  Please  Judge  rate  that  are used  how accurately  h o w accurately  If  5 slightly accurate  BOLD.  each  the n u m b e r  is a  t o it,  to it, a n d so on..  6. quite accurate  H o w accurately  y o u think  6  If  a  word  to each  describes  scale:  quite inaccurate  example,  find  characteristics.  3  very inaccurate  describe  y o u will  personal  this  "6"  is  next  a  does  7_ very accurate  that  8. extremely accurate  word  quite accurate  to it:  B O L D  tlightly inaccurate d e s c r i p t i o n i f i t i s very inaccurate w r i t e  of you,  write  the number  "2"  1  1  3  extremely inaccurate  very inaccurate  quite inaccurate  (001)  Introverted  (002) Assertive  4  5  6  slightly inaccurate  sligKtry accurate  quite accurate  (022) Anxious ___(023) Abstract-thinking  S very accurate  extremely accurate  (043) Forceful (044) Wily  (003)  Timid  (024) Philosophical  (045) Undisciplined  (004)  Unargumentative  (025) Tender  (046) Sly  (005)  Organized  (026) Hardhearted  (047) Systematic  (006)  Boastful  (027) Unneighbourly  (048) Self-conscious  (007)  Softhearted  (026) Worrying  (049) Ironhearted  (008)  Ruthless  (029) Literary  (050) Thorough  (030) Uncharitable  (051) Untidy  1009) Kind (010)  Tense  (031) Uncunning  (052) Unbold  (011)  Highstrung  (032) Hypersensitive  (053) Neighbourly  (012)  Cheerful  _(033) Extraverted  (054) Unorderly  (013)  Unsparkling  (014) Tricky  _(034) Unphilosophical  (055) Shy  (035) At ease  (056) Undemanding  (015)  Unconventional  (036) Orderly  (057) Meek  (016)  Inefficient  (037) Cocky  (058) Reflective  (017)  Unaggressive  (036) Planful  (059) Inquisitive  (018)  Unreflective  (039)  (019)  Relaxed  (040) Unsearching  (061) Unsystematic  (04}) Anti-social  (062) Self-assured  (042) Perky  (063) Dissocial  (020) Calculating __(021) Unmoody  Dominant  (060) Unwiry  228  1  2  3  4  5  extremely inaccurate  very inaccurate  quite inaccurate  sligHtry inaccurate  slightly accurate  «  quTle accurate  L  very accurate  8. extremely accurate  (064)  Jovial  (085)  Friendly  (106)  Unrefiable  (065)  r*>mineering  (086)  Cunning  (107)  Outgoing  (066)  Neat  (087)  Self-confident  (108)  Sympathetic  (067)  Unabstract  (088)  Unauthoritative  (109)  Boastless  (089)  Uncrafty  Unworrying  (090)  Unsympathetic  Unimaginative  (091)  Charitable  _(112)  Imaginative  Coldhearted  _(113)  Persistent  (068)  _(069)  (070)  _(071)  Tenderhearted  Tidy  _(092)  _(110)  (111)  Unnervous  UnBterary  J072)  Warmthless  (093)  Guilt-prone  J114)  Reliable  J073)  Unsly  (094)  Nervous  _(115)  Crafty  _(074)  Enthusiastic  (095)  Broadminded  __(116)  Unagitated  (096)  Distant  J117)  Stable  (097)  Forceless  _(118)  Uninquisitive  (096)  Efficient  _(119)  Unsociable  (099)  Fretful  J120)  Unarastic  Overexdtable  _(121)  SeK-disciplined  _(122)  Forgetful  _(123)  Cruel  J075) Rrm  _(076)  Impractical  _(077)  Uncalculating  __(078)  Questioning  _(079)  Accommodating  _(080)  Uncheery  __(081)  Uncomplex  __(082)  Calm  (103)  Unplanful  (083)  Conventional  (104)  Unanxious  (084)  Individualistic  (105)  UnseK-conscious  _(100)  (101)  Centlehearted  _(102)  Disorganized  (124)  Bashful  229  Male  I.D.#.  230  Female. Age.  POND Instructions: For each of the 66 statements located on the following pages, please indicate your level of agreement or disagreement by circling one of the scale categories to the right of each statement. Use the scale as shown below:  Strongly Disagree 1  Disagree  Slightly Disagree  Slightly Agree  Agree  Strongly Agree  2  3  4  5  6  For example, consider the statement "I'm quite a flirt." 2  3  4  5  6  If you mostly disagree, you would circle:  ©  3  4  5  6  If you disagree slightly more than agree, you would circle:  2  (3)  4  5  6  If you agree slightly more than disagree, you would circle:  2  3  ®  5  6  If you mostly agree, you would circle:  2  3  4  ©  6  If you strongly agree, you would circle:  2  3  4  5  ©  If you strongly disagree, you would circle:  (T)  Please try and make use of all levels of the scale in your answers. Read each item carefully and circle the scale level that best reflects your actual opinion about yourself. Please do not skip any items. The questionnaire takes about 10-15 minutes.  231  1. The thought of ruling the world frightens the hell out of me. 2.  1 2  If I have to, I'll He to get out of a sticky situation.  3  4  5 6  1 2 3 4 5 6  3. I sometimes take advantage of other people.  1 2  3  4  5  6  4.  1 2  3  4  5  6  5. I like it when members of the opposite sex "check me out".  1 2  3  4  5  6  6. I'm probably a bit of a flirt.  12  3  4 5  7. I believe that honesty is always the best policy.  1 2  3  4  5  6  8. I like to dress in trendy fashions.  1 2  3  4  5  6  1 2  3  4  5  6  I don't particularly like to show off my body.  I am generous with my friends. 10. I get upset when people don't notice how I look when I go out in public.  1 2 3 4 5 6 •.  11. I must admit I'm a pretty fair talker. 12. I'm not very good at telling jokes and stories. I quickly get over and recover from being startled.  6  v  1 2 3 4 5 6 1 2  3  4  5  1 2  3  4  14. Power for its own sake doesn't interest me.  1 2  3  4  15. Being an authority doesn't mean that much to me.  1 2  3  4  5  6  16. It's pretty easy for me to get other people to do what I want them to do.  1 2  3  4  5  6  I enjoy dealing with new and unusual situations.  1 2  3  4  5  6  18. I don't like it, but sometimes I'm pretty selfish.  1 2  3  4  5  6  19. I can't stand losing at games. I usually succeed in making a favorable impression  1 2  3  4  5  6  ®  5  6  5  6 6  21.  on people. I think that I'm pretty good-looking.  1 2 3 4 5 6 1 2 3 4 5 6  22.  Compliments embarrass me.  1 2  3  4  5  6  232  23.  I'll do anything within reason to get ahead.  1  2  3  4  5 6  ^ 4 ^ I enjoy trying new foods I have never tasted before.  1  2  3  4  5 6  1  2  3  4  5 6  1  2  3  4  5 6  1  2  3  4  5 6  25.  I am assertive.  26.  I believe in the survival of the fittest.  .,  ^27^ I am regarded as a very energetic person. 28.  I see myself as a good leader.  1  2  3  4  5 6  29.  I'd like to be successful, but it's not a big deal to me.  1  2  3 4  5 6  30.  It's my ambition to somehow "leave my mark" on the world.  1  2  3  4  5 6  31.  It bothers me when people don't appreciate my accomplishments.  1  2  3  4  5 6  1  2  3  4  5 6  1  2  3  4  5 6  1  2  3  4  5 6  ^32^ I like to take different paths to familiar places. 33.  I probably wouldn't make a very good leader.  ^ 4 ^ I am more curious than most people. 35.  I don't pay very much attention to new trends or fashions.  1  2  3  4  5 6  36.  If I spend a lot of time getting dressed-up to go out, it bugs me if people don't notice how I look.  1  2  3  4  5 6  Most of the people I meet are likable.  1  2  3  4  5 6  38.  When I pass a mirror, I usually "check myself out'.  1  2  3  4  5 6  39.  I really believe that it's a "dog eat dog" world.  1  2  3  4  5 6  40.  People seem to instinctively trust my judgement.  1  2  3  4  5 6  41.  I am not good at influencing people.  1  2  3  4  5 6  1  2  3  4  5 6  ^2^) I usually think carefully about something before acting. 43.  If people let you manipulate them, it's their own darn fault.  1  2  3  4  5 6  44.  I'm not particularly ambitious.  1  2  3  4  5 6  233  234  45^ I like to do new and different things.  1 2  3  4 5 6  46.  I like to look at my body.  1 2  3 4 5 6  47.  I think that I have a pretty good body.  1 2  3 4  5 6  48.  I can usually talk my way out of anything.  1 2 3 4  5 6  49.  I don't like it when I find myself manipulating people.  1 2 3 4  5 6  My daily life is full of things that keep me interested.  1 2  5 6  3 4  51.  I think I am a special person.  1 2 3  4  5 6  52.  It's important to me to get the respect that I deserve.  1 2 3 4  53.  I have a natural talent for influencing people.  1 2  3 4 5 6  54.  When people compliment me I sometimes get embarrassed.  1 2  3 4  55.  My friends would probably describe me as competitive.  1 2  3 4  56.  I'm not always sure of myself when it comes to making decisions.  1 2  3 4 5 6  57.  I really resent it when someone else does better that I do.  58.  I get uncomfortable when members of the opposite sex "check me out."  5 6  5 6 5 6  1 2 3 4 5 6  1 2 3 4 5 6  59J I would be willing to describe myself as a pretty "strong" — ' personality. 60.  I am an extraordinary person.  61.  I usually accept defeat gracefully.  62.  The thought of talking in front of a group terrifies me.  63. Taking advantage of other people can never be justified. 64^ I get over my anger at someone reasonably quickly. 65. 66.  I spend a lot of time on my appearance. I work out mainly to look good.  1 2.3  4 5 6  1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1  2  3 4 5 6  1 2 3 4 5 6 1 2 3 4 5 6  Fenigstein  Scale  Using the s c a l e below as a guide, w r i t e a number b e s i d e each statement to i n d i c a t e how much you agree with i t .  +  +  +___  +  +  +_  +  1 .  2  3  4  5  6  7  not t r u e  somewhat  1)  Large groups make me  2)  I'm  very true  nervous.  aware o f the way my mind works when I work  through a problem.  _____  3)  I'm  always t r y i n g t o f i g u r e myself out.  4)  Im concerned about my  5)  G e n e r a l l y , I'm  6)  I t takes me  s t y l e o f doing t h i n g s .  not v e r y aware o f myself.  time t o overcome my  shyness i n new  situations.  .  7)  I r e f l e c t about myself a l o t .  8)  I'm  concerned about the way  9)  I'm  o f t e n the s u b j e c t o f my own  I present m y s e l f . fantasies.  10)  I have t r o u b l e working when someone i s watching me.  11)  I never s c r u t i n i z e m y s e l f .  12)  I get embarassed v e r y e a s i l y .  13.  I'm  14)  I don't f i n d i t hard t o t a l k to s t r a n g e r s .  15)  I'm  16)  I u s u a l l y worry about making a good i m p r e s s i o n .  17)  I'm  18)  I f e e l anxious when I speak i n f r o n t of a group.  s e l f - c o n s c i o u s about the way  I look.  g e n e r a l l y a t t e n t i v e t o my inner  c o n s t a n t l y examining my  feelings.  motives.  19)  One of the t h i n g s I do b e f o r e I leave my look i n the m i r r o r .  20)  I sometimes have the f e e l i n g t h a t I'm somewhere watching m y s e l f .  21)  I'm me.  concerned  22)  I'm  a l e r t to changes i n my mood.  23)  I'm  u s u a l l y aware o f my  house i s  off  about what other people t h i n k o f  appearance.  237  I.D#  Male_  Female_ Age  I N V E N T O R Y  O F  I N T E R P E R S O N A L  P R O B L E M S  Shortform-C  Listed below are a variety of common problems that people report in relating to other people. Please read each one and consider whether that problem has been a problem for you with respect to any significant person in your life. Then select the number that describes how distressing that problem has been, and circle that number. •  EXAMPLE  How much have you been distressed by this problem?  i* It is hard  00.  for me to...  get along with my relatives  /  /  Part  I.  The  It i s hard trust  2.  say  "no"  3.  join  in  4.  keep  5..  let  •  other  things  yon  find  hard  to  do  with  people.  to  on  other  a  people.  groups.  things  other  tell  are  other  people.  for me to...  1.  6.  following  private  people  person  from  know  to  what  stop  myself  people.  I  want.  bothering  me.  3  0  I  2  3  0  I  2  3  0  L  2  3  0  L  2  3  o  I  2  3  0  ]I  2  3  0  ]L  2  3  0  ]L  2  3  0  ]L  2  3  0  ]L  2  3  0  ]1  2  3  0  ]L  2  3  0  ']L  2  3  0  1L  2  3  0  IL  2  3  0  1L  2  3  d  iL  2  3  0  ]L  2  3  0  1I  2  3  0  1  2  3  0  1  2  3  0  ]  2  3  0  1  2  3  0  1  2  3  0  1  2  3  0  1  2  3  0  1  2  3  8.  confront  9.  be  assertive  10.  let  other  11.  make  12.  be  another  13.  be  aggressive  14.  socialize  15.  show  16.  get  17.  understand  18.  express  19.  be  20.  experience  21.  set  22.  be  23.  feel  24.  really  care  about  25.  argue  with  another  26.  spend  time  alone.  27.  give  28.  let myself feel angry at somebody  29.  put somebody else's needs before my o w n .  0  1  2  3  30.  stay out of other people's  0  1  2  3  31.  take instructions from people who have authority over me.  0  1  2  3  with  people  know  along  other  limits  I  feeling  on  other  to  gift  of  I'm  angry. to  someone  another  person.  when  the  situation  calls  for  people.  person's to  point  other  of  of  people  view. directly.  be.  to  love  for  another  person.  people.  another  other  to  when  need  a  supportive  person.  people.  feelings  when  up.  people.  another  my  close  to  come  boss.  toward  with  that  commitment  person's  with  firm  problems  another  long-term  affection  a  with  people.  2  introduce  a  new  :  1  7.  people  to  other  0  person's  goals  in  life.  people. other  people's  problems.  person.  another  person. I  like.  business.  it.  239  32.  feel  good  33.  ask  other  34.  feel  angry  35.  open  36.  forgive  37.  attend  38.  be  assertive  39.  be  self-confident  Part  about  another  people  up  at  tell  another my  happiness.  together  socially  my  feelings  person  I  following  another  I've  been  when  worrying  when  somebody  about  am  with  are  things  people  other  that  I  feel  too  responsible  42.  I  am  too  easily  43.  I  open  44.  I  am  too  independent.  45.  I  am  too  aggressive  46.  I  try  to  47.  I  clown  48.  I  want  to  49.  I  trust  other  50.  I  try  to  51.  I  put  other  52.  I  try  to  53.  I  am  too  54.  I  am  overly  55.  I  am  too afraid  56.  I  am  too suspicious  57.  I  manipulate  58.  I  tell  59.  I  argue  60.  I  keep other people at a  61.  I  let  62.  I feel embarrassed in front  63.  I am affected  64.  I want to get revenge against people too  people  please  too  around  too  feelings.  yon  do  too  other  other  people's  other too  too too  other  problems.  people.  people. much.  people's  much. much.  people  needs  other  too  before  people  too  much. my  own  too  much.  much.  gullible. generous  personal  of  other  to  other  other of  people  things  to  people.  too other  with other people  other people take  people.  people.  other  too  much people  get  too  what  I  want.  much.  much.  distance  advantage of  to  too of  much. me  too  much.  other people too  by another person's misery too  much. much.  much.  0  ]L  2  3  0  ]L  2  3  0  1L  2  3  0  1L  2  3  0  ]L  2  3  0  JL  2  3  0  ]L  2  3  0.  1L  2  3  0  I  2  3  4  0  1L  2  3  4  0  1L  2  3  4  0  ]L  2  3  4  0  ]L  2  3  4  o  :1  3  4  0  ]L  2  3  4  0  1L  2  3  4  o  :L  2  3  4  0  ]L  2  3  4  0  )L  2  3  4  0  ]L  2  3  4  0  :I  2  3  4  0  JI  2  3  4  much.  much.  people  change  other's  much.  people  noticed  control  by  toward  other  needy.  much.  solving  persuaded  is  people.  41.  for  else  hurting  fight  be  too  angry.  I  to  me.  person.  40.  up  other  to  after  welfare  without  with  with  people.  own  The  n.  get  other  and  to  to  person's  2  0  L  2  3  4  0  ]L  2  3  4  0  ]L  2  3  4  0  ]L  2  3  4  0  ]L  2  3  4  0  ]L  2  3  4  o  :L  2  3  4  0  I  2  3  4  0  L  2  3  4  0  1  3  4  0  L  3  4  2 2  Cheek Revised scale Using the s c a l e below as a guide, w r i t e a number b e s i d e each statement t o i n d i c a t e how much you agree w i t h i t .  +______+_  1 not t r u e  2  + 3  +  +.  4 neutral  5  +_ 6  _+  7 very t r u e  1.  I f e e l tense when I'm with people I don't well.  know  2.  I am s o c i a l l y somewhat awkward.  3.  I do not f i n d i t d i f f i c u l t t o ask o t h e r people f o r information.  4.  I am o f t e n uncomfortable s o c i a l functions.  5.  When i n a group of people, I have t r o u b l e t h i n k i n g o f the r i g h t t h i n g s t o t a l k about.  6.  I t does not take me long to overcome my. shyness i n new s i t u a t i o n s .  7.  I t i s h a r d f o r me t o a c t n a t u r a l when I am meeting w i t h new people.  8.  I f e e l nervous when speaking t o someone i n authority.  9.  I have no doubts about my s o c i a l competence.  at p a r t i e s and o t h e r  10.  I have t r o u b l e l o o k i n g someone r i g h t i n the eye.  11.  I feel inhibited i n social  12.  I do not  13.  I am more shy with members o f the o p p o s i t e sex.,  situations.  f i n d i t hard t o t a l k t o s t r a n g e r s .  Male  I.D.#.  Female. Age  RRQ Instructions: For each of the statements located on the next two pages, please indicate your level of agreement or disagreement by circling one of the scale categories to the right of each statement. Use the scale as shown below:  Strongly Disagree  Disagree  Neutral  Agree  SD  D  N  A  Strongly Agree SA  For example, consider the statement "I'm often absent-minded. If you strongly disagree, you would circle:  (SD)  D  N  A  SA  N  A  SA  If you mostly disagree, you would circle:  SD  If you are in the middle or neutral, you would circle:  SD  D  SA  If you mostly agree, you would circle:  SD  D  SA  If you strongly agree, you would circle:  SD  D  S  A  Read each item carefully and circle the scale category that best reflects your actual opinion about yourself. Please be sure to answer every item.  Strongly Disagree  Disagree  Neutral  Agree  Strongly Agree  SD  D  N  A  SA  1. My attention is often focused on aspects of myself I would like to stop thinking about. 2. It is easy for me to put unwanted thoughts out of my mind. 3. Often I'm playing back over in my mind how I acted in a past situation. 4. Sometimes it is hard for me to shut off thoughts about myself. 5. I don't waste time re-thinking things that are over and done with. 6. I often reflect on episodes in my life that I should no longer concern myself with. 7. Often my thoughts drift to some concern or worry about myself. 8. I often ruminate or dwell over aspects of my life.  242  SD  D  N  A  SA  SD  D  N  A  SA  SD  D  N  A  SA  SD D  N  A SA  SD  D  N  A  SA  SD  D  N  A  SA  SD D  N  A SA  SD  D  N  A  SA  SD  D  N  A  SA  10. My mind rarely gets preoccupied by personal worries.  SD  D  N  A  SA  11. My thoughts rarely dwell on self-concerns.  SD  D  N  A  SA  12. I often find myself re-evaluating something I've done.  SD  D  N  A  SA  SD  D  N  A  SA  SD  D  N  A  SA  SD  D  N  A  SA  9. Long after an argument or disagreement is over with, my thoughts keep going back to what happened.  13. I often wish I could stop from "re-hashing" in my mind my personal doubts and insecurities. 14. When I make mistakes, I tend to put them out of my mind quickly and move on. 15. I rarely think about past embarrassments or disappointments.  Strongly Disagree  Disagree  SD  D  Neutral  _ ' Agree  Strongly Agree  N  A  SA  1. Philosophical or abstract thinking doesn't appeal to me that much.  243  SD  D  N  A  SA  2. I'm not really a meditative type of person.  SD  D  N  A  SA  3. I. love analyzing why I do things.  SD  D  N  A  SA  4 . I find analyzing life a bit boring.  SD  D  N  A  SA  5. I don't like to analyze people's motives.  SD  D  N  A  SA  6. I love to meditate on the nature and meaning of things.  SD  D  N  A  SA  7. I love exploring my "inner" self.  SD  D  N  A  SA  SD  D  N  A  SA  SD  D  N  A  SA  10. I'm very self-inquisitive by nature.  SD  D  N  A  SA  11. I don't really care for introspective or self-reflective thinking.  SD  D  N  A  SA  12. I often love to look at my life in philosophical ways.  SD  D  N  A  SA  13. Contemplating myself isn't my idea of fun.  SD  D  N  A  SA  8. Most people would describe me as quite a "deep" or introspective type of person.  9. I love to contemplate life's mysteries.  STAI - Forii'S  For each statement belov, please c i r c l e a nuaber vhlch seeas to describe hov you generally f e e l . There are no r i g h t or wrong answers. Do not spend too auch tlae on any one statement but give the answer vhlch seeas to best describe your feelings. Use the following s c a l e :  Alaost Never  Soaetiaes -1 1  e  e  Alaost Always 1 3  -  2  3  I lack  0  1 2  3  I aa a steady person.  8  1 2  3  I f e e l nervous and r e s t l e s s .  2  3  I feel  1 2  3  I f e e l s a t i s f i e d with ayself.  2  3  I wish I could be as happy as others seea to be.  0  1 2  3  I feel like a failure.  0  1 2  3  I get In a state of tension or t u r a o l l as I think over ay recent concerns and i n t e r e s t s .  2  3  I worry too auch over soaethlng that does not r e a l l y aatter.  1 2  3  I f e e l secure.  0  i  Often 1 2  1  0 0  0 0  1  1  self-confidence.  Inadequate.  Physiological Questionnnairei Height:__ Eye Color (check one):  Weight: brown  blue  other  Aspects of Identity Questionnaire  Instructions: These items describe different aspects of identity. Please read each item carefully and consider how it applies to you. Circle a number to the left of each item by choosing a number from the scale below:  0= N o t important to my sense of who I am.  1=  Slightly  important to my sense of who I am.  2=  Somewhat  3=  Very  4=  Extremely  important t o my sense of who I am.  important to my sense of who I am. important to my sense of who I am.  2  3  4  My personal values and moral standards.  2  3  4  My popularity with other people.  2  3  4  My dreams and imagination.  2  3  4  The ways in which other people react to what I say and d o .  2  3  4  My personal goals and hopes for the future.  2  3  4  My physical appearance: my height, weight, and the shape of my body.  2  3  4  My emotions and feelings.  2  3  4  My reputation, what others think of me.  2  3  4  My thoughts and ideas.  2  3  4  My attractiveness to other people.  2  3  4  The ways I deal with my fears and anxieties.  2  3  4  My gestures and mannerisms, the impression I make on others.  2  3  4  My feeling of being a unique person, being distinct from others.  2  3  4  My social behavior, such as the way I act when meeting people.  2  3  4  Knowing that I continue t b be essentially the same inside even though life involves many external changes.  2  3  4  My self-knowledge, my ideas about what kind of person I really am.  2  3  4  My personal self-evaluations, the private opinion I have of myself.  j jyf  Male  Female.  Age  BSI  Instructions:  Below is a list of problems and complaints that people sometimes have. Read each one carefully, and circle one of the numbered descriptions to the right of each item that best describes how much discomfort that problem has caused you during the past month, including today. Do not skip any items, and circle your number clearly. If you change your mind, erase your first number completely.  Example: How much were you distressed by:  00. Body aches  0  1 ©  3  4  If this is moderately a problem, you would circle "2". If this is quite a bit of a problem, you would circle "3". If this is extremely a problem, you would circle " 4 " .  . and so on.  247  HOW MUCH WERE YOU DISTRESSED BY: 30. Hot or cold spells  . . . . . . .0  1 2 3 4  31. Having to avoid certain things, places, or activites because they frighten you.. 32. Your mind going blank.. . . . . . .  0  1 2  3,4  .0  1 2  34  33. Numbness or tingling in parts of your b o d y . . . . . . . . . . . . . . .0 1  2 3 4  34. The idea that you should be punished for your sins  .0 1 2 3 4  35. Feeling hopeless about the future  .0 1 2 3 4  36. Trouble concentrating.  .0  37. Feeling weak in parts of your body...  .0  1 2 3 4 1 23  38. Feeling tense or keyed u p . . . . . . . . . . . . . . . . . . . . . . . . . . . .0 1 2 39. Thoughts of death or dying..  0  4 34  1 2 3 4  40. Having urges to beat, injure, or harm someone........... D 1 2 3 4 41. Having urges to break or smash things  0  42. Feeling very self-conscious with others.  .0  12  43. Feeling uneasy in crowds........  0  1 2 3 4  44. Never feeling close to another person.  0  45. Spells of terror or panic.... 46. Getting into frequent arguments.. .  0 /  1 2 3 4  12 12  3 4 3 4 3 4  . . X) 1 2 3 4  47. Feeling nervous when you are left alone.  0  1 2 3 4  48. Others not giving you proper credit for your achievements  ..0  1 2 3 4  49. Feeling so restless you couldn't sit still  .0  1 2 3 4  50. Feelings of worthlessness  .0  1 2 3 4  0  12  51. Feeling that people will take advantage of you if you let them 52. Feelings of guilt  .......  3 4  .0 1 2 3 4  53. The idea that something is wrong with your m i n d . . . . . . . . . . 0  1 2  34  248  HOW MUCH W E R E YOU DISTRESSED BY:  1. Nervousness or shakiness inside  0  2. Faintness or dizziness  0 1 2 3 4  3. The idea that someone else can control your thoughts 4. Feeling others are to blame for most of your troubles  0 ..0  1 2 3 4 1 2 3 4 1 2 3 4  5. Trouble remembering things  0  6. Feeling easily annoyed or Irritated  0 1 2 3 4  7. Pains in heart or chest.  .0  12  3 4  12  3 4  8. Feeling afraid in open spaces  0 1 2 3 4  9. Thoughts of ending your life. .  0  1 2 3 4  10. Feeling that most people cannot be trusted  .0 1 2 3 4  11. Poor appetite  0 12  3 4  12. Suddenly scared for no reason  0 12  3 4  13. Temper outbursts that you could not control..  0 12  14. Feeling lonely even when you are with people  0  15. Feeling blockedtagetting things done 16. Feeling lonely. 17. Feelingblue.  ...0  3 4  12-34 1 2 3 4  0 1 2 3 4 .........  ......  0 1 2 3 4  18. Feeling no interest In things.  .0 1 2  3 4  19. Feeling fearful.. .  0 1 2 3 4  20. Your feelings being easily hurt  0 12  3 4  21. Feeling that people are unfriendly or dislike you  0  12  3 4  12  3 4  22. Feeling inferior to others. 23. Nausea or upset stomach  .0  D 1 2 3 4  24. Feeling that you are watched or talked about by others...... 0  1 2 3 4  25. Trouble falling asleep  .0 1 2 3 4  26. Having to check and doublecheck what you do......'  0 1 2 3 4  27. Difficulty making decisions..  0 1 2  28. Feeling afraid to travel on buses, subways, or trains........ 0 1 2 29. Trouble getting your breath.  34 3 4  0 1 23  4  249  Rosenberg S e l f - E s t e e m Using the s c a l e below as a guide, w r i t e a number beside each statement to i n d i c a t e how much you agree w i t h i t .  +  +  1  2  .  +  3  +  4  +  5  •  +  •  .6  +  7  extremely inaccurate  extremely accurate  1)  I f e e l t h a t I'm a person o f worth, a t l e a s t on equal b a s i s with o t h e r s .  2)  I f e e l t h a t I have a number o f good  3)  A l l i n a l l , I am i n c l i n e d  4)  I am a b l e to do things as w e l l as most other people.  5)  I f e e l I do not have much t o be proud o f .  6)  I take a p o s i t i v e a t t i t u d e toward m y s e l f .  7)  On the whole, I am s a t i s f i e d w i t h  8)  I wish I c o u l d have more r e s p e c t f o r myself.  9)  I c e r t a i n l y f e e l u s e l e s s at t i m e s .  10)  8  At times  qualities.  t o f e e l t h a t I am a f a i l u r e .  myself.  I t h i n k I am no good a t a l l .  250  BIDR Version 6 - Form 40A Using the scale below as a guide, write a number beside each statement to indicate how true it is.  +  1 not true  + 2  + 3  +  4 somewhat  +____ 5  +_  +  6  7 very true  1. My first impressions of people usually turn out to be right. 2. It would be hard for me to break any of my bad habits. 3.1 don't care to know what other people really think of me. 4.1 have not always been honest with myself. 5.1 always know why I like things. 6. When my emotions are aroused, it biases my thinking. 7. Once I've made up my mind, other people can seldom change my opinion. 8.1 am not a safe driver when I exceed the speed limit. 9.1 am fully in control of my own fate. 10. It's hard for me to shut off a disturbing thought. 11.1 never regret my decisions. 12.1 sometimes lose out on things because I can't make up my mind soon enough. 13. The reason I vote is because my vote can make a difference. 14. My parents were not always fair when they punished me. 15.1 am a completely rational person. 16.1 rarely appreciate criticism. 17.1 am very confident of my judgments 18.1 have sometimes doubted my ability as a lover. .. ' • 19. It's all right with me if some people happen to dislike me. 7  - 20.1 don't always know the reasons why I do the things I do.  ••  251  Using the scale below as a guide, write a number beside each statement to indicate how true it is.  +  1 not true  + 2  + 3  +  4 5 somewhat  +  +  6  +  7 very true  21.1 sometimes tell lies if I have to. 22.1 never cover up my mistakes. 23. There have been occasions when I have taken advantage of someone. 24.1 never swear. 25.1 sometimes try to get even rather than forgive and forget. 26.1 always obey laws, even if I'm unlikely to get caught. 27.1 have said something bad about a friend behind his/her back. 28. When I hear people talking privately, I avoid listening. 29.1 have received too much change from a salesperson without telling him or her. 30.1 always declare everything at customs. 31. When I was young I sometimes stole things. 32.1 have never dropped litter on the street. 33.1 sometimes drive faster than the speed limit. 34.1 never read sexy books or magazines. 35.1 have done things that I don't tell other people about. 36.1 never take things that don't belong to me. 37.1 have taken sick-leave from work or school even though I wasn't really sick. 38.1 have never damaged a library book or store merchandise without reporting it. 39.1 have some pretty awful habits. 40.1 don't gossip about other people's business.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0087680/manifest

Comment

Related Items