This article was downloaded by: [University Of Pittsburgh]

On: 26 November 2014, At: 11:23

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Journal of Personality Assessment

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/hjpa20

The Role of Defense Mechanisms in Borderline and

Antisocial Personalities
a a b
Michelle D. Presniak , Trevor R. Olson & Michael Wm. MacGregor
a
Institute of Community and Family Psychiatry at SMBD Jewish General Hospital,
Department of Psychiatry , McGill University , Montreal, Quebec, Canada
b
Department of Psychology , University of Saskatchewan , Saskatoon, Saskatchewan,
Canada
Published online: 10 Feb 2010.

To cite this article: Michelle D. Presniak , Trevor R. Olson & Michael Wm. MacGregor (2010) The Role of Defense Mechanisms
in Borderline and Antisocial Personalities, Journal of Personality Assessment, 92:2, 137-145, DOI: 10.1080/00223890903510373

To link to this article: http://dx.doi.org/10.1080/00223890903510373

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Journal of Personality Assessment, 92(2), 137–145, 2010
Copyright
C Taylor & Francis Group, LLC
ISSN: 0022-3891 print / 1532-7752 online
DOI: 10.1080/00223890903510373

The Role of Defense Mechanisms in Borderline and Antisocial

Personalities
MICHELLE D. PRESNIAK,1 TREVOR R. OLSON,1 AND MICHAEL WM. MACGREGOR2

1
Institute of Community and Family Psychiatry at SMBD Jewish General Hospital, Department of Psychiatry, McGill University,
Montreal, Quebec, Canada
2
Department of Psychology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

We examined whether borderline personality disorder (BPD) and antisocial personality disorder (APD) could be differentiated based on defense
mechanisms as measured by observer (Defense-Q; MacGregor, Olson, Presniak, & Davidson, 2008) and self-report (Defense Style Questionnaire;
Andrews, Singh, & Bond, 1993) measures. We conducted 2 studies whereby nonclinical participants were divided into borderline and antisocial
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014

groups based on scores from the Personality Assessment Inventory (Morey, 1991). Multivariate analysis of variance results revealed significant
overall group differences in defense use. Univariate analyses further showed group differences on several individual defenses (e.g., acting out,
denial, and turning against self). Together, the findings suggest that in BPD, the defenses may emphasize interpersonal dependency and a tendency
to direct aggression toward the self; whereas in APD, the defenses may emphasize egocentricity, interpersonal exploitation, and a tendency to
direct aggression toward others. Overall, this study demonstrates important differences in defense use between borderline and antisocial personality
groups across both observer and self-report measures.

Recently, there has been a renewed interest in the empirical study
of defense mechanisms and the role they play in many aspects
of psychosocial functioning (e.g., Hilsenroth, Hibbard, Nash,
& Handler, 1993; MacGregor, Davidson, Barksdale, Black, &
MacLean, 2003). Defenses are now included in the Diagnos-
tic and Statistical Manual of Mental Disorders (4th ed., text
rev. [DSM–IV–TR]; American Psychiatric Association, 2000)
as a diagnostic axis for further study. Empirical evidence has
emerged demonstrating the role of defenses in childhood and
adolescent development (Cramer, 1997), adult social function-
ing (Vaillant, 1977), physical health (MacGregor et al., 2003),
and psychopathology (Andrews, Singh, & Bond, 1993). De-
fenses may also be helpful in differentiating between similar
mental disorders (Spinhoven & Kooiman, 1997). Cluster B per-
sonality disorders are some of the most difficult disorders to
differentiate, with literature having shown overlapping criteria,
high comorbidity, and low discriminant validity (e.g., Becker,
Grilo, Edell, & McGlashan, 2000). One of the highest comor-
bidity rates within this cluster is between borderline person-
ality disorder (BPD) and antisocial personality disorder (APD;
Becker et al., 2000; Zanarini et al., 2004). The goal of these stud-
ies was to investigate the utility of defenses in differentiating
borderline and antisocial personalities.
CONCEPTUALIZATION AND ASSESSMENT OF DEFENSE
MECHANISMS
Conceptualization
Current psychodynamic authors have defined defenses as un-
conscious mental operations that function to defend against ex-
Received February 1, 2009; Revised October 21, 2009.
Address correspondence to Michelle D. Presniak, Institute of Community and
Family Psychiatry at SMBD Jewish General Hospital, Department of Psychiatry,
McGill University, 4333 Chemin de la cote Ste-Catherine, Montreal, Quebec,
H3T 1E4, Canada. Email: michelle.presniak@usask.ca
137
cessive anxiety (e.g., Vaillant, 1994). Defenses are viewed as
both protectors against unconscious feelings and as cognitive
and relational behaviors that protect the self (Cooper, 1998).
At present, there is no consensus on the number of defenses
that exist. Anna Freud (1936/1986) discussed 10 defenses in
“The Ego and the Mechanisms of Defense,” and there has since
been an expansion of recognized defenses. For example, Vaillant
(1977) included 18 in his original hierarchy of defenses, whereas
the DSM–IV–TR (American Psychiatric Association, 2000) in-
cludes a list of 31 defenses. Despite disagreement in the number
of defenses that exist, most authors agree on the role of defenses.
Defense mechanisms work to alter internal psychological states
(e.g., emotions, thoughts). They alter meaning or significance
of perceived threats, create the perception of control, reduce or
eliminate the experience of conscious anxiety, and thus protect
and enhance the self (Ihilevich & Gleser, 1995).
Assessment of Defense Mechanisms
One area of debate in defense mechanism literature centers
on how best to assess defenses. For example, theorists who
see defenses predominantly as intrapsychic processes have ar-
gued against the use of self-report measures to assess defenses,
whereas others argue that self-report assessment has some utility
(Bond, Gardener, Christian, & Sigal, 1983; Cooper, 1998).
Self-report defense measures require individuals to report on
their thoughts, feelings, and behaviors, which are believed to
represent aspects of various defenses (Cramer, 2006). Advan-
tages of this method are that it limits researcher bias and allows
timely collection of data in large samples. A criticism, however,
is that individuals must consciously report on their unconscious
behavior (Shedler, Mayman, & Manis, 1993). Because self-
report assessment requires an awareness of thoughts, feelings,
and behaviors, these measures are problematic for assessment
of individuals with limited self-awareness or who frequently use
defenses that distort self-evaluation (e.g., denial; Shedler et al.,
 138
1993). As such, most self-report defense measures use items in-
tended to assess conscious derivatives of defenses (Bond, 2004).
Other researchers have focused on objective observational
assessment of defenses (e.g., MacGregor, Olson, Presniak, &
Davidson, 2008; Perry, 1990). This usually involves the use of
coding manuals to assist trained observers in the identification
and assessment of the defenses. These measures allow for use
of some inference by coders, which facilitates the assessment of
unconscious material. One limitation of this approach, however,
is that it is often labor intensive and makes data collection in
large samples difficult. Additionally, although good interrater
reliability can often be achieved for overall defensive function-
ing or defense levels, it can be difficult to achieve good reliability
at the individual defense level.
Assessing defenses with more than one method of measure-
ment can provide additional information regarding an individ-
ual’s defense use. However, problems may arise when trying
to compare results from different measures. One common dif-
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014
ficulty is differences in the conceptualization of a particular
defense mechanism between the measures. For example, de-
valuation, as assessed by some measures (e.g., Defense Style
Questionnaire [DSQ]; Andrews et al., 1993), combines both
the devaluation of self and the devaluation of others, whereas
other measures only include one direction of devaluation, such
as devaluating others (e.g., Defense-Q; MacGregor et al., 2008).
Additionally, the behaviors an individual can or will self-report
in the context of privately filling out a questionnaire may dif-
fer from behaviors evident to an external observer viewing a
video-recorded interview (Bond, 1990). In turn, these behaviors
may differentially relate to various aspects of psychopathology
or personality to which the defense scores are compared. Be-
cause of these conceptualization, measurement, or contextual
differences, it is difficult to generalize results across measures.
As such, it is important to utilize multiple methods of assessing
defenses in any study (Davidson & MacGregor, 1998).
DIFFERENTIATING BPD AND APD BASED ON
DEFENSE MECHANISMS
Since the introduction of personality disorders into the DSM–
III (American Psychiatric Association, 1980), there have been
many criticisms of the diagnoses, especially of those disor-
ders within Cluster B (e.g., Herkov & Blashfield, 1995; Widi-
ger, 1993). One major criticism is the high comorbidity and
poor discriminant validity between the Cluster B personality
disorders due to their overlapping criteria (Blais & Norman,
1997). BPD has one of the highest comorbidity rates with other
personality disorders, especially with APD (e.g., 26%: Becker
et al., 2000; Herkov & Blashfield, 1995).
Theoretical Role of Defenses in BPD and APD
According to psychodynamic theory, BPD and APD share
many similarities in personality structure. Both disorders fall
within Kernberg’s (1984) borderline personality organization,
thus sharing primary use of primitive defenses, a lack of an in-
tegrated identity, and maintained reality testing. The primitive
defenses are thought to protect the ego by separating contra-
dictory experiences of self and others, thus preventing anxiety
related to these conflicts (Kernberg, 1984).
Despite the similarities in psychodynamics (e.g., identity dis-
integration) of BPD and APD, many argue that these groups
PRESNIAK, OLSON, MACGREGOR
differ in the expression of the underlying dynamics. For exam-
ple, Perry and Cooper (1986) suggested that BPD is expressed
through interpersonal dependency, whereas APD is expressed
through interpersonal exploitation and antisocial behaviors (e.g.,
stealing). Also, although both disorders are characterized by
manipulative behavior, in BPD, it is used in interpersonal rela-
tionships to gain concern from others; whereas in APD, it is used
to establish and gain material goods and power (American Psy-
chiatric Association, 2000). Finally, although both disorders are
characterized by aggression, Gacono, Meloy, and Berg (1992)
suggested that the aggression may be more strongly directed
toward the self in BPD and toward others in APD.
Although individuals with BPD and APD both exhibit prim-
itive defense use (e.g., splitting, denial, devaluation), many au-
thors have also stated that these groups will differ in their relative
use of some defenses. For example, BPD is considered more se-
vere than APD within borderline personality organization and
should therefore be associated with higher reliance on maladap-
tive defenses (Cramer, 1999). Perry and Cooper (1986), how-
ever, proposed that two dimensions of primitive defenses exist.
The first dimension includes splitting of self, splitting of others,
and projective identification and would be more associated with
BPD because it leads to interpersonal instability and depen-
dency. The second dimension includes omnipotence, primitive
idealization, and devaluation and would be more associated with
APD because it protects the self from low self-esteem and feel-
ings of anger.
Gacono et al. (1992) suggested some differences in the way
these defenses relate to BPD and APD. First, Gacono et al.
suggested that projection may be more characteristic of APD
because the primary way these individuals can relate to others
is through a projection of themselves. Second, unlike Perry and
Cooper (1986), Gacono et al. argued that many of the broader
conceptualized defenses will be associated with both BPD and
APD such as splitting and devaluation. However, Gacono et al.
suggested that the way in which the defense is utilized differs be-
tween groups. For example, although both groups use splitting,
those with BPD tend to alternate between idealization of others
and devaluation. The devaluation, however, is predominantly
turned inward with negative self-thoughts, feelings, and behav-
iors (i.e., devaluation of self, turning against self). Although at
times aggression in BPD can be directed toward others, this
often results in an exacerbated view of the self as damaged. In
those with APD, splitting is evident through a poorly integrated
identity in which they present with an overt grandiosity of the
self and tendency to devalue others. Splitting, therefore, occurs
between a grandiose and devalued self, but this latter portion is
defended against through grandiosity and devaluation of others.
Gacono and Meloy (1988) also argued that denial is an es-
pecially integral defense for individuals with APD because it
prevents them from thinking about the consequences of their
antisocial behaviors. They additionally note that, in its more
developmentally advanced form, denial may develop into ra-
tionalization. This occurs as the individuals mature and, rather
than deny their behaviors, they will justify them through dis-
torted explanations.
Empirical Evidence of Defense Use in BPD and APD
Research has supported Kernberg’s (1984) hypotheses that
BPD is positively associated with primitive defenses (e.g.,
 DEFENSE MECHANISMS AND PERSONALITY
denial, devaluation, idealization, projective identification, split-
ting; Berg, 1990; Hilsenroth et al., 1993). Studies have also
shown a positive association between APD and primitive de-
fense use (e.g., devaluation, projective identification, splitting;
Chabrol & Leichsenring, 2006; Leichsenring, Kunst, & Hoyer,
2003). Although both disorders are associated with primitive
defenses, some findings have supported that BPD may demon-
strate higher levels of maladaptive defenses, such as acting out
and passive aggression, compared to other personality disorder
groups (Bond, 1990; Perry & Cooper, 1986).
Few studies have examined the association of defenses with
BPD and APD concurrently (e.g., Cramer, 1999; Gacono et al.,
1992). The results of these studies have shown that BPD is
positively associated with acting out, denial, principalization,
reversal, and turning against self (TAS), and APD is positively
associated with acting out, denial, intellectualization, princi-
palization, projection, and turning against object (Berman &
McCann, 1995; Cramer, 1999; Lingiardi et al., 1999; Perry &
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014
Cooper, 1986). In addition, Perry and Cooper (1986) tested their
hypothesis that two dimensions of primitive defenses existed.
They found that the first dimension (i.e., splitting of self, split-
ting of others, and projective identification) was associated with
BPD symptoms (r = .36), whereas the second (i.e., omnipo-
tence, primitive idealization, and devaluation) was associated
with APD symptoms (r = .23). Perry and Cooper (1986) also
found that the action defense level (most maladaptive level) was
associated with BPD symptoms (r = .26), and the disavowal
defense level (e.g., denial) was associated with APD symptoms
(r = .22).
Despite different patterns of defense use emerging in these
correlational studies, only two studies have statistically com-
pared the mean differences in defense use between BPD and
APD groups. Perry and Cooper (1986) compared defenses be-
tween 10 patients with BPD and 8 patients with APD. Defenses
were assessed with an early version of the Defense Mechanism
Rating Scales (DMRS; Perry, 1990). Results from a canonical
discriminant function analysis revealed no significant differ-
ences between groups. In 1992, Gacono et al. examined the
results of previous studies that used the Lerner Defense Scales
(Lerner & Lerner, 1980) to assess defenses. One-way analy-
ses of variance (ANOVAs) revealed no significant differences
between BPD (n = 18) and APD (n = 21) groups, although
there was a trend toward higher primitive idealization scores in
the BPD group. Together, the results of these two studies sug-
gest that these disorders could not be differentiated based on
defenses.
CURRENT RESEARCH STUDIES
Most studies that have investigated defenses and personality
disorders have examined correlations between disorders and de-
fenses. To date, however, only Gacono et al. (1992) and Perry
and Cooper (1986) have statistically explored differences in
defense use between BPD and APD groups. Although no sig-
nificant results were found in either study, both studies had very
small samples sizes. Additionally, the chosen analyses may have
restricted potential findings. For example, Perry and Cooper
(1986) compared the groups on five defense styles rather than
individual defenses, whereas many argue that differences be-
tween the groups will be seen on individual defenses and not on
broad defense styles (e.g., Gacono et al., 1992).
139
Only Gacono et al. (1992) previously examined differences in
individual defense mean scores between BPD and APD groups.
In their study, however, Gacono et al. acknowledged that they
were limited by the defense measure used, the Lerner Defense
Scales, because the measure requires human responses on the
Rorschach. Gacono et al. (1992) noted that “too few scores were
produced to analyze these indices meaningfully” (p. 41).
In these studies, we examined mean differences in individual
defense mechanisms between BPD and APD groups. In Study 1,
we assessed defenses using a self-report measure, the DSQ;
and in Study 2, the DSQ and an observer report measure, the
Defense-Q, were used. In both studies, mean defense scores
were compared between BPD and APD groups using multivari-
ate ANOVA (MANOVA) analyses. To avoid the small sample
limitation of previous studies, we used a self-report personal-
ity measure in two large samples to identify potential partici-
pants that might approximate participants diagnosed with BPD
or APD. The Personality Assessment Inventory (PAI; Morey,
1991) was chosen based on its good overall validity and reliabil-
ity indexes as well as its demonstrated validity at assessing BPD
pathology in nonclinical samples (Boone, 1998; Kurtz, Morey,
& Tomarken, 1993; Trull, 1995). Group membership was as-
signed based on clinical cutoff scores from the PAI. Defense
use comparisons between groups were conducted to extend pre-
vious research, which has primarily examined the correlations
between many defenses and personality disorders.
Hypotheses
Based on the theoretical and empirical literature described
previously, 12 hypothesized defense use differences were tested
between the BPD and APD groups. Five defenses were pro-
posed to be higher in the BPD group (acting out, idealization,
passive aggression, splitting, and TAS), and seven defenses were
proposed to be higher in the APD group (denial, devaluation
of other, grandiosity, intellectualization, projection, rationaliza-
tion, and turning against others [TAO]).
STUDY 1
Method
Participants. A total of 674 undergraduate students from
a Canadian university participated in this study. All students
were enrolled in an introductory psychology class and received
partial course credit for participation. All participants provided
informed consent, and the study was approved by the Research
Ethics Board of the University of Saskatchewan. The mean age
of the sample was 19.4 years (range = 17–40; SD = 3.0). The
sample was 71.0% female, and 87.1% of the participants were
Caucasian.
Measures
PAI (Morey, 1991). The PAI is a 344-item self-report mea-
sure of mental health and personality functioning, which in-
cludes four Validity scales, 11 Clinical scales, five Treatment
scales, and two Interpersonal scales (Morey, 1991). The Bor-
derline Features (BOR) and Antisocial Features (ANT) Clini-
cal scales were used in this study. The PAI has demonstrated
good reliability and validity in census, college, and clinical sam-
ples (e.g., Boyle & Lennon, 1994; Kurtz et al., 1993; Morey,
1991; Walters, Duncan & Geyer, 2003). The ANT and BOR
 140
TABLE 1.—Study 1 means and standard deviations for the Defense Style Questionnaire.
Defense BPD M (SD)a
Acting Out 33.68 (8.07)
Denial 5.82 (2.84)
Idealization 15.68 (4.72)
Passive Aggression 38.91 (9.13)
Projection 33.44 (9.17)
Rationalization 4.20 (2.25)
Splitting 11.65 (4.78)
Note. BPD = borderline personality disorder group; APD = antisocial personality disorder group.
a
n = 34. b n = 38. c Reliability calculated using Cronbach’s alpha. Reliability for rationalization is not calculated because it consists of one item.
scales have demonstrated good concurrent validity. For exam-
ple, the BOR scale correlated with borderline scales from other
self-report measures (e.g., Minnesota Multiphasic Personality
Inventory [Hathaway & McKinley, 1943]; Kurtz et al., 1993),
and the ANT scale correlated with the Psychopathy Checklist–
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014
Revised (Hare, 1991) total scale, Factor 1 (Affective and Inter-
personal Features), and Factor 2 (Behavioral Features; Walters
et al., 2003). Finally, the PAI has demonstrated good validity
in nonclinical samples (Boone, 1998; Kurtz et al., 1993; Trull,
1995). For example, nonclinical participants who scored ≥70t
on the BOR scale demonstrated more interpersonal problems,
maladaptive coping methods, academic difficulties, and psy-
chopathology in general compared to those scoring <70t (Trull,
1995; Trull, Useda, Conforti, & Doan, 1997).
DSQ (Andrews et al., 1993). The DSQ is a 72-item self-
report questionnaire that assesses the use of 20 defense mech-
anisms categorized into three defense styles: mature, neurotic,
and immature (Andrews et al., 1993). Participants rate their
agreement of each item on a scale ranging from 1 (no agreement)
to 9 (strong agreement). Only the hypothesized defenses of act-
ing out, denial, idealization, passive aggression, projection, ra-
tionalization, and splitting were used in analyses. Devaluation
was not included because the DSQ combines both devaluation
toward others and toward the self; therefore, the directional dif-
ferences proposed by Gacono et al. (1992) could not be tested.
Other defenses hypothesized to be related to BOR and ANT
(e.g., TAS) are not assessed by the DSQ. Internal consistency
for individual defenses has ranged from .07 to .82 (Andrews
et al., 1993). Concurrent validity of the DSQ is evidenced by
differences in individual defense use between individuals with
clinical disorders and healthy controls (Andrews et al., 1993).
Procedure
We asked participants to complete the PAI and the DSQ.
Participants who did not complete both measures were not in-
cluded, resulting in a sample of 446 participants. These par-
ticipants were screened using PAI scores. Invalid profiles were
identified according to Morey (1991). We removed any profiles
that evidenced inconsistent (Inconsistency scale; t score > 73),
careless (Infrequency scale; t score > 74), exaggerated nega-
tive (Negative Impression Management scale; t score > 91), or
exaggerated positive responding (Positive Impression Manage-
ment scale; t score > 68), resulting in a sample of 428. Second,
we divided participants into BPD and APD groups based on
the following criteria: 34 participants met criteria for the BPD
group (i.e., scored ≥ 70t on the PAI BOR scale and < 70t on
PRESNIAK, OLSON, MACGREGOR
APD M (SD)b F p d Reliability (No. of Items)c
27.58 (6.12) 13.23 .001 +.87 .69 (6)
7.92 (3.32) 8.18 .006 −.68 .19 (2)
13.95 (4.82) 2.35 .129 +.37 .32 (3)
32.34 (7.59) 11.10 .001 +.80 .54 (8)
29.53 (10.84) 2.70 .105 +.39 .74 (10)
5.34 (2.35) 4.38 .040 −.50 — (1)
11.42 (4.98) 0.04 .845 +.05 .56 (3)
the ANT scale), and 38 participants met criteria for the APD
group (i.e., score ≥ 70t on the ANT scale and < 70t on the
BOR scale). We excluded participants who had scores ≥ 70t on
both the ANT and BOR scales.
The mean t score for BOR in the BPD group was 74.1 (SD =
4.0), and in the APD group it was 58.4 (SD = 7.3). The mean t
score for ANT in the APD group was 76.1 (SD = 5.3); and in the
BPD group, it was 56.5 (SD = 7.4). The mean age of the grouped
sample was 20.0 years (range = 18–33; SD = 3.0), and 93.2%
were Caucasian, with no significant age or racial differences
between groups, t(58) = 1.62, p = .11 and χ 2 (2, N = 72) =
2.31, p = .32, respectively. There were significant differences
in percentage of each sex between groups, χ 2 (1, N = 72) =
17.51, p < .001, with more women in the BPD group (85.3%)
compared to the APD group (36.8%). Many of the participants
in each group had elevations on other scales on the PAI. The
most common concurrent elevations in the BPD group were on
the Suicide (38.2%) and Depression scales (47.1%); and in the
APD group, the most common elevations were on the Alcohol
Problems (31.6%) and Aggression (15.8%) scales.
Results
MANOVA results. Due to the significant differences in sex
between the two groups, we conducted a preliminary factorial
ANOVA whereby the interaction between sex and defenses was
tested. There was no significant interaction, F (7, 62) = 0.22,
p = .98; therefore, the MANOVA was conducted with men and
women combined.
We conducted a MANOVA between the APD and BPD groups
on the following defenses: Acting Out, Denial, Idealization, Pas-
sive Aggression, Projection, Rationalization, and Splitting. The
mean and standard deviation for each defense mechanism is pre-
sented in Table 1.1 Wilks’ Lambda indicates there was an overall
significant difference between the BPD and APD groups on de-
fenses, F (7, 64) = 5.01, p < .001. This effect accounted for
35.4% (η2 ) of the variance. The univariate results revealed that
Acting Out and Passive Aggression were significantly higher
in the BPD group, and Denial and Rationalization were signif-
icantly higher in the APD group. Acting Out accounted for the
most variance (partial ω2 = 15.9%) followed by Passive Ag-
gression (13.7%), Denial (10.5%), and Rationalization (5.9%).
1 After examining the variables for normal distribution, one variable (Pro-
jection) was significantly skewed (z > 3.33). We conducted a square root trans-
formation, which resulted in a normal distribution. Because the results with
the transformed score confirmed the results with the nontransformed score, the
latter results were reported.
 DEFENSE MECHANISMS AND PERSONALITY 141

TABLE 2.—Study 1 correlations between seven Defense Style Questionnaire TABLE 3.—Study 2 means and standard deviations for the Defense-Q.
defenses and the Borderline Features and Antisocial Features scales from the
Personality Assessment Inventory. BPD M APD M
Defense (SD)a (SD)b F p d Reliabilityc
Defense Borderline Features Antisocial Features z p
Acting Out 2.84 (0.83) 3.27 (1.05) 3.92 .051 −.45 .73
Acting Out .53∗∗ .35∗∗ 4.05 <.001 Devaluation 3.86 (0.80) 4.32 (0.91) 5.74 .019 −.54 .54
Denial −.00 .22∗∗ 4.29 <.001 Grandiosity 3.94 (0.76) 4.33 (0.81) 4.90 .030 −.50 .64
Idealization .01 −.10∗ 2.11 .017 Idealization 3.82 (0.68) 3.84 (0.48) 0.03 .857 −.04 .52
Passive Aggression .47∗∗ .30∗∗ 3.54 <.001 Intellectualization 3.07 (1.23) 3.55 (1.54) 2.28 .135 −.34 .85
Projection .50∗∗ .34∗∗ 4.36 <.001 Neurotic Denial 4.18 (1.14) 4.63 (1.22) 2.79 .099 −.38 .72
Rationalization .08 .23∗∗ 2.96 .002 Passive Aggression 4.18 (0.86) 4.18 (0.77) 0.00 .995 .00 .60
Splitting .31∗∗ .25∗∗ 1.28 .100 Projection 4.31 (0.57) 4.01 (0.55) 6.16 .015 +.56 .45
Rationalization 5.41 (0.86) 5.26 (0.84) 0.66 .420 +.18 .52
Note. N = 428. z scores indicate differences in magnitude of correlations according to Splitting 3.39 (0.84) 3.19 (0.69) 1.37 .246 +.26 .28
Meng, Rosenthal, and Rubin (1992). Turning Against Others 4.66 (0.77) 4.36 (0.79) 2.89 .093 +.38 .51
∗
p < .05. ∗∗ p < .001. Turning Against Self 4.62 (0.97) 4.12 (0.99) 5.31 .024 +.52 .62

Note. BPD = borderline personality disorder group; APD = antisocial personality

Correlation results. Because a larger sample of partici- disorder group.
a
n = 34. b n = 49. c Defense-Q reliability calculated using one-way random effects
pants had completed both the PAI and DSQ (of whom many did model intraclass correlation coefficients.
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014

not meet inclusion criteria), we also conducted correlation anal-

yses between the individual defenses with the BOR and ANT
scales of the PAI in this larger sample (n = 428). The results
of the Pearson correlations are reported in Table 2.2 Results in-
dicated four defenses were positively correlated with the BOR
scale: Acting Out, Passive Aggression, Projection, and Splitting.
Six defenses were positively correlated with the ANT scale:
Acting Out, Denial, Passive Aggression, Projection, Rational-
ization, and Splitting. One defense, Idealization, was negatively
correlated with the ANT scale.
STUDY 2
Method
Participants. A total of 1,539 undergraduate students from
a Canadian university participated in this study. All participants
provided informed consent, and the study was approved by the
Research Ethics Board of the University of Saskatchewan. The
mean age of the participants was 19.7 years (range = 17–40;
SD = 2.7); 68.0% were female, and 87.0% were Caucasian. All
participants received partial course credit for their participation.
Measures
The PAI and DSQ as described in Study 1 were used in
Study 2.
Defense-Q (MacGregor et al., 2008). The Defense-Q is
an observational measure used to assess the relative use of 25
defense mechanisms (MacGregor & Davidson, 1998; MacGre-
gor et al., 2008). It is a Q-sort instrument based on a system
of rank ordering: one defense is identified as most characteris-
tic, two as quite characteristic, five as somewhat characteristic,
nine as neither uncharacteristic nor characteristic, five as some-
what uncharacteristic, two as quite uncharacteristic, and one as
most uncharacteristic. The categorization results in a seminor-
mal distribution representing a participant’s overall pattern (or
profile) of defense use. The Defense-Q allows for the exami-
nation of individual defense use as well as provides a profile
2 Three variables (denial, projection, and splitting) were significantly skewed
(z > 3.33) and we conducted transformations, resulting in a normal distribution.
Because the results with the transformed scores confirmed the results with the
nontransformed scores, the latter results were reported.
of overall defense use. The individual defense scores from the
Defense-Q coders were averaged together to produce a mean
score for each of the 12 defenses used in this study. We used
the defenses of Acting Out, Devaluation, Grandiosity, Idealiza-
tion, Intellectualization, Neurotic Denial, Passive Aggression,
Projection, Rationalization, Splitting, TAO, and TAS in this
study.
The Defense-Q has demonstrated good convergent and con-
current validity (see MacGregor & Olson, 2005). Interrater reli-
ability for the Defense-Q has been shown to range from poor to
excellent for individual defenses and excellent for profile com-
parisons (MacGregor & Olson, 2005). The interrater reliability
for each defense from this study are reported in Table 3.
Procedure
Participants completed the PAI in the initial meeting. Of the
1,539 participants, we called back 171 for this study. We con-
tacted participants if they had elevated scores on BOR or ANT.
We asked participants to return to complete the DSQ and the Ex-
panded Structured Interview (ESI; Hall, Davidson, MacGregor,
& MacLean, 1998). The ESI is a 12 to 15 min semistructured in-
terview designed to elicit information related to stress, anxiety,
and coping (Hall et al., 1998). Participants are asked questions
related to school and work and how they react in certain stress-
ful situations. The ESI is based on a Type A interview (used to
assess hostility, competitiveness, time urgency, etc.) adapted for
use with a university population and modified to assess defense
use and emotional expression. Defenses have been assessed us-
ing the ESI in previous research on defenses, personality, and
health (e.g., MacGregor et al., 2003). The entire interview is
viewed separately by each coder, providing them with a sam-
ple of responses and behaviors from which they observationally
assess defenses using the Defense-Q.
Of the 171 participants who were called back, 92 participants
returned to complete the ESI and DSQ. The most common rea-
sons for not returning were a failure to respond to repeated
contact attempts or a response that they had already reached
their maximum allotted research credits. Once we removed par-
ticipants for incomplete data or invalid PAI profiles (see Study 1
for criteria), the resulting sample was 83 participants, of which
34 met criteria for the BPD group (i.e., scored ≥ 70t on the PAI
 142
TABLE 4.—Study 2 means and standard deviations for the Defense Style Questionnaire.
Defense BPD M (SD)a
Acting Out 33.77 (7.27)
Denial 5.97 (3.08)
Idealization 15.62 (4.57)
Passive Aggression 38.77 (7.92)
Projection 35.06 (12.24)
Rationalization 5.06 (2.51)
Splitting 12.29 (4.47)
Note. BPD = borderline personality disorder group; APD = antisocial personality disorder group.
a
n = 34. b n = 49. c Reliability calculated using Cronbach’s alpha. Reliability for rationalization is not calculated because it consists of one item.
BOR scale and < 70t on the ANT scale), and 49 met criteria for
the APD group (i.e., score ≥ 70t on the ANT scale and < 70t
on the BOR scale). All variables were normally distributed with
no outliers. Of the total sample, 93.4% were Caucasian, and
the mean age was 19.5 years (range = 18–26; SD = 1.6), with
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014
no significant age, t(59) = 0.04, p = .97, or racial differences,
χ 2 (3, N = 61) = 3.88, p = .28 between the groups. There
were significant differences in the sex of participants between
groups, χ 2 (1, N = 61) = 8.15, p = .006, with more women in
the BPD group (84.6%) compared to the APD group (48.4%).
Many of the participants in each group had elevations on other
scales on the PAI. The most common concurrent elevations in
the BPD group were on the Suicide (26.5%) and Depression
scales (38.2%); and in the APD group, the most common eleva-
tions were on the Alcohol Problems (34.7%) and Drug Problems
(26.5%) scales.
Defense coder training. Data were collected over 4 con-
secutive years. Each year’s participant interviews were coded by
three Defense-Q coders, some of whom coded multiple years.
The coders from the initial 2 years had additionally coded de-
fense mechanism data for another published manuscript (Olson,
Presniak, & MacGregor, 2009). All coders were senior under-
graduate students who received standard training for the Defense
Q. This included a minimum of 20 hr of training, which involved
a discussion of theory, how to rate each defense mechanism (i.e.,
definition, examples of its manifestation, how to look for evi-
dence of its presence or absence), and observation of trainers
(current authors) coding defenses from video-recorded inter-
views as well as coders independently practicing coding and
then reviewing their assessment in a group meeting (see David-
son & MacGregor, 1996, or MacGregor & Olson, 2005, for a
more detailed discussion of training procedure). After coding
defenses for this data began, coders met together with trainers
weekly to discuss coding and maintain reliability.
Results
DSQ results. Due to the significant differences in sex be-
tween the two groups, we conducted a preliminary factorial
ANOVA whereby we tested the interaction between sex and
defenses. There was no significant interaction for the DSQ anal-
yses, F (7, 47) = 0.58, p = .77; therefore, we conducted the
MANOVA with men and women combined.
The mean and standard deviation for each defense in the anal-
ysis is presented in Table 4. Results of the Wilks’ Lambda multi-
variate test indicated a significant overall difference between the
BPD and APD groups on defenses, F (7, 75) = 2.68, p = .016.
PRESNIAK, OLSON, MACGREGOR
APD M (SD)b F p d Reliability (No. of Items)c
29.27 (7.09) 7.92 .006 +.64 .55 (6)
6.84 (3.09) 1.58 .212 −.28 .07 (2)
15.45 (4.26) 0.30 .864 +.04 .14 (3)
33.88 (6.44) 9.57 .003 +.70 .22 (8)
30.76 (12.71) 2.37 .127 +.35 .83 (10)
5.45 (2.07) 0.60 .442 −.18 − (1)
11.86 (5.05) 0.17 .686 +.09 .44 (3)
This effect accounted for 18.8% (η2 ) of the variance. Univariate
results revealed that Acting Out and Passive Aggression were
higher in the BPD group compared to the APD group. Passive
Aggression accounted for 10.6% (partial ω2 ) of the variance,
and Acting Out accounted for 8.9%.
Defense-Q results. Due to the significant differences in sex
between the two groups, we conducted a preliminary factorial
ANOVA whereby we tested the interaction between sex and
defenses. There was no significant interaction for the Defense-
Q analyses, F (12, 42) = 0.58, p = .85; therefore, we conducted
the MANOVA with men and women combined.
The mean and standard deviation for each defense mechanism
used in the Defense-Q analysis is presented in Table 3. There
was an overall significant difference between the BPD and APD
groups on defenses, F (12, 70) = 3.08, p = .002. This effect
accounted for 34.5% (η2 ) of the variance. Results from univari-
ate analyses revealed that Projection and TAS were higher in
the BPD group, and Devaluation and Grandiosity were signifi-
cantly higher in the APD group. Devaluation accounted for the
most variance (partial ω2 = 6.6%), followed by TAS (6.2%),
Grandiosity (5.7%), and Projection (1.5%).
DISCUSSION
Overall group differences were demonstrated across the two
studies. Of the five defenses hypothesized to be higher in the
BPD group, two were supported (passive aggression and TAS),
one was inconsistently supported (acting out), and two were not
supported (idealization and splitting). Of the seven defenses hy-
pothesized to be higher in the APD group, two were supported
(devaluation of others and grandiosity), one was partially sup-
ported (denial; all effects in the right direction, but only one
of three was significant), one was inconsistently supported (ra-
tionalization), and three were not supported (intellectualization,
TAO, and projection).
Consistent with our hypothesis and previous theory and re-
search (e.g., Bond, 1990; Cramer, 1999), we found partial sup-
port that the BPD group would use the maladaptive defenses,
acting out, and passive aggression more than the APD group.
These results, however, were not consistent across measure-
ment type.3 The Defense-Q results indicated no difference in
Passive Aggression scores and showed a moderate effect size
for lower Acting Out scores in the BPD group. One explanation
for this discrepancy could be the differences in the measurement
3 For a correlation table of the convergence between the DSQ and Defense-Q
defenses from this study, please contact M. D. Presniak.
 DEFENSE MECHANISMS AND PERSONALITY
of these defenses between the DSQ and Defense-Q. For exam-
ple, the DSQ’s Acting Out predominantly assesses impulsivity
and limited aggression, which may be more likely to be self-
reported by individuals with BPD. In comparison, APD groups
may be less likely to admit exploitative and criminal acts on
self-report measures, but their acting out behaviors may be ob-
served by coders. As such, it would be valuable to further assess
the types of acting out and passive aggressive behaviors used
by each group and to examine whether these differences are
reflected in measurement type or differentially related to each
group.
Contrary to our hypothesis, splitting was not used signif-
icantly more by the BPD group. Our results, however, were
consistent with researchers who previously have found no dif-
ference in splitting use between BPD and APD groups (Gacono
et al., 1992). Another explanation for the unsupported hypoth-
esis could be related to the low reliability of Splitting scores.
In Study 2, interrater reliability of Splitting on the Defense-Q
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014
was poor (.28), therefore likely affecting the validity of these
results. The hypotheses, however, were not supported with the
DSQ either. The construct validity of assessing splitting through
self-report may have impacted the findings. For example, the
DSQ asks participants to report the extent to which they view
things as polar opposites (e.g., “Sometimes I think I’m an angel
and other times I think I’m a devil”). Accurate self-report of this
behavior may be challenging for individuals who rely heavily
on splitting because it requires a self-awareness of an inability
to integrate that is inherently problematic for them.
Gacono et al. (1992) also proposed an alternate theory of
splitting in BPD and APD, and our results provide support for
this conceptualization. First, Grandiosity scores were higher in
the APD group, supporting the notion that this group splits their
identity into a grandiose and devalued self, of which only the
grandiose self is evident. Second, our findings also partially
supported Gacono et al.’s (1992) suggestion that the direction
aggression is aimed will differ between groups. TAS was sig-
nificantly higher in the BPD group compared to the APD group.
This finding was consistent with previous findings showing that
BPD is positively associated with TAS (Berman & McCann,
1995). In addition, devaluation of others was used significantly
more by the APD group. Lingiardi et al. (1999) had previously
found no correlation between APD scores and devaluation on
the DMRS, although this research was conducted using a pre-
vious version of the DMRS that combined the constructs of
devaluation of self and others, thus obscuring the directional
relationships.
Although a large portion of Gacono et al.’s (1992) theory
on splitting was supported (i.e., use of devaluation, grandiosity,
and TAS), our results did not show support for significant dif-
ferences between the groups for use of idealization or TAO. The
results for TAO, which showed no significant differences but a
small to medium effect size indicating a lower score in the APD
group, were contrary to previous theory and research. Using
the self-report Defense Mechanisms Inventory (DMI; Ihilevich
& Gleser, 1986), Berman and McCann (1995) previously found
APD was positively associated with the individual defense Turn-
ing Against Object. The DMI, however, uses a forced-choice
self-report format in which participants are required to choose
between five reactions to multiple situations. Each reaction rep-
resents a defense, and as such, participants are unable to have
elevated scores on all five defenses. It may be that blaming others
143
is characteristic of both BPD and APD, but that when required
to choose one’s most likely reaction to situations, participants
with APD are more likely to choose TAO. Our results also did
not support our hypothesis that idealization would be used more
by the BPD group. However, a moderate effect size emerged
indicating idealization scores on the DSQ were higher in the
BPD group and, in addition, there was a significant negative
correlation between Idealization on the DSQ and the ANT scale
of the PAI. This latter finding supports Gacono et al.’s (1992)
suggestions that individuals with APD are incapable of idealiz-
ing others. Therefore, there is partial evidence that idealization
may not be characteristic of APD and may be used more by
the BPD group. The findings are consistent with the results of
Gacono et al. (1992), but additional research is needed to further
understand whether differences in the use of idealization exist
between BPD and APD groups.
We hypothesized that denial would be used more by the APD
group based on Gacono and Meloy (1988) who suggested that
denial is an important defense for APD. Gacono and Meloy
explained that intrinsic to APD is a “shut-off” mechanism that
allows the exclusion of personality aspects that could diminish
their grandiose self-view or lead to a recognition of their ex-
ploitative and criminal behaviors. Denial is a key component of
this proposed mechanism. In our studies here, Denial and Neu-
rotic Denial scores were higher, with small to medium effect
sizes, in the APD group compared to the BPD group, although
the difference was statistically significant in only one of the three
tests. These findings provide partial support for the hypothesis
and are consistent with previous findings showing APD to be
associated with disavowal level defenses including denial (Perry
& Cooper, 1986).
We also hypothesized that the APD group would use more
rationalization. Gacono and Meloy (1988) suggested that, in its
developmentally advanced form, denial would develop into ra-
tionalization. This occurs when individuals become more aware
of their antisocial behaviors but justify them through distorted
self-serving explanations. The results supported this hypothesis
in Study 1 but not in Study 2. The Defense-Q results showed that
both groups were rated by observers as using the defense above
average (M > 4). It has been noted by many that college stu-
dents tend to rationalize maladaptive behaviors (e.g., cheating
on tests, violence; McCabe, 1992; Pezza & Bellotti, 1995). It
may be that elevated use of rationalization in this university sam-
ple obscured our ability to examine this hypothesis. Lingiardi
et al. (1999) also suggested that intellectualization can capture
the tendency to minimize negative experiences. Lingiardi et al.
previously found that use of intellectualization positively cor-
related with APD in a sample of clinical outpatients. Based on
this finding, we expected intellectualization on the Defense-Q
to be higher in the APD group; however, the hypothesis was
not supported. Further research is needed to explore the use of
intellectualization and rationalization in APD.
Finally, the hypothesis that projection would be used more by
the APD group was not supported. This result is also inconsistent
with the theory proposed by Gacono et al. (1992) that, in APD,
the world is seen through a projected view of themselves as well
as inconsistent with the results of Cramer (1999) who found that
projection was a predictor of APD but not BPD. Our results are
consistent, however, with the theory and findings of Perry and
Cooper (1986), who suggested that projective identification,
which is a more primitive form of projection, was associated
 144
with BPD. Therefore, additional research is needed to examine
the role of projection in both of these groups.
Overall, although all hypotheses were not supported, the find-
ings support the conceptualization of BPD and APD as having
separate and diverse defensive profiles. Supporting Perry and
Cooper’s (1986) proposition, the defenses in BPD appear to
emphasize interpersonal relatedness or dependency. These in-
clude defenses that lead to difficulties expressing one’s needs
and wishes to others (passive aggression), projecting internal
selves onto others (projection), and the splitting of the ego,
which involves predominantly a devaluation of the self (TAS).
Supporting many of Gacono et al.’s (1992) and Gacano and
Meloy’s (1988) theoretical work, the defensive profile in APD,
on the other hand, appears to emphasize egocentricity, inter-
personal exploitation, and the expression of aggression toward
others. Specifically, the defenses contribute to a splitting of the
ego, which results in a grandiose view of the self (grandiosity);
the tendency to minimize, distort, or avoid the consequences of
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014
their behavior (denial and rationalization); and a tendency to
protect the self by devaluing others (devaluation).
Strengths and Limitations
One limitation to these studies is the weaknesses in reliabil-
ity and validity that is characteristic of some defense measures,
which may have decreased the likelihood of finding significant
results. For example, the interrater reliability for Splitting on
the Defense-Q was poor for Study 2, and the Acting Out items
and behaviors measured by the DSQ and Defense-Q may cap-
ture different aspects of acting out. Additionally, some of the
group comparisons made using the DSQ may have been atten-
uated by the low internal consistency observed for several of
these scales. These difficulties capturing defenses can compro-
mise our ability to accurately assess defenses and their relation
to personality disorders, which complicates the comparison of
results across studies. To mitigate this problem, we used both
observer and self-report measures of defenses in this study to
get a more complete and detailed defense assessment. However,
further measurement development is still needed to improve the
assessment of defenses.
Another limitation to these studies is the use of a nonclinical
sample for the BPD and APD groups. Nonclinical samples may
not display the degree of symptom severity and behaviors that
would be found in a clinical group. Although it will be impor-
tant to replicate these results in a clinical sample, our findings
suggest that in BPD, the defenses may emphasize interpersonal
relatedness or dependency, whereas in APD, the defenses may
emphasize egocentricity and interpersonal exploitation.
ACKNOWLEDGMENT
This study was conducted with the support of a Canada Grad-
uate Scholarship awarded to M. D. Presniak by the Social Sci-
ences and Humanities Research Council of Canada.
REFERENCES
American Psychiatric Association. (1980). Diagnostic and statistical manual of
mental disorders (3rd ed.). Washington, DC: Author.
American Psychiatric Association. (2000). Diagnostic and statistical manual of
mental disorders (4th ed., text rev.). Washington, DC: Author.
Andrews, G., Singh, M., & Bond, M. (1993). The Defense Style Questionnaire.
The Journal of Nervous and Mental Disease, 181, 246–256.
PRESNIAK, OLSON, MACGREGOR
Becker, D. F., Grilo, C. M., Edell, W. S., & McGlashan, T. H. (2000). Comor-
bidity of borderline personality disorder with other personality disorders in
hospitalized adolescents and adults. American Journal of Psychiatry, 157,
2011–2016.
Berg, J. L. (1990). Differentiating ego functions of borderline and narcissistic
personalities. Journal of Personality Assessment, 55, 537–548.
Berman, S. M., & McCann, J. T. (1995). Defense mechanisms and personal-
ity disorders: An empirical test of Millon’s theory. Journal of Personality
Assessment, 64, 132–144.
Blais, M. A., & Norman, D. (1997). A psychometric evaluation of the DSM–
IV personality disorder criteria sets. Journal of Personality Disorders, 11,
168–176.
Bond, M. (1990). Are “borderline defenses” specific for borderline personality
disorders? Journal of Personality Disorders, 4, 251–256.
Bond, M. (2004). Empirical studies of defense style: Relationships with psy-
chopathology and change. Harvard Review Psychiatry, 12, 263–278.
Bond, M., Gardener, S. T., Christian, J., & Sigal, J. J. (1983). Empirical study
of self-rated defense styles. Archives of General Psychiatry, 40, 333–338.
Boone, D. (1998). Internal consistency reliability of the Personality Assessment
Inventory with psychiatric patients. Journal of Clinical Psychology, 54, 839–
843.
Boyle, G. J., & Lennon, T. J. (1994). Examination of the reliability and validity
of the Personality Assessment Inventory. Journal of Psychopathology and
Behavioral Assessment, 16, 173–187.
Chabrol, H., & Leichsenring, F. (2006). Borderline personality organization
and psychopathic traits in nonclinical adolescents: Relationships of identity
diffusion, primitive defense mechanisms and reality testing with callousness
and impulsivity traits. Bulletin of the Menninger Clinic, 70, 161–170.
Cooper, S. H. (1998). Changing notions of defense within psychoanalytic theory.
Journal of Personality, 66, 947–964.
Cramer, P. (1997). Evidence for change in children’s use of defense mechanisms.
Journal of Personality, 65, 233–247.
Cramer, P. (1999). Personality, personality disorders, and defense mechanisms.
Journal of Personality, 67, 535–554.
Cramer, P. (2006). Protecting the self: Defense mechanisms in action. New
York: Guildford.
Davidson, K., & MacGregor, M. Wm. (1996). The reliability of the Defense-Q.
Journal of Personality Assessment, 66, 624–639.
Davidson, K., & MacGregor, M. Wm. (1998). A critical appraisal of self-report
defense mechanism measures. Journal of Personality, 66, 965–992.
Freud, A. (1986). The ego and the mechanisms of defense. In J. D. Sutherland
(Ed. & Trans.), The writings of Anna Freud (Vol. 2). London, England:
Hogarth Press. (Original work published 1936)
Gacono, C. B., & Meloy, J. R. (1988). The relationship between cognitive style
and defensive process in the psychopath. Criminal Justice and Behavior, 15,
472–483.
Gacono, C. B., Meloy, J. R., & Berg, J. L. (1992). Object relations, defen-
sive operations, and affective states in narcissistic, borderline, and antisocial
personality disorder. Journal of Personality Assessment, 59, 32–49.
Hall, P., Davidson, K., MacGregor, M. Wm., & MacLean, D. (1998). Expanded
Structured Interview administration training manual (Tech. Rep. No. 1).
Halifax, Nova Scotia, Canada: Heart Health of Nova Scotia.
Hare, R. D. (1991). The Hare Psychopathy Checklist–Revised. North
Tonawanda, NY: Multi-Health Systems.
Hathaway, S. R., & McKinley, J. C. (1943). The Minnesota Multiphasic Per-
sonality Inventory. Minneapolis, MN: University of Minnesota Press.
Herkov, M. J., & Blashfield, R. K. (1995). Clinician diagnoses of personal-
ity disorders: Evidence of a hierarchical structure. Journal of Personality
Assessment, 65, 313–321.
Hilsenroth, M. J., Hibbard, S. R., Nash, M. R., & Handler, L. (1993). A
Rorschach study of narcissism, defense, and aggression in borderline, narcis-
sistic, and cluster c personality disorders. Journal of Personality Assessment,
60, 346–361.
Ihilevich, D., & Gleser, G. C. (1986). Defense mechanisms: Their classifica-
tion, correlates, and measurement with the Defense Mechanism Inventory.
Owosso, MI: DMI associates.
 DEFENSE MECHANISMS AND PERSONALITY
Ihilevich, D., & Gleser, G. C. (1995). The defense mechanisms inventory: Its
development and clinical applications. In H. R. Conte & R. Plutchik (Eds.),
Ego defenses: Theory and development (pp. 221–246). New York, NY: Wiley.
Kernberg, O. F. (1984). Severe personality disorders: Psychotherapeutic strate-
gies. New Haven, CT: Yale University Press.
Kurtz, J. E., Morey, L. C., & Tomarken, A. J. (1993). The concurrent validity of
three self-report measures of borderline personality. Journal of Psychopathol-
ogy and Behavioral Assessment, 15, 255–266.
Leichsenring, F., Kunst, H., & Hoyer, J. (2003). Borderline personality organi-
zation in violent offenders: Correlations of identity diffusion and primitive
defense mechanisms with antisocial features, neuroticism, and interpersonal
problems. Bulletin of the Menninger Clinic, 67, 314–327.
Lerner, P., & Lerner, H. (1980). Rorschach assessment of primitive defenses in
borderline personality structure. In J. Kwawer, A. Sugarman, P. Lerner, & H.
Lerner (Eds.), Borderline phenomena and the Rorschach test (pp. 257–274).
New York, NY: International Universities Press.
Lingiardi, V., Lonati, C., Delucchi, F., Fossati, A., Vanzulli, L., & Maffei, C.
(1999). Defense mechanisms and personality disorders. Journal of Nervous
and Mental Disease, 187, 224–228.
Downloaded by [University Of Pittsburgh] at 11:23 26 November 2014
MacGregor, M. Wm., & Davidson, K. (1998). The Defense-Q: An idiographic Q-
Sort measure of defense mechanisms. A manual for clinicians and researchers
(Tech. Rep. No. 4). Halifax, Nova Scotia, Canada: Heart Health of Nova
Scotia.
MacGregor, M. Wm., Davidson, K., Barksdale, C., Black, S., & MacLean, D.
(2003). Adaptive defense use and resting blood pressure in a population-based
sample. Journal of Psychosomatic Research, 55, 531–541.
MacGregor, M. Wm., & Olson, T. R. (2005). Defense mechanisms: Their rela-
tion to personality and health: An exploration of defense mechanisms assessed
by the Defense-Q. In A. Columbus (Ed.), Advances in psychology research
(Vol. 36, pp. 95–141). New York, NY: Nova Science.
MacGregor, M. Wm., Olson, T. R., Presniak, M. D., & Davidson, K. (2008).
The Defense-Q: An idiographic Q-Sort measure of defense mechanisms: A
manual for clinicians and researchers (2nd ed.). Unpublished manuscript.
McCabe, D. L. (1992). The influence of situational ethics on cheating among
college students. Sociological Inquiry, 62, 365–374.
Meng, X., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated corre-
lation coefficients. Psychological Bulletin, 111, 172–175.
145
Morey, L. C. (1991). The Personality Assessment Inventory. Lutz, FL: Psycho-
logical Assessment Resources.
Olson, T. R., Presniak, M. D., & MacGregor, M. Wm. (2009). Differentiation of
depression and anxiety groups using defense mechanisms. Journal of Nervous
and Mental Disease, 197, 834–840.
Perry, J. C. (1990). The Defense Mechanism Rating scales (5th ed.). Cambridge,
MA: Cambridge Hospital.
Perry, J. C., & Cooper, S. H. (1986). A preliminary report on defenses and con-
flicts associated with borderline personality disorder. Journal of the American
Psychoanalytic Association, 34, 863–893.
Pezza, P. E., & Bellotti, A. (1995). College campus violence: Origins, impacts,
and responses. Educational Psychology Review, 7, 105–123.
Shedler, J., Mayman, M., & Manis, M. (1993). The illusion of mental health.
American Psychologist, 48, 1117–1131.
Spinhoven, P., & Kooiman, C. G. (1997). Defense style in depressed and anxious
psychiatric outpatients: An explorative study. Journal of Nervous and Mental
Disease, 185, 87–94.
Trull, T. J. (1995). Borderline personality disorder features in nonclinical
young adults: 1. Identification and validation. Psychological Assessment, 7,
33–41.
Trull, T., Useda, J. D., Conforti, K., & Doan, B. (1997). Borderline personality
disorder features in nonclinical young adults: 2. Two-year outcome. Journal
of Abnormal Psychology, 106, 307–314.
Vaillant, G. E. (1977). Adaptation to life. Cambridge, MA: Harvard University
Press.
Vaillant, G. E. (1994). Ego mechanisms of defense and personality psy-
chopathology. Journal of Abnormal Psychology, 103, 44–50.
Walters, G. D., Duncan, S. A., & Geyer, M. D. (2003). Predicting disciplinary
adjustment in inmates undergoing forensic evaluation: A direct comparison
of the PCL-R and the PAI. Journal of Forensic Psychiatry and Psychology,
14, 382–393.
Widiger, T. A. (1993). The DSM–III–R categorical personality disorder diag-
noses: A critique and an alternative. Psychological Inquiry, 4, 75–90.
Zanarini, M. C., Frankenburg, F. R., Dubo, E. D., Sickel, A. E., Trikha, A., Levin,
A., & Reynolds, V. (2004). Axis II comorbidity of borderline personality
disorder: Description of 6-year course and prediction to time-to-remission.
Acta Psychiatrica Scandinavica, 110, 416–420.