Structure of personality pathology in normal and clinical samples: Spanish validation of the DAPP-BQ

Journal of Personality Disorders, 22(4), 389–404, 2008 2008 The Guilford Press

STRUCTURE OF PERSONALITY PATHOLOGYIN NORMAL AND CLINICAL SAMPLES:SPANISH VALIDATION OF THE DAPP-BQ

Jose Alfonso Gutierrez-Zotes, PhD, Fernando Gutierrez, PhD,Joaquın Valero, MD, Emma Gallego, PhD, Eva Bailles, PhD,Xavier Torres, PhD, Antonio Labad, MD,and W. John Livesley, MD, PhD

Given that the DSM taxonomy of personality disorders is flawed by se-vere classificatory problems, the development of alternative classifica-tory systems, such as the Dimensional Assessment of Personality Pa-thology—Basic Questionnaire (DAPP-BQ), has now become a priority.This study examined the internal consistency, second-order factorstructure, and criterion validity of a Spanish translation of the DAPP-BQ in two samples: subjects with personality disorder (n = 155) andsubjects from the general population (n = 300). Alpha coefficientsranged satisfactorily from .75 to .93. Four second-order factors of Emo-tional Dysregulation, Dissocial Behavior, Inhibitedness, and Compul-sivity were obtained, which were replicable between samples and identi-cal to those reported in the literature. Finally, disordered subjectsscored significantly higher than normal subjects on 17 of the 18 DAPP-BQ traits. Some pending issues in the construction of an alternativetaxonomy of personality disorders are discussed.

The DSM Axis II is still the most widely used taxonomy for the descriptionand classification of personality disorders. Nonetheless, it is generallyagreed that this model is inadequate for classificatory purposes, due tosevere drawbacks in two main areas. First, the true nature of personalitytraits (whether normal or pathological) is not categorical but dimensional,and imposing arbitrary dichotomies produces both substantial informa-tion loss and classificatory problems (Clark, Livesley, & Morey, 1997; First

From Institut Pere Mata and Rovira I Virgili University, Reus, Tarragona (J. A. G.-Z., J. V.,E. G., A. L.); Hospital Clınic of Barcelona (F. G., E. B., X. T.); IDIBAPS (F. G.); Institut Univers-itari Dexeus (E. B., X. T.); and The University of British Columbia (W. J. L.).

This work was partially supported by grants from Spain’s Ministerio de Educacion y Ciencia(FIS 03/0464) awarded to F. Gutierrez and (F1S 06/0857) awarded to A. Labad.

Address correspondence to Fernando Gutierrez, Servei Psicologia, Institut Clınic Neurocien-cies, Hospital Clınic de Barcelona, Villarroel 170, 08036 Barcelona, Spain. E-mail: [email protected]

389

390 GUTIERREZ-ZOTES ET AL.

et al., 2002; Livesley, Schroeder, Jackson, & Jang, 1994; Widiger, 1993).Second, personality features are not really arranged in the ten diagnosticentities included in the DSM. Most diagnoses incorporate traits and symp-toms that do not show particular co-variation and which are shaped byseveral disorders. These heterogeneous entities, which are grouped ac-cording to expert opinion, present considerable content overlap and diag-nostic comorbidity, and do not replicate in multivariate analysis (Clarket al., 1997; Nurnberg, Woodbury, & Bogenschutz, 1999; Widiger, Trull,Clarkin, Sanderson, & Costa, 1994). Finding a more appropriate paradigmfor organizing personality pathology has thus become a priority issue(Livesley, 2003, 2005; McCrae, Lockenhoff, & Costa, 2005).

Livesley and colleagues have proposed an alternative personality taxon-omy, measured by the Dimensional Assessment of Personality Pathology—Basic Questionnaire (DAPP-BQ), which addresses these limitations. TheDAPP-BQ preserves the methodological strengths of dimensional person-ality inventories such as the NEO-PI-R (Costa & McCrae, 1992a) or the TCI(Cloninger, Przybeck, Svrakic, & Wetzel, 1994), but contains items assess-ing personality dysfunction. During its construction, a wide sample of de-scriptors of personality pathology was first identified based on extensivereviews of the literature, content analysis, and clinicians’ judgments(Livesley 1986, 1987). The authors then developed self-report scales tomeasure the resulting 100 descriptors, which were later refined and fac-tor-analyzed (Livesley, Jackson, & Schroeder, 1989, 1991, 1992). Thiseventuated in the development of 18 basic dimensions contained in theDAPP-BQ (Schroeder, Wormworth, & Livesley, 1992).

The DAPP-BQ dimensions capture a considerable amount of the vari-ance of the DSM individual personality disorders—between 29% and 63%,with a median of 44% (Bagge & Trull, 2003)—and are also consistentlyrelated to the dimensional models of Eysenck, Zuckerman, and Costa andMcCrae (Jang, Livesley, & Vernon, 1999; Larstone, Jang, Livesley, Vernon,& Wolf, 2002; Schroeder et al., 1992; Wang, Du, Wang, Livesley, & Jang,2004). The DAPP-BQ thus seems to provide the necessary bridge betweennormal and abnormal personality models (Livesley, 2005). However, a con-siderable amount of work remains to be done before it can be consideredan alternative to the current taxonomic system. The criteria that a classifi-catory system should fulfill (Clark et al., 1997; Livesley, 1998; Livesley &Jackson, 1992), as well as the requirements of its underlying theory (Costa& McCrae, 1992b; Eysenck, 1991; Zuckerman, 1992), have been dis-cussed elsewhere. Four fundamental issues still under study deserve spe-cial attention, and are raised in the present paper.

Firstly, constructs forming a suitable classificatory system should be in-ternally consistent, i.e., they should be formed by descriptors that co-varysufficiently. Unlike the DSM, the consistency of the 18 DAPP-BQ dimen-sions has been found to be quite good, with alpha coefficients generallyabove .80 or .90 (Clark, Livesley, Schroeder, & Irish, 1996; Schroeder etal., 1992; van Hiel, Mervielde, & de Fruyt, 2004; van Kampen, 2002). How-

STRUCTURE OF PERSONALITY PATHOLOGY 391

ever, moderate departures from these figures have also been reported, andso the issue requires further examination (Wang et al., 2004; Zheng et al.,2002).

Secondly, the system has to have a replicable and theoretically consis-tent factor structure. Four higher-order factors have been proposed as un-derlying the DAPP-BQ dimensions (Livesley, Jang, & Vernon, 1998): Emo-tional Dysregulation represents anxiousness, mood instability, andinsecure, submissive interpersonal behavior; Dissocial Behavior capturesa set of impulsive, uninhibited, and exploitative behaviors; Inhibitednessreflects restricted emotional expression, interpersonal inhibition, and re-duced social drive and enjoyment; and Compulsivity is a narrower factormarked by rigidity, and orderliness, and self-control. These four factorshave generally met with support (Bagge & Trull, 2003; Pukrop, Gentil,Steinbring, & Steinmeyer, 2001; van Hiel et al., 2004; van Kampen, 2002;Zheng et al., 2002) and seem to be similar across clinical and nonclinicalsamples, thus suggesting a structural continuity between normal andpathological personality (Livesley et al., 1998). However, there is not com-plete agreement on this point. For example, comparisons with the originaldimensions have generally been made by visual inspection (Bagge & Trull,2003; Pukrop et al., 2001; van Hiel et al., 2004). When some type of con-gruence analysis has been conducted, confirmation has been incomplete,either with respect to the original factors (van Kampen, 2002; Zheng et al.,2002) or between normal and pathological samples (Pukrop et al., 2001).Furthermore, a not inconsiderable number of studies have obtained five-factor structures (Clark et al., 1996; Goldner, Srikaweswaran, Schroeder,Livesley, & Birmingham, 1999; Jang et al., 1999; Larstone et al., 2002;Schroeder et al., 1992; Wang et al., 2004).

Thirdly, a taxonomic system designed to organize the personality pathol-ogy domain should be able to differentiate personality-disordered fromnormal subjects. To date, few authors have presented DAPP-BQ scores inhealthy populations, and those reported have been very disparate (Micha-lak et al., 2004; Wang et al., 2004; Reuber, Pukrop, Bauer, Derfuss, &Elger, 2004; Wang et al., 2005). Furthermore, the only two published stud-ies in personality-disordered subjects have not provided mean scores(Livesley et al., 1998), or have only done so for the higher-order dimensions(Pukrop, 2002), and no study has yet tested whether the 18 lower-ordertraits are able to differentiate the two populations.

A final point to note that unites the preceding three is that an acceptabletaxonomic system requires generalizability and maintain its propertiesacross different populations, languages, and cultures. German, Dutch,and Chinese adaptations have been reported for the DAPP-BQ, usuallywith satisfactory results (Pukrop et al., 2001; van Kampen, 2002; Zheng etal., 2002). However, internal consistency has rarely been studied in thesetranslations or has been unsuitable (Wang et al., 2004; Zheng et al., 2002),while factor replications have been incomplete (van Kampen, 2002; Zhenget al., 2002). In addition, if the DAPP-BQ seeks to differentiate pathological


from normal samples it should also maintain its psychometric propertiesin both populations, although the scores are expected to be different. Apartfrom the original study (Livesley et al., 1998), the factor structure of theDAPP-BQ has only been studied in two patient samples which reporteduneven results: Goldner et al. (1999) obtained a five-factor solution in 136eating disordered patients, while Pukrop et al. (2001) were unable to repli-cate the entire higher-order structure in 81 personality-disordered pa-tients.

Thus, the aim of the present study was to analyze whether the Spanishversion of the DAPP-BQ shows (a) suitable internal consistency, (b) a factorstructure in congruence with those reported by the authors and replicablebetween a clinical and a nonclinical sample, and (c) the ability to differenti-ate between healthy and personality-disordered subjects.

METHODSUBJECTS

The clinical sample comprised 155 outpatients (52.3% male) mean age34.3 years (SD 10.8; range 16–64), consecutively referred for personalityassessment to the Psychology Service of Barcelona’s Hospital Clınic.Though the major indication for referral was in every case a suspectedpersonality disorder, 29.0% of the clinical sample concurrently presenteda depressive disorder, 7.7% an anxiety disorder, and 17.5% other diagno-ses with frequencies below 5%. Conditions that might distort personalityprofiles were excluded: severe affective disorder, psychosis, and dementia,with a prevalence below 3% in our referred sample. These diagnoses weremade according to DSM-IV by the referring psychiatrist and again by anexperienced clinician (FG) through clinical interviews. With respect to AxisII, most outpatients (83.2%) received at least a diagnosis of personalitydisorder on the Personality Diagnostic Questionnaire-4+ (Hyler, 1994),though this study focused on the DAPP-BQ questionnaire alone. All per-sonality disorders were represented: paranoid (32.3%), schizoid (16.1%),schizotypal (27.1%), antisocial (5.8%), borderline (39.4%), histrionic (12.9%),narcissistic (7.7%), avoidant (48.4%), dependent (20.0%), obsessive-compul-sive (60.0%), passive-aggressive (20.0%), and depressive (54.2%), with amean of 3.5 (SD 2.9) disorders per subject. It should be noted, however,that self-reports such as the PDQ-4+ usually result in higher prevalencesthan do structured diagnostic interviews. The nonclinical sample com-prised 300 subjects (35.7% male) of mean age 27.9 years (SD 11.1; range18–66) and recruited from the general population. All subjects gave in-formed consent to participate in the study.

INSTRUMENTS

The Dimensional Assessment of Personality Pathology–Basic Question-naire (DAPP-BQ) is a 290-item, self-report questionnaire rated on a 5-point


scale ranging from 1 (very unlike me) to 5 (very like me). The DAPP-BQ18 dimensions: Submissiveness, Affective Lability, Anxiousness, InsecureAttachment, Cognitive Distortion, Identity Problems, Social Avoidance,Oppositionality, Narcissism, Stimulus Seeking, Callousness, Rejection,Conduct Problems, Restricted Expression, Intimacy Problems, Compulsiv-ity, Suspiciousness, and Self-Harm. The DAPP-BQ was translated into Span-ish by two independent teams of three translators each, who later reachedagreement on a common version. This version was backtranslated by anEnglish native speaker and compared to the original before the final ver-sion was completed.

DATA ANALYSIS

Descriptive statistics for the DAPP-BQ were obtained for both samples andcompared through Student’s t tests. Personality pathology scales were re-lated to age by means of Pearson’s r coefficients. Internal consistency wasassessed through Cronbach’s alpha coefficients and the correlations ofeach item with their corrected scale (ri-s). This second coefficient is irre-spective of scale length. To study the DAPP-BQ second-order factor struc-ture, a series of exploratory factor analyses (EFA) were performed usingtwo different extraction methods (principal components and common fac-tors) and retaining different numbers of factors (from two to six). As thefactors were nonorthogonal, oblimin rotations were applied. Usual rules-of-thumb for the number of factors (Kaiser, scree test, parallel test), aswell as criteria of simple structure, psychological interpretability, and rep-licability between samples were considered in order to retain the best solu-tion. Replicability was tested through congruence coefficients (Wrigley &Neuhaus, 1955). SPSS v.12 was used for all analyses.

RESULTSThe DAPP-BQ mean scores for both samples are shown in Table 1. Differ-ences were found for each of the 18 scales, the clinical sample having con-siderably more personality pathology. These differences were maintainedwhen the between-groups imbalance in age and gender was controlled for,with the exception of Compulsivity, on which the differences were attribut-able to age. Identity Problems, Anxiousness, and Affective Lability showedthe greatest differences between samples.

Alpha coefficients ranged from .78 to .93 (mean .87) in the clinical sam-ple and from .75 to .92 (mean .86) in the nonclinical sample (Table 2).There were no differences in internal consistency between samples. Cor-rected item-scale correlations (ri-s) were within the desired range, and itemsgenerally maintained the highest correlation with their theoretically as-signed scale.

Scale intercorrelations for the DAPP-BQ are reported in Table 3, for both


TABLE 1. Mean (SD) Scores on the DAPP-BQ Traitsand the Significance of the Differences Between

a Clinical and a Nonclinical Sample

Clinical Nonclinical(n = 155) (n = 300) p

Submissiveness 41.6 (11.1) 34.4 (9.7) .000Affective Lability 51.8 (13.7) 37.7 (11.1) .000Anxiousness 52.3 (15.3) 36.8 (12.7) .000Insecure Attachment 45.8 (14.0) 42.4 (11.7) .010Cognitive Distortion 38.2 (13.0) 29.1 (9.6) .000Identity Problems 47.9 (14.8) 30.1 (10.4) .000Social Avoidance 45.3 (13.6) 34.9 (11.3) .000Oppositionality 44.9 (13.3) 36.4 (10.0) .000Narcissism 45.8 (12.9) 38.2 (10.7) .000Stimulus Seeking 40.5 (11.0) 37.6 (9.8) .004Callousness 32.9 (9.9) 29.7 (7.8) .000Rejection 43.8 (10.5) 37.1 (9.5) .000Conduct Problems 27.5 (10.4) 23.9 (7.6) .000Restricted Expression 44.1 (12.7) 37.1 (10.0) .000Intimacy Problems 33.4 (9.6) 29.2 (7.1) .000Compulsivity 53.1 (13.0) 50.2 (9.7) .015a

Suspiciousness 32.9 (11.1) 27.9 (8.8) .000Self-Harm 22.6 (12.8) 13.9 (4.8) .000aSignificance is attributable to age differences between samples.

samples separately. Correlation coefficients were on average .07 higher inthe nonclinical sample. Affective Lability and Anxiousness showed associ-ations close to .80. High associations, in the range .50 to .79, were alsoobserved among distress-related scales: Affective Lability, Anxiousness,Insecure Attachment, Cognitive Distortion, Identity Problems, SocialAvoidance, Oppositionality, Narcissism, and Suspiciousness.

TABLE 2. Internal Reliability of the DAPP-BQ Traitsin a Clinical and a Nonclinical Sample: Cronbach’s �and Corrected Item-Scale Correlations (Median ri-s)

Clinical Nonclinical(n = 155) (n = 300)

� ri-s � ri-s

Submissiveness .84 .46 .87 .53Affective Lability .89 .59 .88 .53Anxiousness .93 .68 .92 .62Insecure Attachment .90 .62 .89 .56Cognitive Distortion .88 .52 .86 .51Identity Problems .91 .61 .89 .59Social Avoidance .89 .56 .89 .56Oppositionality .88 .55 .85 .51Narcissism .89 .57 .89 .57Stimulus Seeking .83 .47 .84 .48Callousness .82 .47 .80 .43Rejection .81 .42 .84 .45Conduct Problems .85 .50 .84 .46Restricted Expression .86 .51 .84 .44Intimacy Problems .78 .38 .75 .36Compulsivity .89 .54 .84 .45Suspiciousness .88 .58 .87 .55Self-Harm .93 .77 .89 .61

TABLE 3. Intercorrelations Among the DAPP-BQ Traits in a Clinical and a Nonclinical Sample

Subm

issi

ven

ess

Aff

ect.

Lab

ilit

y

Anxio

usn

ess

Inse

c.A

ttac

h.

Cogn

.D

isto

rtio

n

Iden

tity

Pro

b.

Soci

alA

void

.

Opposi

tion.

Nar

ciss

ism

Sti

m.See

kin

g

Cal

lousn

ess

Rej

ecti

on

Conduct

Pro

b.

Res

tr.E

xpre

ss.

Inti

mac

yPro

b.

Com

puls

ivit

y

Susp

icio

usn

ess

Sel

f-H

arm

Submissiveness — .55 .66 .49 .62 .59 .67 .56 .50 .20 .31 .20 .24 .31 .08 .01 .40 .25Affective Lability .34 — .83 .61 .75 .69 .58 .62 .70 .47 .47 .52 .45 .19 .05 .20 .66 .42Anxiousness .55 .77 — .57 .76 .79 .65 .64 .66 .35 .41 .38 .38 .30 .10 .13 .60 .44Insecure Attachment .41 .56 .49 — .54 .41 .46 .47 .59 .25 .25 .42 .29 .07 −.17 .25 .52 .25Cognitive Distortion .38 .68 .68 .38 — .72 .62 .65 .60 .43 .40 .38 .39 .37 .19 .08 .62 .48Identity Problems .46 .56 .75 .32 .64 — .69 .61 .56 .30 .41 .32 .35 .42 .31 .00 .52 .54Social Avoidance .55 .37 .55 .36 .48 .59 — .58 .59 .21 .49 .32 .31 .58 .21 .08 .53 .39Oppositionality .47 .47 .58 .30 .61 .56 .53 — .55 .47 .47 .39 .49 .28 .07 −.22 .41 .33Narcissism .36 .48 .40 .53 .32 .25 .41 .31 — .48 .55 .64 .45 .24 −.04 .20 .58 .35Stimulus Seeking .11 .40 .26 .29 .34 .14 .14 .35 .41 — .44 .39 .53 .17 −.03 −.08 .35 .15Callousness .12 .28 .22 .21 .35 .28 .29 .29 .42 .31 — .58 .62 .43 .20 .03 .51 .27Rejection −.06 .36 .14 .31 .19 .08 −.03 .03 .38 .34 .43 — .50 .16 −.03 .19 .50 .24Conduct Problems −.02 .38 .27 .16 .42 .22 .20 .39 .38 .46 .48 .47 — .20 .02 −.11 .37 .30Restricted Expression .29 .18 .34 .03 .39 .42 .58 .35 .05 .02 .25 −.07 .17 — .45 .08 .35 .18Intimacy Problems .21 .10 .24 −.10 .40 .41 .25 .27 −.20 −.04 .26 −.12 .08 .49 — −.01 .18 .22Compulsivity .15 .19 .17 .18 .10 .14 −.03 −.26 .05 −.21 −.08 .18 −.11 .09 .03 — .29 −.02Suspiciousness .33 .54 .54 .45 .57 .47 .38 .28 .37 .29 .49 .35 .37 .29 .26 .28 — .37Self Harm .26 .38 .45 .24 .52 .54 .27 .39 .07 .25 .19 .18 .31 .26 .31 .03 .34 —

Note. Clinical sample below the diagonal (p < .05 since .15; p < .01 since .21); nonclinical sample above the diagonal (p < .05 since .11; p < .01 since .14).Pearson’s r coefficients ≥ .40 are underlined.


In order to study the second-order structure, 17 DAPP-BQ scales wereseparately factor-analyzed in each sample after principal components ex-traction. Self-harm was seriously skewed and showed low communalities(h2 = .41 and .33), so it was excluded from the analysis. KMO coefficientswere .85 for the clinical and .91 for the nonclinical sample. The Kaiser-Guttman criterion and the parallel test suggested a four-factor solution,whereas the scree test indicated a four- or five-factor solution. Due to thearbitrary nature of these rules-of-thumb, between two and six factors weresuccessively retained, rotated to oblimin, and examined. Indeed, only thefour-factor solutions, explaining 68.6% and 73.5% of the total variancein the clinical and nonclinical samples, respectively, approached a simplestructure and were interpretable. Pattern structures for these solutionsare shown in Table 4.

A large first factor was found that comprised Submissiveness, Anxious-ness, Social Avoidance, Identity Problems, Insecure Attachment, Opposi-tionality, Affective Lability, Cognitive Distortion, and Narcissism. This is afactor of general distress, hyperemotionality, instability, and interpersonaluneasiness that has been previously referred to as Emotional Dysregula-tion. The second factor encompassed Intimacy Problems and RestrictedExpression, plus an opposite loading of around .40 for Insecure Attach-ment. It is characterized by social introversion, low attachment tenden-cies, reserve and, in short, schizoid tendencies, and has been labeled In-hibitedness. The third factor, Dissocial Behavior, included ConductProblems, Rejection, Callousness, Stimulus Seeking, and a secondary

TABLE 4. Second-Order Oblimin Factor Pattern of the DAPP-BQ in a Clinicaland a Nonclinical Sample

Clinical (n = 155) Nonclinical (n = 300)

F1 F2 F3 F4 h2 F1 F2 F3 F4 h2

Submissiveness .84 .01 .29 .01 .63 .92 .01 −.20 −.10 .70Anxiousness .83 −.08 −.06 .11 .78 .88 −.02 .03 .05 .82Social Avoidance .74 −.22 .06 −.16 .64 .75 .31 .00 .05 .74Identity Problems .69 −.35 −.05 .08 .70 .80 .25 .01 −.08 .76Insecure Attachment .68 .42 −.11 .20 .67 .68 −.36 .05 .29 .70Oppositionality .66 −.15 −.13 −.47 .76 .68 −.04 .31 −.36 .77Affective Lability .65 .11 −.33 .18 .71 .71 −.11 .25 .16 .80Cognitive Distortion .57 −.29 −.33 .04 .69 .79 .08 .11 .01 .75Narcissism .52 .44 −.36 −.04 .65 .50 −.11 .43 .23 .74Intimacy Problems .05 −.84 −.10 .05 .73 .00 .84 −.05 .00 .69Restricted Expression .29 −.67 −.05 −.01 .60 .18 .77 .12 .08 .72Conduct Problems −.03 −.09 −.82 −.19 .69 .01 .01 .84 −.17 .71Rejection −.15 .20 −.79 .30 .72 −.01 −.06 .74 .33 .71Callousness −.02 −.25 −.75 −.04 .61 −.01 .33 .78 .07 .75Stimulus Seeking .20 .25 −.56 −.35 .59 .06 −.10 .74 −.18 .60Suspiciousness .36 −.19 −.48 .37 .67 .41 .16 .31 .40 .67Compulsivity .10 −.06 .06 .89 .82 −.04 .04 −.10 .93 .84Sums of squared loadings 5.6 2.3 3.8 1.5 7.0 2.0 5.0 1.8

Note. Factors F2 and F3 are reverse oriented between the clinical and the nonclinical sample.Loadings ≥ .40 are underlined.


loading of Narcissism. It essentially reflects characteristics of low empa-thy, exploitation, interpersonal conflict, tough-mindedness, impulsivity,and need for stimulation. The fourth factor, including Compulsivity and anegative secondary loading of Oppositionality, is characterized by obses-sive perfectionism, orderliness and self-exigency. The Suspiciousnessscale maintained loadings over .30 on Emotional Dysregulation, DissocialBehavior, and Compulsivity, and changed its primary location from one toanother sample, as reported by Livesley et al. (1998). Emotional Dysregu-lation and Dissocial Behavior correlated .33 in the clinical and .48 in thenonclinical sample, whereas the remaining correlations were below .17.Congruence coefficients between samples for the homologous pairs of fac-tors were high (.99, .93, .97, and .92), with the nonhomologous coefficientsbeing .22 or below (mean of .12). Congruence seriously decreased, how-ever, when solutions below or above four factors were tested, indicatingthat these structures were not replicable.

A number of complementary analyses were conducted in order to ensurethe validity and generalizability of our solution. Firstly, as positive skewwas detected (two-fold above the standard error) for 9 DAPP-BQ scales inthe clinical sample and for 16 in the nonclinical sample, scales were re-factored after appropriate transformation (√x, ln x or 1/x). Only Self-harmscores in the nonclinical sample could not be normalized. Congruence co-efficients between the new factors and those in Table 3 were homoge-neously 1.00. Secondly, scales were reanalyzed after principal axes factor-ing extraction (PAF) instead of the widely used principal componentsanalysis (PCA). PAF is more appropriate when factors are to be interpretedas underlying dimensions and then generalized (Fabrigar, Wegener, Mac-Callum, & Strahan, 1999; Widaman, 1993). Solutions from both extrac-tion methods had congruence coefficients of between .99 and 1.00. Finally,our second-order structure and that of the original Canadian sample(complete matrix in van Kampen, 2002, p. 245) were compared. The con-gruence coefficients were .99, .92, .99, and .96 (clinical) and .99, .97, .97,and .83 (nonclinical), with a median of .95.

DISCUSSIONINTERNAL CONSISTENCY

The main purpose of the present study was to analyze the internal consis-tency, factor structure, and criterion validity of the Spanish version of theDAPP-BQ in both clinical and general populations. We found that theDAPP-BQ maintained its psychometric properties in a different languageand culture. Internal consistencies of the 18 traits ranged from .78 to .93in the clinical sample and from .75 to .92 in the normal sample, and wereequivalent to those published for the original version (Clark et al., 1996;Schroeder et al., 1992; van Hiel et al., 2004). Consistencies also equaledthose of the Dutch version (van Kampen, 2002), and did not encounter the


problems reported with the Chinese version (Wang et al., 2004; Zheng etal., 2002). Item-scale correlations were also within the desirable range(Comrey, 1988; Smith & McCarthy, 1995).

FACTOR ANALYSIS

The factor structure of the Spanish DAPP-BQ was equivalent between ourclinical and nonclinical samples. Congruences were high (.92 to .99), eventhough both matrices were independently rotated to oblimin, a more strin-gent similarity criterion than Procrustean rotations. This equivalence wasobserved in the original factor study (Livesley et al., 1998), but not in theonly published replication: Pukrop et al. (2001) reported only moderatecongruences, ranging from .70 to .80, for Dissocial Behavior, Inhibited-ness, and Compulsivity. Our finding of good replicability between sampleschallenges the notion that personality disorders form “qualitatively dis-tinct clinical syndromes” (American Psychiatric Association, 1994, p. 633)and provides additional support for the increasingly accepted view that allpersonality traits vary on a continuum from normalcy to pathology. In-deed, there are no pathological characteristics in our disordered subjectsthat are not present and organized identically in our healthy subjects, de-spite differences between the two groups in terms of the intensity of traits.

Furthermore, in both groups our results accurately replicated the origi-nal factors of Emotional Dysregulation, Dissocial Behavior, Inhibitedness,and Compulsivity obtained by Livesley et al. (1998). The median congru-ence between them was .95, and complementary analyses were unsuc-cessful in identifying further solutions that were both psychologically in-terpretable and replicable. Other replications of this structure, supportedby some type of congruence analysis, have been previously reported. Withthe Dutch version of the DAPP-BQ, van Kampen (2002) obtained excellentcongruences except for the Inhibitedness factor, which was only loadedby the Intimacy Problems trait. Zheng et al. (2002) obtained Pearson’s rcoefficients of .89 or above, and van Hiel et al. (2004) congruences rangingfrom .88 to .97 with the original Canadian sample. In those studies notreporting congruences (Bagge & Trull, 2003; Pukrop et al., 2001), visualcomparisons also revealed notable similarities.

FOUR DIMENSIONS OF PERSONALITY PATHOLOGY

These four factors make sense from a theoretical perspective, and haveshown in the literature consistent relationships with both normal person-ality models and the DSM personality disorders. Emotional Dysregulationis a broad factor of general distress—oversensitivity, strength, and perva-siveness of diverse negative emotions—which is ubiquitous in the person-ality literature (Austin & Deary, 2000; Claridge & Davis, 2001; Mulder &Joyce, 1997; Watson, Clark, & Harkness, 1994). It spans Neuroticism,


Harm Avoidance, and Negative Emotionality, but also additional compo-nents of Submissiveness, Identity Problems, Cognitive Dysregulation, In-secure Attachment, Oppositionality, Suspiciousness, and Narcissism,which stem from the psychopathological tradition and in fact go beyondthese constructs. In short, Emotional Dysregulation seems to capture alldimensions that are sufficiently laden with emotional distress. The find-ings published to date show that this factor is clearly aligned with Neuroti-cism in the EPQ-R, the NEO-PI-R, and the ZKPQ (Jang et al., 1999; Lar-stone et al., 2002; Schroeder et al., 1992; Wang et al., 2004). It alsounderlies most personality disorders, notably the Avoidant, Borderline,Dependent, and Depressive diagnoses (Bagby, Marshall, & Georgiades,2005; Bagge & Trull, 2003; Pukrop et al., 2001). According to Livesley andJang (2000), Anxiousness and Affective Lability are the core features ofEmotional Dysregulation. However, Submissiveness has shown equivalentloadings on this factor (Bagge & Trull, 2003; Goldner et al., 1999; Livesleyet al., 1998; van Hiel et al., 2004), which is not unexpected from an evolu-tionary perspective. Some authors have argued convincingly for the essen-tial—even etiological—role that status losses and submissive strategiesplay in negative emotionality (Gilbert & Allan, 1998; Sloman, Gilbert, &Hasey, 2003).

Dissocial Behavior is a factor of tough-mindedness, callousness, impul-sivity, and disinhibition. As expected, it is aligned with EPQ-Extraversion,ZKPQ-Impulsive Sensation Seeking, ZKPQ-Aggression/Hostility, and (low)NEO–Agreeableness (Jang et al., 1999; Larstone et al., 2002; Schroeder etal., 1992; Wang et al., 2004), and is associated with Antisocial (in therange .47 to .76), Narcissistic, and Paranoid disorders (Bagby et al., 2005;Bagge & Trull, 2003; Pukrop et al., 2001). Dissocial Behavior was indivisi-ble in our study, even when five or six factors were extracted. However, itsplits into two factors of Conduct Problems/Stimulus Seeking and Cal-lousness/Rejection whenever five-factor solutions have been reported, andalso when the DAPP-BQ has been factorized together with other instru-ments. This separation is characteristic of most normal personality mod-els, whereas personality pathology models usually put both componentstogether (e.g., Austin & Deary, 2000; Mulder & Joyce, 1997; O’Connor,2005).

Inhibitedness is a factor reflecting both low affiliative tendencies andrestricted emotional expression. It is negatively aligned with EPQ- andNEO-Extraversion, NEO-Openness, and ZKPQ-Sociability (Jang et al.,1999; Larstone et al., 2002; Schroeder et al., 1992; Wang et al., 2004).It also underlies the Schizoid and Avoidant personality disorders, and isnegatively aligned with the Narcissistic and Histrionic disorders (Bagby etal., 2005; Bagge & Trull, 2003; Pukrop et al., 2001). Inhibitedness hasbeen considered narrower and more specific than Introversion (Livesley &Jang, 2000).

Finally, Compulsivity reflects characteristics of orderliness, meticulous-


ness, and self-control. It is related to NEO-Conscientiousness (Larstone etal., 2002; Schroeder et al., 1992) and loads slightly on the ZKPQ-Activitydimension, a domain poorly captured by the DAPP-BQ (Wang et al., 2004).Its relationships to the Axis II disorders are unclear: it has positive but notimpressive correlations with Obsessive-Compulsive personality disorder(in the range .34 to .45), and negative correlations with the Antisocial dis-order (Bagby et al., 2005; Bagge & Trull, 2003; Pukrop et al., 2001). Com-pulsivity is considered too discrete to form a higher-order factor, so it isonly loaded by Compulsivity and (low) Oppositionality (Livesley & Jang,2000). However, a separate anankastic or compulsive factor has consis-tently emerged in abnormal personality research (Austin & Deary, 2000;Mulder & Joyce, 1997; O’Connor, 2005) and, under the labels of Con-straint, Conscientiousness, or Persistence, it also constitutes a higher-order dimension in most normal personality models (Costa & McCrae,1992a; Cloninger et al., 1994; Tellegen, 1985).

These higher-order factors are not exclusive to the DAPP-BQ. Four nota-bly similar factors, labeled the “four As” (Asthenic, Antisocial, Asocial, andAnankastic), have emerged independently from the factorization of DSMsymptoms and syndromes and have been put forward as the ultimate or-ganizing axes of personality pathology (Austin & Deary, 2000; Mulder &Joyce, 1997). This structure has received further support in a recent meta-analysis of 33 studies (O’Connor, 2005) in which the same factors of Neu-roticism, Low Agreeableness, Introversion, and Conscientiousness aroseirrespective of whether the model was FFM- or DSM-based, the samplewas clinical or nonclinical, or the information was obtained by interviewor self-report.

THE DAPP-BQ IN HEALTHY AND DISORDERED SUBJECTS

The third point in our analysis involved characterizing and comparing theDAPP-BQ profiles of normal and disordered subjects, a point on which theliterature offers little information. Our normal sample had similar scoresto those of the 287 nonclinical subjects of Michalak et al. (2004), with amean difference of z = .31 (range .03 to .76). However, notable differenceswere observed with regard to the other published studies (Reuber et al.,2004; Wang et al., 2004, 2005). For example, the Chinese subjects fromWang were strikingly closer to our clinical subjects than to our nonclinicalsubjects for most traits. The few studies carried out on personality-disor-dered subjects (Livesley et al., 1998; Pukrop, 2002) have not providedmean scores for the lower-order traits. Our study revealed differences for17 of the 18 DAPP-BQ dimensions between clinical subjects and the gen-eral population. The standardized (z) differences ranged from .29 (InsecureAttachment) to 1.81 (Self-Harm). As the DAPP-BQ aims to measure andorganize personality pathology, the ability to differentiate healthy subjectsfrom subjects seeking treatment for personality problems is an essential


component of its criterion validity. The exception was Compulsivity, a traitwith an unclear status in personality pathology taxonomies. Moderatecompulsive features are highly prevalent in the population (Nestadt et al.,1991) and, even when a obsessive-compulsive personality disorder can bediagnosed, it causes only a minimal degree of impairment: the risk in-creases for anxiety and decreases for substance abuse disorders, the inter-personal area is hardly affected, and work/academic performance mayeven improve (Costa, Samuels, Bagby, Daffin, & Norton, 2005). Thus, nor-mal subjects may be quite compulsive without seeking treatment and,conversely, a clinical sample may not be especially laden with dysfunc-tional compulsive features.

TOWARD AN ALTERNATIVE TAXONOMY

In short, in our study the DAPP-BQ has shown good reliability, a robustfactor structure, and adequate discrimination between healthy and disor-dered subjects in a different language and culture. However, if it is to beaccepted as a suitable alternative to the current taxonomy, several issuesrequire further attention. Firstly, although the DAPP-BQ explains consid-erable variance for most normal dimensions and DSM disorders, it cannotcapture the complete domain of normal and abnormal personality, anddimensions such as Conscientiousness-Persistence, Activity Level, Domi-nance, Positive Temperament, or Perceptual-Cognitive Aberration havegreater relevance or are more broadly represented in other models (e.g.,Bouchard & Loehlin, 2001; Buss & Plomin, 1984; Cloninger et al., 1994;Costa & McCrae, 1992a; Leary, 1957; Meehl, 1990; Siever & Davis, 1991).Secondly, the confirmation of the DAPP-BQ structure at the first-orderlevel is still pending, this being a far more problematic task than the sec-ond order for any personality model. Thirdly, an instrument measuringpsychopathology needs some criterion of disorder. On the one hand, theDAPP-BQ still lacks normative values which could indicate to what extenta trait is abnormally intense if compared with the reference population.On the other hand, intensity is neither the sole nor a better criterion fordetermining the presence of personality pathology, so an additional, inde-pendent criterion of dysfunction or problems in living will be required(Livesley & Jang, 2000; Parker & Barrett, 2000; Parker et al., 2002). Fi-nally, some issues concerning assessment strategy need to be resolved.The DAPP-BQ shares certain limitations with most self-report inventories.It assesses the intensity of a trait reliably, but not its persistence during alifetime or its pervasiveness, which are more easily assessed by interview.Additional clinician- or informant-based assessment forms would alsohelp to disentangle more effectively the confounding effects of present AxisI disorders (e.g., affective disorders), as well as the voluntary or involun-tary distortions many personality disordered subjects produce when theydescribe themselves.


REFERENCES

American Psychiatric Association. (1994). Di- PI-R). Odessa: Psychological Assess-ment Resources.agnostic and statistical manual of men-

tal disorders: DSM-IV. Washington, Costa, P. T., & McCrae, R. R. (1992b). Fourways five factors are basic. PersonalityDC: Author.

Austin, E. J., & Deary, I. J. (2000). The “four and Individual Differences, 13, 653–665.As”: A common framework for normal

and abnormal personality? Personality Costa, P. T., Samuels, J., Bagby, R. M.,Daffin, L., & Norton, H. (2005). Obses-and Individual Differences, 28, 977–

995. sive-Compulsive personality disorder:A review. In M. Maj, H. S. Akiskal,Bagby, R. M., Marshall, M. B., & Georgiades,

S. (2005). Dimensional personality J. E. Mezzich, & A. Okasha (Eds.), Per-sonality disorders. Chichester: Wiley.traits and the prediction of DSM-IV

personality disorder symptom counts Eysenck, H. J. (1991). Dimensions of per-sonality: 16, 5 or 3? Criteria for a taxo-in a nonclinical sample. Journal of Per-

sonality Disorders, 19, 53–67. nomic paradigm. Personality and Indi-vidual Differences, 12, 667–673.Bagge, C. L., & Trull, T. J. (2003). DAPP-BQ:

Factor structure and relations to per- Fabrigar, L. R., Wegener, D. T., MacCallum,R. C., & Strahan, E. J. (1999). Evaluat-sonality disorder symptoms in a non-

clinical sample. Journal of Personality ing the use of exploratory factor analy-sis in psychological research. Psycho-Disorders, 17, 19–32.

Bouchard Jr, T. J., & Loehlin, J. C. (2001). logical Methods, 4, 272–299.First, B. M., Bell, C., Cuthbert, B., Krystal,Genes, evolution and personality. Be-

havior Genetics, 31, 243–273. J. H., Malison, R., Offord, D. R., et al.(2002). Personality Disorders and Re-Buss, A. H., & Plomin, R. (1984). Tempera-

ment: Early developing personality traits. lational Disorders. In D. J. Kupfer,M. B. First, & D. A. Regier (Eds.), A Re-Hillsdale, NJ: Lawrence Erlbaum As-

sociates, Inc. search Agenda for DSM-V. Washing-ton: American Psychiatric Association.Claridge, G., & Davis, C. (2001). What’s the

use of neuroticism? Personality and In- Gilbert, P., & Allan, S. (1998). The role of de-feat and entrapment (arrested flight) individual Differences, 31, 383–400.

Clark, L. A., Livesley, W. J., & Morey, L. depression: An exploration of an evo-lutionary view. Psychological Medicine,(1997). Personality disorder assess-

ment: The challenge of construct valid- 28, 585–598.Goldner, E. M., Srikaweswaran, S., Schroeder,ity. Journal of Personality Disorders,

11, 205–231. M. L., Livesley, W. J., & Birmingham,C. L. (1999). Dimensional assessmentClark, L. A., Livesley, W. J., Schroeder, M. L.,

& Irish, S. L. (1996). Convergence of of personality pathology in patientswith eating disorders. Psychiatry Re-two systems for assessing specific

traits of personality disorder. Psycho- search, 85, 151–159.Hyler, S. E. (1994). PDQ-4+ Personality Ques-logical Assessment, 8, 294–303.

Cloninger, C. R., Przybeck, T. R., Svrakic, tionnaire. New York: Author.Jang, K. L., Livesley, W. J., & Vernon, P. A.D. M., & Wetzel, R. D. (1994). The tem-

perament and character inventory (1999). The relationship between Eys-enck’s P-E-N model of personality and(TCI): A guide to its development and

use. St. Louis, Missouri: Center for traits delineating personality disor-ders. Personality and Individual Differ-Psychobiology of Personality, Wash-

ington University. ences, 26, 121–128.Larstone, R. M., Jang, K. L., Livesley, W. J.,Comrey, A. L. (1988). Factor-analytic meth-

ods of scale development in personal- Vernon, P. A., & Wolf, H. (2002). Therelationship between Eysenck’s P-E-Nity and clinical psychology. Journal of

Consulting and Clinical Psychology, 56, model of personality, the five-factormodel of personality and traits deline-754–761.

Costa, P. T., & McCrae, R. R. (1992a). Re- ating personality disorder. Personalityand Individual Differences, 33, 25–37.vised NEO Personality Inventory (NEO


Leary, T. (1957). Interpersonal diagnosis of disorder: Implications for classifica-tion. Journal of Abnormal Psychology,personality. New York: Ronald Press.

Livesley, W. J. (1986). Trait and behavioral 103, 6–17.McCrae, R. R., Lockenhoff, C. E., & Costa,prototypes of personality disorder.

American Journal of Psychiatry, 143, P. T. (2005). A step toward DSM-V:Cataloguing personality-related prob-728–732.

Livesley, W. J. (1987). A systematic approach lems in living. European Journal of Per-sonality, 19, 269–286.to the delineation of personality disor-

ders. American Journal of Psychiatry, Meehl, P. E. (1990). Toward an integratedtheory of schizotaxia, schizotypy and144, 772–777.

Livesley, W. J. (1998). Suggestions for a schizophrenia. Journal of PersonalityDisorders, 4, 1–99.framework for an empirically based

classification of personality disorder. Michalak, E. E., Jang, K. L., Tam, E. M.,Yatham, L. N., Lam, R. W., & Livesley,Canadian Journal of Psychiatry, 43,

137–147. W. J. (2004). A comparison of person-ality function among patients withLivesley, W. J. (2003). Diagnostic dilemmas

in classifying personality disorders. In seasonal depression, nonseasonal de-pression, and nonclinical participants.K. A. Phillips, M. B. First, & H. A. Pin-

cus (Eds.), Advancing DSM dilemmas Journal of Personality Disorders, 18,448–458.in psychiatric diagnosis. Washington:

American Psychiatric Association. Mulder, R. T., & Joyce, P. R. (1997). Temper-ament and the structure of personalityLivesley, W. J. (2005). Behavioral and molec-

ular genetic contributions to a dimen- disorder symptoms. Psychological Medi-cine, 27, 99–106.sional classification of personality dis-

order. Journal of Personality Disorders, Nestadt, G., Romanoski, A. J., Brown, C. H.,Chahal, R., Merchant, A., Folstein,19, 131–155.

Livesley, W. J., & Jackson, D. (1992). Guide- M. F., et al. (1991). DSM-III compul-sive personality disorder: An epidemio-lines for developing, evaluating and re-

vising the classification of personality logic survey. Psychological Medicine,21, 461–471.disorders. Journal of Nervous and Men-

tal Disease, 180, 609–618. Nurnberg, H. G., Woodbury, M. A., & Bogen-schutz, M. P. (1999). A mathematicalLivesley, W. J., Jackson, D., & Schroeder, M.

(1989). A study on the factorial struc- typology analysis of DSM-III-R person-ality disorder classification: Grade ofture of personality pathology. Journal

of Personality Disorders, 3, 292–306. membership technique. Comprehen-sive Psychiatry, 40, 61–71.Livesley, W. J., Jackson, D. N., & Schroeder,

M. L. (1991). Dimensions of personal- O’Connor, B. P. (2005). A search for consen-sus on the dimensional structure ofity pathology. Canadian Journal of

Psychiatry, 36, 557–562. personality disorders. Journal of Clini-cal Psychology, 61, 323–345.Livesley, W. J., Jackson, D. N., & Schroeder,

M. L. (1992). Factorial structure of Parker, G., & Barret, E. (2000). Personalityand personality disorder: Current is-traits delineating personality disorders

in clinical and general population sues and directions [Editorial]. Psycho-logical Medicine, 30, 1–9.samples. Journal of Abnormal Psychol-

ogy, 101, 432–440. Parker, G., Both, L., Olley, A., Hadzi-Pavlovic,H., Irvine, P., & Jacobs, G. (2002). De-Livesley, W. J., & Jang, K. L. (2000). Toward

an empirically based classification of fining disordered personality function-ing. Journal of Personality Disorders,personality disorder. Journal of Per-

sonality Disorders, 14, 137–151. 16, 503–522.Pukrop, R. (2002). Dimensional personalityLivesley, W. J., Jang, K. L., & Vernon, P. A.

(1998). Phenotypic and genetic struc- profiles of borderline personality disor-ders in comparison with other person-ture of traits delineating personality

disorder. Archives of General Psychia- ality disorders and healthy controls.Journal of Personality Disorders, 16,try, 55, 941–947.

Livesley, W. J., Schroeder, M. L., Jackson, 135–147.Pukrop, R., Gentil, I., Steinbring, I., & Stein-D. N., & Jang, K. L. (1994). Categorical

distinctions in the study of personality meyer, E. (2001). Factorial structure of


the German version of the dimensional between the Zuckerman-KuhlmanPersonality Questionnaire and traitsassessment of personality pathology—

Basic questionnaire in clinical and delineating personality pathology. Per-sonality and Individual Differences, 36,nonclinical samples. Journal of Per-

sonality Disorders, 15, 450–456. 155–162.Wang, W., Yang, T., Zhu, H., Mao, F., Lives-Reuber, M., Pukrop, R., Bauer, J., Derfuss,

R., & Elger, C. E. (2004). Multidimen- ley, W. J., Larstone, R. M., et al.(2005). Disordered personality traits insional assessment of personality in pa-

tients with psychogenic non-epileptic primary headaches. Social Behaviorand Personality, 33, 495–502.seizures. Journal of Neurology, Neuro-

surgery and Psychiatry, 75, 743–748. Watson, D., Clark, L. A., & Harkness, A. R.(1994). Structures of personality andSchroeder, M. L., Wormworth, J. A., & Lives-

ley, W. J. (1992). Dimensions of per- their relevance to psychopathology.Journal of Abnormal Psychology, 103,sonality disorder and their relation-

ships to the big five dimensions of 18–31.Widaman, K. F. (1993). Common factor anal-personality. Psychological Assessment,

4, 47–53. ysis versus principal components anal-ysis: Differential bias in representingSiever, L. J., & Davis, K. L. (1991). A psycho-

biological perspective on the personal- model parameters? Multivariate Be-havioral Research, 28, 263–311.ity disorders. American Journal of Psy-

chiatry, 48, 1647–1658. Widiger, T. A. (1993). The DSM-III-R categor-ical personality disorder diagnoses: ASloman, L., Gilbert, P., & Hasey, G. (2003).

Evolved mechanisms in depression: critique and an alternative. Psychologi-cal Inquiry, 4, 75–90.the role and interaction of attachment

and social rank in depression. Journal Widiger, T. A., Trull, T. J., Clarkin, J. F.,Sanderson, C., & Costa, P. T. (1994). Aof Affective Disorders, 74, 107–121.

Smith, G. T., & McCarthy, D. M. (1995). description of the DSM-III-R and DSM-IV personality disorders with the five-Methodological considerations in the

refinement of clinical assessment in- factor model of personality. In P. T.Costa & T. A. Widiger (Eds.), Personal-struments. Psychological Assessment,

7, 300–308. ity disorders and the five-factor modelof personality. Washington, DC: Amer-Tellegen, A. (1985). Structures of mood and

personality and their relevance to as- ican Psychological Association.Wrigley, C. S., & Neuhaus, J. O. (1955). Asessing anxiety, with an emphasis on

self-report. In A. H. Tuma & J. D. Ma- comparison of component and factorpatterns: A Monte Carlo approach.ser (Eds.), Anxiety and the anxiety dis-

orders. Hillsdale, NJ: Erlbaum. Multivariate Behavioral Research, 17,371–388.van Hiel, A., Mervielde, I., & de Fruyt, F.

(2004). The relationship between mal- Zheng, W., Wang, W., Huang, Z., Sun, C.,Zhu, J., & Livesley, W. J. (2002). Theadaptive personality and right wing

ideology. Personality and Individual structure of traits delineating person-ality disorder in a Chinese sample.Differences, 36, 405–417.

van Kampen, D. (2002). The DAPP-BQ in the Journal of Personality Disorders, 16,477–486.Netherlands: Factor structure and re-

lationship with basic personality di- Zuckerman, M. (1992). What is a basic factorand which factors are basic? Turtlesmensions. Journal of Personality Disor-

ders, 16, 235–254. all the way down. Personality and Indi-vidual Differences, 13, 675–682.Wang, W., Du, W., Wang, Y., Livesley, W. J.,

& Jang, K. L. (2004). The relationship

Structure of personality pathology in normal and clinical samples: Spanish validation of the DAPP-BQ

Documents