See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225346482

The Personality Belief Questionnaire-Short Form: Development and Preliminary

Findings

Article  in  Cognitive Therapy and Research · June 2007

DOI: 10.1007/s10608-006-9041-x

CITATIONS READS
60 9,103

3 authors, including:

Andrew C Butler

21 PUBLICATIONS   3,293 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Andrew C Butler on 14 May 2014.

The user has requested enhancement of the downloaded file.

 Elsevier Editorial System(tm) for Behaviour Research and Therapy
Manuscript Draft

Manuscript Number:

Title: Beliefs in personality disorders: An overview of the Personality Beliefs Questionnaire

Article Type: Full Length Article

Keywords: Personality disorders; assessment; measures

Corresponding Author: Dr Sunil Singh Bhar, PhD

Corresponding Author's Institution: Swinburne University of Technology

First Author: Sunil Singh Bhar, PhD

Order of Authors: Sunil Singh Bhar, PhD; Aaron T Beck, MD; Andrew C Butler

Abstract: The Personality Beliefs Questionnaire (PBQ) is a 126-item self-report measure of beliefs
associated with personality disorders. This paper presents an overview of the measure's psychometric
properties, refinements and research applications. Across both non-clinical and psychiatric
populations, the PBQ has demonstrated high internal consistency and test-retest reliability. Concurrent
validity has been demonstrated in psychiatric samples for scales measuring beliefs of avoidant,
paranoid, obsessive-compulsive, narcissistic, borderline and dependent personality disorders. Factor
analysis has empirically supported a ten-factor structure of the PBQ for psychiatric patients. The
original PBQ has been expanded to include a scale measuring beliefs of borderline personality
disorder. A short form version has also been developed and validated. The sensitivity to change of the
PBQ has been demonstrated in treatment outcome research. Its application in such research has
subjected to empirical testing the theoretical conjectures of cognitive theory regarding the pivotal role
of beliefs in personality disorders. The PBQ is a promising instrument for identifying beliefs pertinent
to a range of personality disorders.
Covering letter

SSB 7/22/09 Page 1 of 1

May 5, 2010

Dear Professor Wilson

Please find attached a copy of the manuscript titled, “Beliefs in personality disorders: An
overview of the Personality Beliefs Questionnaire” that we are submitting to “Behaviour Research
and Therapy” for blind review and for publication.

The manuscript is 25 pages long and includes one table. The manuscript is original, not previously
published and not under consideration elsewhere.

My coauthors and I do not have interests that might be interpreted as influencing the research, and
APA ethical standards were followed in the conduct of the study.

I will be serving as the corresponding author for this manuscript. All the authors listed in the
byline have agreed to the byline order and to submission of the manuscript in the present form. I
have assumed responsibility for keeping my coauthors informed of our progress through the
editorial review process, the content of the reviews, and any revisions made.

Please feel free to contact me if you have any questions or concerns about the manuscript. We
appreciate your time and effort in reviewing this manuscript and we are looking forward to
hearing from you.

Sincerely,

Sunil S. Bhar, Ph.D.

Faculty of Life and Social Sciences

Swinburne University of Technology
Mail H24, PO Box 218,
Hawthorn, VIC 3122, Australia
Phone + 613 9214 8371
Fax + 613 9819 0821,
email [email protected]
*Manuscript
Click here to view linked References

Running head: Personality Beliefs Questionnaire

Beliefs in personality disorders: An overview of the Personality Beliefs Questionnaire

Sunil S. Bhar1, Aaron T. Beck2 & Andrew C. Butler3

1
Swinburne University of Technology, 2University of Pennsylvania, 3Private Practice

*Correspondence concerning this article should be addressed to Sunil S. Bhar, Faculty of

Life and Social Sciences, Swinburne University of Technology, Mail H24, PO Box 218

Hawthorn, VIC 3122, Australia. Fax + 613 9819 0821, Email: [email protected].

-------------------------------------------------------
 2

Abstract

The Personality Beliefs Questionnaire (PBQ) is a 126-item self-report measure of

beliefs associated with personality disorders. This paper presents an overview of the

measure’s psychometric properties, refinements and research applications. Across both

non-clinical and psychiatric populations, the PBQ has demonstrated high internal

consistency and test-retest reliability. Concurrent validity has been demonstrated in

psychiatric samples for scales measuring beliefs of avoidant, paranoid, obsessive-

compulsive, narcissistic, borderline and dependent personality disorders. Factor analysis

has empirically supported a ten-factor structure of the PBQ for psychiatric patients. The

original PBQ has been expanded to include a scale measuring beliefs of borderline

personality disorder. A short form version has also been developed and validated. The

sensitivity to change of the PBQ has been demonstrated in treatment outcome research.

Its application in such research has subjected to empirical testing the theoretical

conjectures of cognitive theory regarding the pivotal role of beliefs in personality

disorders. The PBQ is a promising instrument for identifying beliefs pertinent to a range

of personality disorders.

Key words: Personality disorders, assessment, measures

 3

Beliefs in personality disorders: An overview of the Personality Beliefs Questionnaire

A prominent feature of the cognitive theory of personality disorders is its

emphasis on the role of dysfunctional beliefs. According to this theory, each personality

disorder has a characteristic set of dysfunctional beliefs. The behavior patterns of the

different personality disorders are viewed as overt manifestations of the underlying

cognitive structures (Beck, Freeman, Davis, & Associates, 2004). In 1990, Beck and

colleagues, proposed a set of cognitive features believed to represent the dysfunctional

beliefs characteristic of each DSM-III-R personality disorders (Beck, Freeman, &

Associates, 1990). For example, the main beliefs purported to explain the behavior of

patients with avoidant personality disorder were “I am socially inept and socially

undesirable in work or social situations” and “I cannot tolerate unpleasant feelings”,

while those typical of paranoid personality patients included “People will take advantage

of me if I give them the chance” and “I have to be on guard at all times.”

The assessment of beliefs is an important component of cognitive therapy of

personality disorders. Dysfunctional beliefs form the central component of cognitive case

conceptualizations and are prime targets for intervention. When correctly identified, key

dysfunctional beliefs reflect one or more conceptual themes that link a patient’s

developmental history, compensatory strategies and dysfunctional reactions to current

situations. As therapist and patient work together to identify and modify these key

beliefs, improvements may be seen simultaneously across many areas of functioning.

These cognitive features are purported to constitute a primary focus and mechanism of

change in cognitive interventions of personality disorders.

 4

Further, the assessment of such beliefs may also serve a diagnostic function. The

identification of such beliefs arguably form an important source of data for the

differential diagnosis of personality disorders listed in the Diagnostic and Statistical

Manual-IV-Text Revised (DSM-IV-TR) nosology (American Psychiatric Association,

2000). The content of DSM criteria sets, including most of those for personality disorders

primarily consist of behavioral indicators, such as avoidance behavior (e.g., “avoids

occupational activities that involve significant interpersonal contact”), unhelpful

interpersonal styles (e.g., “bears grudges”), situational variables (e.g., “lacks close

friends) and emotional states (e.g., “affective instability”). Although DSM’s criteria for

personality disorders are defined in largely behavioral terms, many criteria refer to beliefs

that may explain the reason for the behaviors. For example, one behavioral criterion for

dependent personality disorder is “difficulty expressing disagreement with others” which

is explained as the result of one’s fear of “loss of support or approval.” Thus, an

assessment of beliefs can purportedly be useful for identifying the presence of such fears,

which in turn can help establish whether the individual meets the behavioral criteria for

the personality disorder.

A number of self-report measures have been developed to assess beliefs

associated with personality disorders including the Dysfunctional attitude Scale ((DAS:

Beck, Brown, Steer, & Weissman, 1991), the Young Schema Questionnaire (YSQ:

Young, 1994) and the Personality Disorder Beliefs Questionnaire (PDBQ: Arntz, Dietzel,

& Dreessen, 1999; Arntz, Dreessen, Schouten, & Weertman, 2004). However, these self-

report measures are limited because they (a) include a mixture of dysfunctional beliefs

and behavior patterns (e.g., YSQ), (b) were not developed to correspond directly with
 5

DSM-IV personality disorders (e.g., YSQ, DAS), (c) have not yet been validated for a

number of DSM-IV personality disorders (e.g., PBDQ).

To address these limitations, Beck and Beck (1991) developed the Personality

Belief Questionnaire (PBQ) to measure the beliefs associated with ten DSM personality

disorders. The PBQ is a 126-item self report measure of beliefs purported by cognitive

theory to be relevant to the following DSM-IV personality disorders: Avoidant,

dependent, obsessive-compulsive, histrionic, passive–aggressive, narcissistic, paranoid,

schizoid, antisocial and borderline.1 The items of the PBQ were based on the clinical

observation of Beck et al (1990).

The present paper provides an overview of the psychometric properties of the

PBQ, its refinements and its applications in treatment outcome research. Given that

nearly 20 years have passed since its development it is timely to consider the

performance of the scales across non-clinical and clinical research studies. The current

paper has three aims: The first is to review the reliability and validity of the PBQ across

these studies and populations. The second is to describe the developments to the PBQ

since its original version. Since its inception in 1991, two refinements to the scale have

been made: One, as noted, an additional scale measuring beliefs of borderline personality

disorder has been identified. Two, a short form of the questionnaire has been developed..

Third, as the PBQ has now been used in several treatment studies, there is an opportunity

to examine the measure’s sensitivity to change, as well as the accumulated evidence for

whether beliefs inform the outcomes and change process in cognitive therapy of

personality disorders, as would be predicted by cognitive theory.

1
The PBQ originally comprised nine scales – items measuring BPD beliefs were later identified from these
nine scales (Butler, Brown, Beck, & Grisham, 2002). Items measuring beliefs corresponding to Schizotypal
Personality Disorder remain to be identified.
 6

Psychometric Properties of the PBQ

Reliability and Validity

Five studies have focused on investigating the reliability and validity of the PBQ.

These studies have demonstrated considerable support for the internal consistency,

stability, criterion validity and construct validity of the measure. In this section, we

review these psychometric properties of the PBQ.

The first study to examine the psychometric properties of the PBQ was conducted

by Trull and colleagues (Trull, Goodwin, Schopp, Hillenbrand, & Schuster, 1993). In this

study, the PBQ was administered to college students (n = 188, mean age = 19.74, SD =

3.73, 64.5% female). Good evidence of internal consistency was found for the PBQ

scales; Cronbach’s alphas (α) ranged from .77 to .93 (see Table 1). Test-retest correlation

coefficients over a one-month interval were high, ranging from .63 (passive-aggressive

scale) to .82 (paranoid scale) with a median of .75. However, the evidence for validity in

this non-clinical sample was less compelling. Modest correlations were obtained between

the PBQ and measures of personality disorders such as the Personality Disorder

Questionnaire-Revised (PDQ-R: Hyler, Skodol, Oldham, Kellman, & Doidge, 1992) and

the Minnesota Multiphasic Personality Inventory-Personality Disorders (Morey, Waugh,

& Blashfield, 1985). These results questioned the criterion validity of the PBQ for non-

clinical PD traits.

However, as noted by Beck et al. (2001), given that the PBQ was designed for use

with psychiatric patients, tests of criterion should evaluate how it performs with its

intended population, rather than with non-clinical individuals. Therefore, in the second

and largest psychometric study of the PBQ, Beck and colleagues (2001) employed a
 7

sample of 756 adult psychiatric outpatients. Due to limited sample sizes for some Axis II

disorders, they focused their investigation on five Axis II diagnoses: Avoidant,

dependent, obsessive-compulsive, narcissistic and paranoid personality disorders (mean

age = 34.73, SD = 11.46; 53% female). The reliability of the PBQ was adequate. The

PBQ scales had satisfactory internal consistency (alphas > .80) (see table 1). Test-retest

correlations for the scales were between .57 (avoidant scale) and .93 (antisocial scale) in

a subset of 15 patients over a period of eight weeks.

In this study, two findings supported the concurrent validity of the five PBQ

scales. First, patients scored higher on their corresponding PBQ scale than on other PBQ

scales. For example, avoidant patients scored significantly higher on the PBQ avoidant

scale than on the dependent, obsessive-compulsive, narcissistic or paranoid scales.

Second, for most comparisons, the highest score on a PBQ scale was obtained by patients

with the clinically diagnosed corresponding personality disorder, compared to patients

with other Axis II disorders. For example, patients with dependent personality disorder

scored higher on the dependent scale than patients with avoidant, obsessive-compulsive,

narcissistic, paranoid or no personality disorder. Exceptions to such findings were with

respect to the obsessive-compulsive and paranoid scales. In both instances, patients with

narcissistic personality disorder scored as highly as did patients with the criterion

personality disorders. However, when analyses were conducted on how well each of the

five PBQ scales discriminated its criterion group from the collection of remaining

personality disorders, the researchers found that across all comparisons, patients with the

criterion personality disorder scored higher on the corresponding belief scale than did the
 8

collection of these other patients. These findings provided support for the criterion

validity of the five PBQ scales when applied to clinical populations.

Given that only five scales were validated by Beck and colleagues (2001), they

recognized that additional studies were needed to validate the PBQ with other personality

disorders. In response to this gap in the literature, a third study was conducted on the

psychometrics of the PBQ. Jones, Burrell-Hodgson and Tate (2007) explored the

criterion validity of three other PBQ scales (passive-aggressive, schizoid, borderline2), as

well as the avoidant and dependent beliefs scales. The researchers found considerable

support for the criterion validity of these scales. Using stepwise regression analyses with

a sample of 164 psychiatric outpatients (mean age = 37.62, SD = 11.95, 60% female),

they examined the association between these scales and the corresponding Axis II

diagnoses as identified by Millon Multiaxial Personality Inventory III (Millon, Davis, &

Millon, 1997). The dependent variable was group membership to a particular personality

disorder (present/absent coded 1 and 0 respectively). The predictors were the 5 specific

PBQ scales. These specific PBQ scales emerged as significant unique predictors for their

criterion personality disorders. For instance, the PBQ avoidant scale significantly

predicted group membership for avoidant personality disorder. Likewise, the PBQ

dependent scale significantly predicted group membership for dependent personality

disorder. Similar results were found for the PBQ passive-aggressive, schizoid and

borderline scales.

In a fourth study on the validation of the PBQ, McMurran and Christopher (2008)

examined the relationship between the PBQ antisocial scale and antisocial personality

disorder. They predicted that individuals with antisocial personality disorder would score
2
The development of the Borderline scale of the PBQ is described below.
 9

higher on the PBQ antisocial scale compared to other PBQ scales and individuals with no

personality disorder. They also predicted that the antisocial scale would best predict the

presence or absence of antisocial personality disorder. Participants (mean age = 33.0, SD

= 8.03; all men) were recruited from three prisons across Wales. Index offences were

violence (44%), acquisitive (25%) and dangerous driving (3%). Personality disorders

were diagnosed with the International Personality Disorder Examination (Loranger,

1999). Seventeen of the participants were diagnosed with antisocial personality only, 14

with antisocial personality plus another personality disorder, and 18 with no personality

disorder diagnosis (controls). The study found that compared to the controls, individuals

with diagnosis of antisocial personality disorder scored significantly higher on the

antisocial scale.

However, the diagnosed group did not score highest on the antisocial scale

compared to other scales on the PBQ. Further in a discriminant function analysis, the

researchers found that avoidant and paranoid scales of the PBQ were better

discriminators of antisocial personality disorder, than was the antisocial scale. These

findings suggest a potential weakness in the utility of the antisocial scale for identifying

individuals with antisocial personality disorders. McMurran and Christopher (2008)

suggest that such individuals may avoid admitting to antisocial beliefs in a deceitful

effort to manage impressions, particularly when incarcerated. Therefore, it is possible

that the face validity of the PBQ scale may compromise its utility with respect to

antisocial beliefs. Further research is required to investigate the utility of this scale in

other contexts with individuals diagnosed with antisocial personality disorder.

 10

In the fifth and most recent published psychometric study of the full PBQ,

Turkcapar and associates (Turkcapar, Orsel, Ugurlu, Sargin, Turhan, Akkoyunlu,

Hatiloglu, & Karakas, 2007) examined the psychometric properties of a Turkish version

of the PBQ in a non-clinical sample of 232 undergraduates. This translated version was

found to have good internal consistency (.67- .90, lowest for the avoidant scale, highest

for paranoid scale) and one-month test retest reliability (.65-.87) (Turkcapar et al., 2007).

The internal consistency found in this study was largely consistent with findings from

other research groups (Beck et al., 2001; Connan, Dhokia, Haslam, Mordant, Morgan,

Pandya, & Waller, 2009; Kuyken, Kurzer, DeRubeis, Beck, & Brown, 2001; Nelson-

Gray, Huprich, Kissling, & Ketchum, 2004) (see table 1).

Factorial Structure

Although there have not been any published studies on the factor structure of the

PBQ using clinical samples, there are to date, two published factor analytic studies using

non-clinical student samples (Trull et al., 1993; Turkcapar et al., 2007) . These studies

have produced a virtually identical two-factor structure for the PBQ. In both studies, the

first factor consisted of passive-aggressive, obsessive-compulsive, antisocial, narcissistic,

schizoid and paranoid scales, and the second factor consisted of avoidant and dependent

scales. Trull et al. suggested that the first factor reflected interpersonal dominance, while

the second, anxious attachment.

In contrast to these two studies, a recent unpublished study with 438 depressed

outpatients (mean age = 43 years, 59% female) found a very different factor structure for

the PBQ (Fournier, DeRubeis, & Beck, 2009). Evidence from this study suggested that

the content of 90 of the 126 items of the PBQ could be captured by 10 empirically
 11

identified components, a factor structure consistent with intended structure of PBQ. Thus,

there is some indication that the structure of the PBQ is different for non-clinical versus

clinical populations.

Developments to the PBQ

The Borderline Scale of the PBQ

A PBQ scale for borderline personality disorder (BPD) was not developed a

priori because the beliefs of BPD patients seemed to transcend the categorization of the

other personality disorders (Beck et al., 2001). Clinical experience indicated that BPD

patients endorsed numerous beliefs that were also characteristic of the other personality

disorders (Beck et al., 1990). Subsequent research with the PBQ confirmed that BPD

patients scored highly on virtually all of the PBQ scales (Butler et al., 2002).

However, a more fine-grained analysis found that BPD patients also preferentially

endorsed certain PBQ items that came from the PBQ dependent, paranoid, avoidant and

histrionic scales. Specifically, Butler et al. (2002) found that fourteen PBQ items

discriminated patients with BPD from patients with other personality disorders. After

cross-validating these findings in a separate sample, a composite scale was constructed

from the 14 items. BPD patients were found to score significantly higher on the newly

constructed PBQ borderline scale than on any other PBQ scale. Further, consistent with

the cognitive model of BPD, these items captured beliefs that were not only dysfunctional

but conflicting as well. The composite scale included items measuring dependency

beliefs (e.g., “I am needy and weak”) as well as distrust (e.g., “I cannot trust other

people”). Sensitivity of the PBQ-Borderline scale to treatment was demonstrated in a

study by Brown and colleagues (Brown, Newman, Charlesworth, Crits-Christoph, &

 12

Beck, 2004). They found significant reductions in borderline scale scores for BPD

patients who responded to cognitive therapy.

Subsequent analysis was conducted to examine the dimensional structure of the

PBQ-Borderline Scale (Bhar, Brown, & Beck, 2008). With a sample of 184 patients

diagnosed with borderline personality disorder (mean age = 33.1, SD = 10.47, 75%

females), exploratory factor analysis found that the 14 items in the PBQ-borderline scale

segregated into three distinct factors: The view of self as helpless and dependent

(Dependent factor, α = .87), a distrust of other people (Distrust factor, α = .87), and

beliefs about the need to act preemptively in order to guard against rejection and distress

(Protection factor, α = .75). The three scales showed discriminant validity with respect to

risk indicators for suicide – depression, hopelessness and suicide ideation. Of the three

factors, distrust was the only significant correlate of suicide ideation (r = .35).

Dependency and distrust were both significantly associated with hopelessness (rs = .30

and .39, respectively), while all factors related significantly to depression (rs = .20 to

.41). Thus, the PBQ allows for the assessment of various beliefs associated with BPD,

and can augment a cognitive formulation of the range of difficulties presented by patients

with BPD.

The Short Form Version of the PBQ

An abbreviated version of the PBQ was recently developed to provide clinicians

and researchers with a brief measure of personality disorder beliefs (Butler, Beck, &

Cohen, 2007). The development for the PBQ – Short Form (PBQ-SF) proceeded in two

stages. In the first stage, archival data from 920 adult psychiatric outpatients (mean age =

36.4, SD = 11.1, 55% female) were used to construct experimental scales comprised of
 13

the 7 PBQ items with the highest item-total correlations for each PBQ scale. In this

sample, there were sufficient numbers of patients with personality disorders to examine

the criterion validity of five PBQ scales: avoidant (n = 79), dependent (n = 26),

obsessive-compulsive (n = 58), narcissistic (n = 26) and paranoid (n = 27). In the second

stage, the experimental scales were administered to a new sample of psychiatric

outpatients (n = 160, mean age = 39.8, SD = 14.2, 58% female), and the reliability and

construct validity of the PBQ-SF were evaluated in this new independent sample.

Results from the first stage showed that the experimental scales had good internal

consistency (see table 1). Patients with the criterion personality disorder tended to score

higher on the corresponding PBQ-SF scale, compared to patients with other personality

disorders, or no personality disorders. For example, patients with dependent personality

disorder scored higher than patients with other personality disorders or no personality

disorders, on the experimental PBQ dependent scale. In only 4 such comparisons were

results non significant. Further, results from within-group analyses showed that the five

personality disorder groups scored higher on their corresponding experimental scale than

on alternative experimental scales.

In stage 2, the researchers examined the internal consistency, test-retest reliability

and construct validity for these scales in an independent sample of psychiatric patients.

Cronbach’s alpha coefficients ranged from .81 (for the avoidant and narcissistic scales)

to.92 (for the paranoid scale; see table 1). Test retest correlation over a 4 week interval

ranged from .57 (Antisocial scale) to .82 (Obsessive-compulsive scale). As expected all

nine PBQ-SF scales correlated in theoretically consistent ways with other clinical

variables including depression, anxiety, dysfunctional attitudes, neuroticism, self-esteem,

 14

and psychosocial functioning. For example, the PBQ-SF scale for avoidant personality

disorder correlated negatively with a measure of self-esteem, and positively with

measures of anxiety, depression and depression-related attitudes. The PBQ-SF scale for

narcissistic personality disorder correlated with the same variables, but in the opposite

direction. The researchers concluded that the PBQ-SF appears to be a practical alternative

as a measure of personality disorder beliefs when it is not feasible to use the longer PBQ.

Applications of the PBQ in Research

Comorbidity Research

Using the PBQ, Connan and colleagues (Connan et al., 2009) examined the

personality disorder beliefs associated with eating disorders. The authors noted that

individuals with eating disorders frequently meet diagnostic criteria for Axis II disorders,

in particular cluster B and C personality disorders. They found that the PBQ beliefs that

were most relevant to eating disorder pathology were those relating to avoidance and

obsessive-compulsive personality disorder. They suggested that these beliefs might

account for the comorbidity between eating disorders and those specific personality

disorders.

Treatment Outcome Research

Ng (2005) used the PBQ to assess the efficacy of cognitive therapy for outpatients

with refractory depression and obsessive compulsive personality disorder (OCPD). All

patients completed the PBQ prior to commencement of cognitive therapy (at enrollment),

at the commencement of treatment (pre-treatment) and at the last session of treatment

(post-treatment). Ng found that post treatment scores on the PBQ obsessive-compulsive

scale were significantly lower compared to scores at enrollment and pre-treatment.

 15

Further, Ng found that the treatment was also successful at significantly reducing the

severity of OCPD symptoms. Thus, in addition to lending support for the efficacy of

cognitive therapy for OCPD, Ng’s study demonstrates that the PQB obsessive-

compulsive scale is sensitive to change.

Brown and colleagues (Brown et al., 2004) used the 14 item borderline scale from

the PBQ in an investigation of the efficacy of cognitive therapy for borderline personality

disorders. A total of 32 patients with borderline personality disorders, who reported

suicide ideation or engaged in self-injury behavior, received weekly cognitive therapy

sessions, as described by Layden, Newman, Freeman and Morse (1993). The results

showed significant and clinically important decreases on the number of borderline

symptoms and dysfunctional beliefs at termination and 18 month follow up assessment.

These results substantiate the sensitivity to change of the PBQ borderline beliefs scale.

Kuyken and colleagues (Kuyken et al., 2001) examined whether personality

disorder beliefs predicted outcomes for cognitive therapy. The outcomes of interest were

depression (as measured by the Beck Depression Inventory-II) and clinician rated global

functioning (GAF rating). In a naturalistic study, 162 depressed outpatients (57%

women, mean age = 33.61, SD = 11.91) with and without a personality disorder were

followed over the course of cognitive therapy. The researchers found that personality

disorder status did not predict response to therapy; however, beliefs associated with

avoidant and paranoid personality disorders predicted variance in outcome. More

specifically, high scores on the PBQ avoidant scale was predictive of greater severity of

depressive symptoms at termination of treatment, while high scores on the PBQ paranoid

scale were predictive of poorer global functioning (i.e., GAF scores) at termination. The
 16

authors of this study suggest that such results demonstrate the moderating role of these

beliefs in the change process.

Summary and Future Directions

The PBQ measures beliefs that are hypothesized to relate to specific DSM

personality disorders. This article was intended to provide researchers and clinicians with

an up-to-date overview of the psychometric properties, developments and applications of

the PBQ in exploring the role of beliefs in personality disorders. .

As shown, the reliability of the PBQ is consistently high in psychiatric and non-

psychiatric samples. Internal consistency estimates ranged from .77 to .94 and test-retest

correlations were found to be greater than .50 in both psychiatric and non-psychiatric

samples. The criterion validity of PBQ was demonstrated in psychiatric samples and

particularly for six of the ten PBQ scales – namely, the borderline, avoidant, dependent,

narcissistic, paranoid and obsessive-compulsive scales (Beck et al., 2001; Butler et al.,

2007; Butler et al., 2002). With few exceptions, the results from these studies

demonstrate specificity in the relationship between beliefs measured by the various PBQ

scales and their corresponding personality disorders. Factorial validity of the PBQ in

psychiatric patients has yet to be established, but there is research in progress in which

the theoretical structure appears to be confirmed (Fournier et al., 2009).

Since the inception of the PBQ in 1991, an additional subscale - that is the

borderline personality beliefs scale – and a short form version of the PBQ have been

developed. These additions have increased the applicability of the measure in two ways.

First, by empirically identifying 14 items that discriminate patients with borderline

personality disorder from those with other personality disorders, the PBQ can be used to
 17

assess beliefs specific to this diagnostic group. Second, the development of an

abbreviated version of the PBQ provides a practical alternative as a measure of

personality disorder beliefs when the full PBQ cannot be used.

The PBQ has been used in four studies, one investigating cognitive basis of

comorbidity between personality disorders and eating disorders, and three on treatment

outcomes. In these latter studies, the PBQ has demonstrated its sensitivity to change in

theoretically consistent ways. Scores on PBQ obsessive-compulsive scale and borderline

scale were significantly reduced for individuals treated for those disorders (Brown et al.,

2004; Ng, 2005). In addition, PBQ avoidant and paranoid scales were found to predict

changes in depressive symptoms and general functioning respectively over the course of

cognitive treatment for depression (Kuyken et al., 2001). These results suggest that the

PBQ is helpful in tracking and predicting cognitive treatment outcomes. As predicted by

cognitive theory, the results also support the proposal that beliefs inform the outcomes

and change process in cognitive therapy.

While some research shows that the beliefs measured by the PBQ are modified

with treatment (Brown et al., 2004; Ng, 2005), more research is required to examine the

extent to which changes in such beliefs mediate the outcomes of cognitive therapy of

personality disorders. For instance, Brown and colleagues (p. 265) found only “small or

negligible” associations between the PBQ borderline scale and number of borderline

criteria. Perhaps as suggested by Bhar et al. (2008), more research attention could be

directed to the relationships between subsets of beliefs within that scale and BPD

symptoms and related psychopathology. Thus, more research is needed to elucidate the
 18

relationships between specific factors within the various PBQ scales and specific

symptoms of personality disorders.

The applicability of the PBQ for non-clinical individuals also requires further

examination. There is evidence that the PBQ is more applicable for psychiatric patients,

than non-clinical individuals. Although in both populations, the internal consistency and

test-retest reliability of the PBQ is strong, its validity and factorial structure may be

different in non-clinical populations compared to psychiatric populations. Current

findings suggest that while the PBQ demonstrates high levels of validity with psychiatric

patients, its applicability in non-clinical populations is less well established.

Like other self-report measures, the PBQ relies exclusively on self-report data.

Some research has suggested that the exclusive reliance on self-report data for measuring

personality functioning is limited. As reviewed in Oltmanns and Turkheimer (2009) peer

reports of personality disorders have been found to better predict adverse outcomes for

individuals with dysfunctional personality traits. Such findings have advocated that

investigators consider data from informants in order to reach a more complete description

of personality disorders and functioning. However, while the support is compelling for

the incremental utility of informant data over self report in predicting adverse outcomes,

the relative validity of self vs. peer data for identifying private internal phenomena such

as beliefs remains to be subjected to empirical investigation.

Finally, there is very limited research on the psychometrics of certain PBQ scales.

Only one study has examined the validity of the antisocial scale (McMurran &

Christopher, 2008) and of the passive-aggressive and schizoid scales (Jones et al., 2007)

respectively. No study to date has examined the validity of the histrionic scale. Further
 19

research is needed to explore whether these scales specifically relate to individuals with

these criterion personality disorders.

In summary, empirical investigations of the PBQ have found it to be a promising

self-report measure of beliefs characteristic of several personality disorders. Various PBQ

scales demonstrate good internal consistency, test-retest reliability, and in psychiatric

samples, adequate levels of validity. The development of a borderline scale and short

form version of the measure has further widened the applicability of the PBQ. Its

applicability in treatment outcome research is demonstrated by studies showing that the

measure includes scales that are sensitive to treatment related changes and predictive of

treatment related outcomes.

Acknowledgements

This research was supported by the National Institute of Mental Health (P30 MH45178).
 20

References

American Psychiatric Association. (2000). Diagnostic and statistical manual - Text

revision (DSM-IV-TR) (4 ed.). Washington, DC: American Psychiatric

Association.

Arntz, A., Dietzel, R., & Dreessen, L. (1999). Assumptions in borderline personality

disorder: Specificity, stability and relationship with etiological factors. Behaviour

Research and Therapy, 37, 545-557.

Arntz, A., Dreessen, L., Schouten, E., & Weertman, A. (2004). Beliefs in personality

disorders: A test with the Personality Disorder Belief Questionnaire. Behaviour

Research and Therapy, 42, 1215-1225.

Beck, A. T., & Beck, J. S. (1991). The Personality Belief Questionnaire. The Beck

Institute for Cognitive Therapy and Research.

Beck, A. T., Brown, G., Steer, R. A., & Weissman, A. N. (1991). Factor analysis of the

Dysfunctional Attitude Scale in a clinical population. Psychological Assessment:

A Journal of Consulting and Clinical Psychology, 3, 478-483.

Beck, A. T., Butler, A. C., Brown, G. K., Dahlsgaard, K. K., Newman, C. F., & Beck, J.

S. (2001). Dysfunctional beliefs discriminate personality disorders. Behaviour

Research & Therapy, 39, 1213-1225.

Beck, A. T., Freeman, A., & Associates. (1990). Cognitive therapy of personality

disorders. New York: The Guildford Press.

Beck, A. T., Freeman, A., Davis, D. D., & Associates. (2004). Cognitive therapy of

personality disorders (2nd ed.). New York: The Guildford Press.

 21

Bhar, S. S., Brown, G. K., & Beck, A. T. (2008). Dysfunctional beliefs and

psychopathology in borderline personality disorder. Journal of Personality

Disorders, 22, 165-177.

Brown, G. K., Newman, C. F., Charlesworth, S. E., Crits-Christoph, P., & Beck, A. T.

(2004). An Open Clinical Trial of Cognitive Therapy for Borderline Personality

Disorder. Journal of Personality Disorders, 18, 257-271.

Butler, A. C., Beck, A. T., & Cohen, L. H. (2007). The Personality Belief Questionnaire-

Short Form: Development and preliminary findings. Cognitive Therapy and

Research, 31, 357-370.

Butler, A. C., Brown, G. K., Beck, A. T., & Grisham, J. R. (2002). Assessment of

dysfunctional beliefs in borderline personality disorder. Behaviour Research &

Therapy, 40, 1231-1240.

Connan, F., Dhokia, R., Haslam, M., Mordant, N., Morgan, G., Pandya, C., & Waller, G.

(2009). Personality disorder cognitions in the eating disorders. Behaviour

Research and Therapy, 47, 77-82.

Fournier, J. C., DeRubeis, R. J., & Beck, A. T. (2009). The structure and validity of the

Personality Belief Questionnaire. Manuscript in preparation.

Hyler, S. E., Skodol, A., Oldham, J., Kellman, H., & Doidge, N. (1992). Validity of the

Personality Diagnostic Questionnaire--Revised: A replication in an outpatient

sample. Comprehensive Psychiatry, 33, 73-77.

Jones, S. H., Burrell-Hodgson, G., & Tate, G. (2007). Relationships between the

Personality Beliefs Questionnaire and self-rated personality disorders. British

Journal of Clinical Psychology, 46, 247-251.

 22

Kuyken, W., Kurzer, N., DeRubeis, R. J., Beck, A. T., & Brown, G. K. (2001). Response

to cognitive therapy in depression: The role of maladaptive beliefs and personality

disorders. Journal of Consulting and Clinical Psychology, 69, 560-566.

Layden, M. A., Newman, C. F., Freeman, A., & Morse, S. B. (1993). Cognitive therapy

of borderline personality disorder. Needham Heights, MA: Allyn & Bacon.

Loranger, A. W. (1999). International personality disorders examination (IPDE).

Odessa, FL: Psychological Assessment Resources.

McMurran, M., & Christopher, G. (2008). Dysfunctional beliefs and antisocial

personality disorder. Journal of Forensic Psychiatry and Psychology, 19, 533-

542.

Millon, T., E., Davis, R., & Millon, C. (1997). The Millon Multiaxial Clinical Inventory-

III. Mineapolis: Computer Systems.

Morey, L. C., Waugh, M. H., & Blashfield, R. K. (1985). MMPI Scales for DSM-III

Personality Disorders: Their Derivation and Correlates. Journal of Personality

Assessment, 49, 245.

Nelson-Gray, R. O., Huprich, S. K., Kissling, G. E., & Ketchum, K. (2004). A

preliminary examination of Beck's cognitive theory of personality disorders in

undergraduate analogues. Personality and Individual Differences, 36, 219-233.

Ng, R. M. K. (2005). Cognitive Therapy for Obsessive-compulsive Personality Disorder -

A Pilot Study in Hong Kong Chinese Patients. Hong Kong Journal of Psychiatry,

15, 50-53.

Oltmanns, T., F. , & Turkheimer, E. (2009). Person Perception and Personality

Pathology. Current Directions in Psychological Science, 18, 32-36.

 23

Trull, T. J., Goodwin, A. H., Schopp, L. H., Hillenbrand, T. L., & Schuster, T. (1993).

Psychometric properties of a cognitive measure of personality disorders. Journal

of Personality Assessment, 61, 536-546.

Turkcapar, M. H., Orsel, S., Ugurlu, M., Sargin, E., Turhan, M., Akkoyunlu, S.,

Hatiloglu, U., & Karakas, G. (2007). Reliability and validity of the turkish version

of personality belief questionnaire. Klinik Psikiyatri, 10, 177-191.

Young, J. (1994). Cognitive therapy for personality disorders: A schema-focussed

approach (Revised ed.). Sarasota: Professional Resource Press.

Table(s)

Table 1 Internal consistency (Cronbach’s alphas) of the PBQ

Study

Scale Trull et Beck et al., Kuyken et Nelson- Butler et al., Butler et al., Connan et

al., 19931 20012 al., 20013 Gray, 20044 2007 (phase 1)5 2007 (phase 2)6 al., 20097

Avoidant .83 .89 .86 .88 .84 .81 .91

Dependent .84 .90 .86 .94 .89 .89 .93

Passive-aggressive .77 .90 N/A NA .86 .85 .89

Obsessive-compulsive .86 .84 .88 .90 .90 .90 .91

Antisocial .87 .93 NA .85 .80 .85 .81

Narcissistic .85 .87 .84 .88 .83 .81 .84

Histrionic .82 .88 N/A .90 .89 .87 .90

Schizoid .79 .81 N/A .81 .79 .83 .83

Paranoid .93 .81 .93 .95 .91 .92 .94

Borderline N/A .898 N/A N/A N/A N/A .90

Note: N/A = not available, 1 = 188 non-clinical college undergraduate students, 2 = 756 psychiatric outpatients, 3 = Depressed
outpatients, 4 = Non depressed undergraduates diagnosed with PDs, 5 = 920 adult psychiatric outpatients, 6 = 160 adult psychiatric
outpatients, 7= 92 eating disorder patients, 8 = 84 outpatients diagnosed with Borderline Personality disorder (Butler et al., 2001).

View publication stats