Content uploaded by Joshua D Miller
Author content
All content in this area was uploaded by Joshua D Miller on Mar 02, 2016
Content may be subject to copyright.
http://asm.sagepub.com
Assessment
DOI: 10.1177/1073191108316888
2008; 15; 450 originally published online May 14, 2008; Assessment
Joshua D. Miller, Eric T. Gaughan and Lauren R. Pryor
Associated With the LSRP Factors
The Levenson Self-Report Psychopathy Scale: An Examination of the Personality Traits and Disorders
http://asm.sagepub.com/cgi/content/abstract/15/4/450
The online version of this article can be found at:
Published by:
http://www.sagepublications.com
can be found at:Assessment Additional services and information for
http://asm.sagepub.com/cgi/alerts Email Alerts:
http://asm.sagepub.com/subscriptions Subscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://asm.sagepub.com/cgi/content/refs/15/4/450 Citations
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
450
Assessment
Volume 15 Number 4
December 2008 450-463
© 2008 Sage Publications
10.1177/1073191108316888
http://asmnt.sagepub.com
hosted at
http://online.sagepub.com
The Levenson Self-Report
Psychopathy Scale
An Examination of the Personality Traits
and Disorders Associated With the LSRP Factors
Joshua D. Miller
Eric T. Gaughan
Lauren R. Pryor
University of Georgia
There are several self-report measures of psychopathy, most of which use a two-factor structure. There is debate
regarding the convergence of these factors, particularly with regard to Factor 1 (F1), which is related to the interper-
sonal and affective aspects of psychopathy; Factor 2 (F2) is related to the social deviance associated with psychopa-
thy. This study examines the relations between the Levenson Self-Report Psychopathy (LSRP) factors and personality
traits and disorders (PDs) in an undergraduate sample (n = 271). LSRP Factor 1 is related to an antagonistic interper-
sonal style (i.e., low Agreeableness; high Narcissistic PD and ratings of prototypical psychopathy), whereas Factor 2
is more strongly related to negative emotionality (i.e., Neuroticism), disinhibition (i.e., low Conscientiousness) and a
broad array of PD symptoms. The authors interpret these findings in the context of alternative measures of psychopa-
thy and suggest that the LSRP is a reasonable, albeit imperfect, measure of psychopathy.
Keywords: psychopathy; Levenson Self-Report Psychopathy Scale; assessment; self-report
D
espite being omitted from recent versions of the
Diagnostic and Statistical Manual of Mental
Disorders (4th ed.) (DSM-IV; American Psychiatric
Association, 1994), psychopathy is one of the most
empirically validated personality disorders (PDs).
The current conceptualization, influenced by Cleckley
(1941) and Hare (1991), includes personality traits
such as manipulativeness, egocentricity, a lack of
remorse or empathy, impulsivity, as well as a pervasive
involvement in criminal behavior (cf. Cooke & Michie,
2001). Psychopathy is related to criminality and
aggression in adults (e.g., Porter, Birt, & Boer, 2001)
and juveniles (e.g., Gretton, Hare, & Catchpole, 2004).
The Psychopathy Checklist (PCL) and its revision
(PCL-R) are the most commonly used measures of
psychopathy among incarcerated offenders. Until
recently (see Cooke & Michie, 2001), PCL/PCL-R
psychopathy was thought to be composed of two fac-
tors that were correlated at approximately .50 (Hare,
1991).
1
Factor 1 (F1) is related to the interpersonal
and affective components of psychopathy (e.g.,
grandiosity, lying, lack of remorse or guilt), whereas
Factor 2 (F2) comprises traits and behaviors indica-
tive of “social deviance” (e.g., early behavior prob-
lems, juvenile delinquency, and impulsivity; Hare,
1991). The PCL/PCL-R factors typically manifest
different patterns of relations with external criteria.
For instance, F1 has been unrelated or negatively
associated with psychological distress and anxiety
(Harpur, Hare, & Hakstian, 1989; Schmitt &
Newman, 1999) and positively associated with social
dominance (Harpur et al., 1989) and emotional
detachment (Patrick, Bradley, & Lang, 1993), whereas
F2 has been positively associated with aggression,
criminality, and antisocial PD (Hare, 1991; Skeem &
Mulvey, 2001), recidivism (Hemphill, Hare, & Wong,
1998), substance use (Taylor & Lang, 2005), and dis-
tress (e.g., Verona, Patrick, & Joiner, 2001).
2
Lynam
and colleagues (Lynam, 2002; Lynam & Derefinko,
Authors’ Note: Correspondence concerning this article should
be addressed to Joshua D. Miller, Department of Psychology,
University of Georgia, Athens, GA 30602-3013; e-mail: jdmiller@
uga.edu.
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
2006; Widiger & Lynam, 1998) have argued that the
psychopathy factors can be understood via an exami-
nation of the basic personality traits that underlie
each factor. Both factors share an antagonistic inter-
personal approach and some degree of impulsivity
(although the relation with impulsivity is stronger for
F2). However, most measures of F2 are also related to
negative emotionality, which typically shows little
relation with measures of F1 (Lynam & Derefinko,
2006).
The PCL-R, however, is not without limitations as
an assessment tool because it requires extensive train-
ing, a lengthy interview, and access to file informa-
tion pertaining to official criminal records and
institutional behavior, which is not typically available
in noninstitutionalized populations. In response to
these limitations, several self-report measures have
been used to assess psychopathy in these populations.
Lilienfeld and Fowler (2006) articulate several
advantages (and disadvantages) to the use of self-
report psychopathy instruments, including “econ-
omy,” the elimination of concerns regarding interrater
reliability, and the ability (of some measures) to
assess for systematic response styles.
Self-report measures of psychopathy differ in the
degree to which they are designed to be congruent with
the PCL-R. For example, Lilienfeld (1990) developed
the Psychopathic Personality Inventory (PPI) by includ-
ing items derived from a variety of conceptualiza-
tions of psychopathy. Although the PPI includes eight
subscales, this measure has been found to conform
to a two-factor structure (Benning, Patrick, Hicks,
Blonigen, & Krueger, 2003; Lilienfeld & Widows,
2005). Additional self-report measures such as the Self-
Report Psychopathy Scale (SRP; Hare, 1985), the
Levenson Self-Report Psychopathy Scale (LSRP;
Levenson, Kiehl, & Fitzpatrick, 1995), and the
Antisocial Process Screening Device (APSD; Frick &
Hare, 2001) were designed to assess psychopathy, but
unlike the PPI, were developed based on the PCL/PCL-
R. Given the prominent role of the PCL-R, and its
factor structure, in the development of these measures,
they too assess psychopathy using a two-factor struc-
ture. Because of their alliance with the PCL-R struc-
ture, the first factor of these models is thought to assess
core interpersonal and affective features. Alternatively,
the second factor focuses on traits and behaviors related
to antisocial behavior (ASB) and social deviance,
including impulsivity. The LSRP’s factors (i.e., primary
vs. secondary) are named after a distinction put forth
by Karpman (1941), who believed that there were two
different types of individuals who engaged in similar
behavior but as a result of different etiologies. Primary
psychopaths were believed to be born with a predispo-
sition toward callousness, whereas secondary psy-
chopaths “behaved badly” as a result of environmental
factors. Although this distinction has been the subject of
increasing (and promising) empirical study of late (e.g.,
Skeem, Johansson, Andershed, Kerr, & Eno Louden,
2007), we refer to the LSRP factors as F1 and F2 to tie
these factors into the PCL-R literature from which they
were derived.
In the current study, we examine the nomological
network surrounding one of these measures, the LSRP.
The LSRP total score has been linked in predictable
ways with ASB and substance use (Brinkley, Schmitt,
Smith, & Newman, 2001; Lynam, Whiteside, & Jones,
1999), basic personality dimensions (e.g., low scores on
Agreeableness, Conscientiousness; Lynam, 2002;
Lynam et al., 1999; Ross, Lutz, & Bailley, 2004), alter-
native measures of psychopathy (Elwood, Poythress, &
Douglas, 2004; Hicklin & Widiger, 2005), and passive
avoidance deficits (Epstein, Poythress, & Brandon,
2006; Lynam et al., 1999). However, there have been
questions surrounding the validity of the LSRP factors
and their congruence with the PCL-R factors. Lilienfeld
and Fowler (2006) suggested that LSRP F1 is “more
highly related to measures of secondary psychopathy
and ASBs than to measures of the core affective and
interpersonal features of psychopathy” (p. 118). This
conclusion is based on findings that (a) LSRP F1 is
equally related to both PCL-R factors (Brinkley et al.,
2001), (b) LSRP F1 is correlated more strongly with F2
scores from certain alternative measures of psychopa-
thy (e.g., PPI or SRP; Lilienfeld & Hess, 2001; Wilson,
Frick, & Clements, 1999), and (c) that the LSRP F1 and
F2 are equally correlated with ASB (e.g., Levenson
et al., 1995; McHoskey, Worzel, & Szyarto, 1998). As
such, it is important to determine the nature of the
LSRP factors.
To characterize the LSRP factors, we examine the
interrelation of the LSRP factor scores and their rela-
tions with two sets of individual differences con-
structs: general personality traits and PDs. These
constructs were chosen because they have consistently
been included as part of the examination of conver-
gent and discriminant validity for psychopathy mea-
sures in general (e.g., Lilienfeld & Windows, 2005),
and the PCL-R specifically (Hare, 1991). We present
both bivariate correlations between the LSRP scores
and the external criteria, as well as correlations using
residualized LSRP factors, to be methodologically
Miller et al. / Levenson Self-Report Psychopathy Scale 451
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
consistent with extant literature using the PCL-R (e.g.,
Verona et al., 2001).
3
We focus primarily on the bivari-
ate correlations in the text, however, because partial
correlations may be difficult to interpret as some have
argued that “partial variables are abstractions, existing
only in the statistical ether” (Miller & Lynam, 2006,
p. 1472). To test the construct validity of the LSRP
factors, particularly LSRP F1, we first review existing
data on the relations between the psychopathy factors
and these constructs (i.e., personality, PDs). We focus
on the extant research that used the PCL-R to assess
psychopathy because it is the largest, most compre-
hensive collection of empirical findings. This is not to
suggest, however, that the PCL-R conceptualization of
psychopathy is the only or best way to conceive of this
construct.
PCL-R Factors and General
Personality Traits
In a recent meta-analysis, Lynam and Derefinko
(2006) examined the relations between psychopathy
factors and a “consensus big four” of personality
traits; of the eight studies used, four used the PCL-R.
For F1, the strongest weighted effect size was for
Agreeableness (r = –.46), followed by Conscientious-
ness (r = –.22). For F2, there were significant nega-
tive effect sizes for Conscientiousness (r = –.45),
Agreeableness (r = –.44), and Extraversion (r = –.12),
and a positive effect size for Neuroticism (r = .34).
These authors argue that the two factors are “rela-
tively highly correlated because of the overlap with
low A (Agreeableness) but are divergent because the
factors differentially assess the remaining three
dimensions” (p. 147).
PCL-R Factors and PDs
Widiger (2006) notes that psychopathy has been
examined most frequently with regard to Antisocial
and Narcissistic PDs and that the factors appear to be
differentially related to the two PDs such that
Antisocial PD is more strongly linked to F2, whereas
Narcissistic PD is more strongly related to F1.
However, it is important to note that both PCL-R fac-
tors are typically significantly related to both PDs. In
fact, the majority of studies that have examined these
relations find that both PCL/PCL-R factors are sig-
nificantly positively associated with all four Cluster B
PDs (e.g., Antisocial, Borderline, Histrionic, and
Narcissistic), as well as Paranoid PD (e.g., Blackburn
& Coid, 1998; Hart, Forth, & Hare, 1991; Hart &
Hare, 1989; Hildebrand & de Ruiter, 2004;
Rutherford, Alterman, Cacciola, & McKay, 1997;
Salekin, Rogers, & Sewell, 1997; Soderstrom,
Nilsson, Sjodin, Carlstedt, & Forsman, 2005).
Correlations between the PCL-R factors and these
PDs show substantial variability: Paranoid (F1: range
of rs: .05-.48; median r: .22; F2: range of rs: .06-.47;
median r: .28), Antisocial (F1: range of rs: .24-.61;
median r: .40; F2: range of rs: .50-.87; median r: .71),
Borderline (F1: range of rs: –.03-.29; median r: .20;
F2: range of rs: .26-.45; median r: .35), Histrionic
(F1: range of rs: .09-.37; median r: .25; F2: range of
rs: .13-.28; median r: .21), and Narcissistic (F1: range
of rs: .08-.57; median r: .47; F2: range of rs: .24-.49;
median r: .28). Based on the median correlations
from previous research, one can conclude that F1 and
F2 typically diverge primarily in their relations to
Antisocial and Borderline PDs (higher in F2), as well
as Narcissistic PD (higher in F1).
Current Study
We believe that it may be premature to draw con-
clusions about the (in)validity of the LSRP F1. Given
the aforementioned criticisms, we focus on the per-
sonality features (traits and disorders) that underlie
the LSRP factors and compare them to the known
relations found for the PCL-R, because the LSRP was
“designed to produce, by means of a self-report pro-
cedure, two factors similar to those produced by the
Hare Psychopathy Checklist” (Levenson et al., 1995,
p. 152). We examine the correlations between LSRP
F1 and F2 and a comprehensive measure of general
personality (i.e., the Revised NEO Personality
Inventory; NEO PI-R; Costa & McCrae, 1992),
expert ratings of prototypical psychopathy using the
NEO PI-R data, as well as a self-report measure of
the DSM-IV PDs and a measure of trait narcissism.
Specifically, we test the following hypotheses.
First, following Lynam and Derefinko’s (2006)
meta-analysis, we hypothesize that LSRP F1 should
be most strongly correlated with Agreeableness (neg-
atively), followed by Conscientiousness (negatively).
Alternatively, LSRP F2 should be significantly nega-
tively related to Agreeableness and Conscientious-
ness (with similar effect sizes for both domains), as
well as Extraversion but to a lesser degree. LSRP F2
should also manifest a significant positive relation
with Neuroticism.
Second, we hypothesize that Agreeableness and, to
a lesser extent, Conscientiousness will account for
452 Assessment
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
the significant correlation between the two LSRP fac-
tors. We will examine the semipartial correlations
between the LSRP factors in two ways, once control-
ling only for Agreeableness and once controlling for
both Agreeableness and Conscientiousness.
Third, on the basis of previous results found for the
PCL-R and PDs, we hypothesize that both LSRP fac-
tors will be correlated with Paranoid PD and the four
Cluster B PDs. We expect, however, that LSRP F1
will be more strongly correlated with Narcissistic PD
and trait narcissism, as well as an expert-rated mea-
sure of prototypical psychopathy, whereas LSRP F2
will be more strongly correlated with Antisocial and
Borderline PDs.
Method
Participants
Participants were 271 undergraduates. Fifty-six
percent of the participants were women, 86% were
White, and the mean age was 19.3 years (SD = 1.3).
Procedure
Participants took part in the 1- to 2-hour experi-
mental protocol during one session, in groups of
approximately 30 individuals. After completion of
informed consent, participants completed all self-
report questionnaires and were debriefed by the
experimenter.
Measures
Levenson Self-Report Psychopathy (LSRP) Scale.
The LSRP is a 26-item self-report measure of psy-
chopathy. Items are scored on a 1 (Disagree Strongly)
to 4 (Agree Strongly) scale. In the current study, coef-
ficient alphas for the LSRP total score (26 items), F1
(16 items), and F2 (10 items) were .83, .82, and .61,
respectively. The lower alpha for F2 is quite consistent
with previous studies of this measure: Hicklin and
Widiger (2005; α=.66). Lynam et al. (1999; α=.68),
Miller, Lynam, Widiger, and Leukefeld (2001; α=
.63), and Ross et al. (2004; α=.62).
NEO PI-R. The NEO PI-R (Costa & McCrae,
1992) is a 240-item self-report measure of the Five-
Factor Model (FFM) of personality, which includes
five broad domains of Neuroticism, Extraversion,
Openness, Agreeableness, and Conscientiousness.
Each of these five domains is underlaid by six facets.
In the current sample, alphas for the domains ranged
from .87 to .92, whereas the facets ranged from .56 to
.85 with a median of .78. These findings are consis-
tent with those reported in the NEO PI-R manual
(Costa & McCrae, 1992).
Personality Disorder Questionnaire-4
+
(PDQ-4
+
).
The PDQ-4+ (Hyler, 1994) is a 99-item self-report
measure of DSM-IV PDs on which items are
answered using a Yes or No response format. PD
symptom counts were computed by summing the
items endorsed for each PD. Widiger and Coker
(2001) suggest that the PDQ-4+ is one of, if not the,
most commonly used self-report measures of PD
symptoms and that it is the most “directly coordi-
nated with the DSM-IV personality disorder diagnos-
tic criteria” (p. 412).
FFM Psychopathy Resemblance Index (PRI). The
FFM PRI was calculated as an intraclass correlation
between participants’ obtained NEO PI-R facets scale
scores and the expert generated facet profile for psy-
chopathy, as described in Miller et al. (2001). The PRI
has demonstrated significant correlations with other
measures of psychopathy (Derefinko & Lynam, 2006;
Miller et al., 2001) and relevant external variables such
as delinquency, substance use, aggression, and risky sex
(Miller & Lynam, 2003).
Narcissistic Personality Inventory (NPI). The NPI
(Raskin & Terry, 1988) is a 40-item self-report mea-
sure of trait narcissism (α=.88 in the current study).
Results
Preliminary Analyses
To control for Type I error, we lowered our signif-
icance level for all analyses to p ≤ .01. We first exam-
ined whether there were gender differences in levels
of psychopathy. For the LSRP total score, there was a
significant difference, t(269) = 3.35, p ≤ .01, d = .39,
such that men had higher scores. This difference,
however, was driven by gender differences on the
LSRP factors. There was no significant gender differ-
ence on F2, t(269) = 1.45, ns, d = .19. However, there
was a significant gender difference for F1, t(269) =
3.74, p ≤ .01, d = .45, such that men had higher scores.
Next, we tested whether the correlations between the
LSRP factors and the external variables were statisti-
cally different across gender. All correlations were
tested separately for men and women (see Cohen &
Cohen, 1983, test for independent rs); of all the pairs
Miller et al. / Levenson Self-Report Psychopathy Scale 453
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
of correlations tested, only three significant differ-
ences were found. As such, we report correlations in
which men and women are combined.
Bivariate Relations Between LSRP
Scores and General Personality Traits
The LSRP total score was significantly positively
correlated with the domain of Neuroticism and nega-
tively with the domains of Extraversion, Agreeableness,
and Conscientiousness (see Table 1). Examining the
correlations between the LSRP factors and the NEO PI-
R reveals substantive differences in the two personality
profiles. The two sets of personality correlates were
tested to see whether the correlations across the LSRP
factors were significantly different (see Cohen &
Cohen, 1983, p. 56, test for dependent rs). Of the 35
pairs of correlations, 18 were significantly different
(51%; p ≤ .01). These differences are indicated in
Table 1.
454 Assessment
Table 1
Relations Between LSRP Scores and NEO PI-R Domains and Facets
LSRP Total LSRP Factor 1 LSRP Factor 2 LSRP Factor 1 Residual LSRP Factor 2 Residual
Neuroticism .31** .17** .43** –.03 .40**
Anxiety .09 .04 .15 –.04 .15
Angry hostility .37** .34** .29** .23** .15
Depression .24** .11 .39** –.08 .38**
Self-consciousness .09 –.03 .26** –.16** .31**
Impulsiveness .23** .14 .31** –.01 .28**
Vulnerability .29** .14 .45** –.08 .44**
Extraversion –.22** –.17** –.24** –.07 –.18**
Warmth –.37** –.33** –.29** –.22** –.15
Gregariousness –.09 –.07 –.09 –.03 –.06
Assertiveness –.14 –.06 –.23** .06 –.23**
Activity –.11 –.04 –.19** .05 –.19**
Excitement-seeking .08 .09 .04 .08 .00
Positive emotions –.36** –.32** –.29** –.21** –.16**
Openness .03 .00 .07 –.03 .08
Fantasy .14 .09 .17** .01 .15
Aesthetics –.04 –.08 .05 –.11 .09
Feelings –.17** –.18** –.08 –.16** .00
Actions .01 –.02 .04 –.04 .05
Ideas .08 .08 .06 .05 .03
Values .07 .10 –.01 .11 –.06
Agreeableness –.62** –.66** –.32** –.58** –.01
Trust –.46** –.44** –.32** –.33** –.13
Straightforwardness –.61** –.65** –.33** –.56** –.03
Altruism –.52** –.52** –.34** –.41** –.11
Compliance –.40** –.40** –.26** –.32** –.08
Modesty –.32** –.43** –.01 –.48** .21**
Tender-mindedness –.34** –.41** –.10 –.42** .11
Conscientiousness –.44** –.23** –.66** .08 –.62**
Competence –.30** –.12 –.51** .13 –.52**
Order –.24** –.07 –.45** .15 –.47**
Dutifulness –.41** –.27** –.50** –.04 –.43**
Achievement striving –.30** –.13 –.49** .11 –.49**
Self-discipline –.37** –.17** –.59** .12 –.57**
Deliberation –.43** –.31** –.49** –.09 –.39**
Adjusted R
2
.58** .52** .57**
Note: LSRP = Levenson Self-Report Psychopathy Scale; NEO PI-R = Revised NEO Personality Inventory. Bolded pairs indicate a sig-
nificant difference (p ≤ .01) exists between the two LSRP factors.
**p ≤ .01.
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
For Neuroticism, both LSRP F1 and F2 were signif-
icantly related at the domain level; however, the corre-
lation between LSRP F2 and Neuroticism was
significantly stronger. The two were also equally sig-
nificantly correlated with the facet score of angry hos-
tility. However, only LSRP F2 was correlated with the
facets of depression, self-consciousness, impulsiveness,
and vulnerability; these sets of correlations were signif-
icantly different across the factors.
For Extraversion, both LSRP factors were equally
negatively correlated with the domain score and the
facets of warmth and positive emotions. LSRP F2
was also significantly negatively correlated with
facets of assertiveness and activity; only the correla-
tions for assertiveness were significantly different
across the LSRP factors.
For Openness, the LSRP factors were each signif-
icantly correlated with one facet. LSRP F1 was sig-
nificantly negatively correlated with openness to
feelings, whereas LSRP F2 was significantly posi-
tively correlated with openness to fantasy. Neither of
these correlations was significantly different across
the LSRP factors.
For Agreeableness, both factors were significantly
negatively correlated with the domain score and the
facets of trust, straightforwardness, altruism, and com-
pliance. Only LSRP F1 was significantly negatively
correlated with the facets of modesty and tendermind-
edness. All these correlations, with the exception of
those for trust and compliance, were significantly dif-
ferent across the factors such that F1 was more strongly
negatively correlated with the domain and facets.
For Conscientiousness, both LSRP factors were sig-
nificantly negatively correlated with the domain score
and the facets of dutifulness, self-discipline, and deliber-
ation. LSRP F2 was also significantly negatively corre-
lated with facets of competence, order, and achievement
striving. The sets of correlations were significantly dif-
ferent between the LSRP factors and the domain and
facets score such that LSRP F2 was more strongly nega-
tively correlated with all facets of Conscientiousness.
Finally, we also ran three simultaneous regression
analyses in which the LSRP scores were regressed on
the 30 NEO PI-R facets. Examining the adjusted
R-squared values, the facets accounted for 52% (F1),
57% (F2), and 58% (LSRP Total) of the variance.
Partial Relations Between LSRP
Scores and General Personality Traits
Next, we examined the correlations between the
residualized LSRP factors and the NEO PI-R
domains and facets. Every significantly different corre-
lation between a NEO PI-R score and the two LSRP
factors in the bivariate correlations (e.g., Neuroticism
and F1 [r = .17]; Neuroticism and F2 [r = .43]) was also
significantly different using the partial correlations
(e.g., Neuroticism and residualized F1 [r = –.03];
Neuroticism and residualized F2 [r = .40]). The primary
differences between the bivariate and partial correla-
tions were as follows: (a) LSRP F1 was no longer pos-
itively related to Neuroticism (except for the facet of
angry hostility) or negatively related to Conscientious-
ness, and (b) LSRP F2 was no longer significantly neg-
atively related to Agreeableness. Overall, the two sets
of findings converge on the primary importance of low
Agreeableness for LSRP F1 and high Neuroticism and
low Conscientiousness for LSRP F2.
Overlap Between LSRP F1 and F2
As expected, LSRP F1 and F2 were significantly
correlated, r = .46, p ≤ .01. Next, we examined the
hypothesis that the two psychopathy factors might be
related because of the shared role of Antagonism and
Conscientiousness (to a lesser degree) that appear to
be central to both. To examine this in the current
data, we examined the semipartial correlation between
LSRP F1 and F2, when controlling for the domain of
Agreeableness alone and then using both Agreeableness
and Conscientiousness. The semipartial correlation
between LSRP F1 and F2, controlling for Agreeable-
ness was .27, p ≤ .01. The semipartial correlation between
LSRP F1 and F2, controlling for Agreeableness and
Conscientiousness, was .20, p ≤ .01.
Bivariate Relations Between LSRP
Scores and Personality Pathology
We examined the correlations between the
LSRP total score and the DSM-IV PD scores obtained
from the PDQ-4+. The correlations ranged from
.04 (Obsessive-Compulsive PD [OCPD]) to .46
(Narcissistic PD) with a median of .30. We also
examined the correlations between the LSRP total
score and two other PD scores: an FFM-based PRI
and NPI narcissism. The LSRP total score was posi-
tively correlated with both scores.
The correlations between LSRP F1 and the DSM-
IV PDs ranged from –.01 (OCPD) to .47 (Narcissistic
PD) with a median correlation of .24 (see Table 2).
The correlations between LSRP F2 and the DSM-IV
PDs ranged from –.11 (OCPD) to .44 (Borderline
PD) with a median of .33. Overall, LSRP F1 was
more strongly correlated with Narcissistic PD,
Miller et al. / Levenson Self-Report Psychopathy Scale 455
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
whereas LSRP F2 was more strongly correlated with
Borderline, Avoidant, and Dependent PDs. It is worth
noting that both factors were significantly positively
correlated with all DSM-IV PDs with the exception of
OCPD. With regard to the alternative measures of
personality pathology, both LSRP factors were corre-
lated with the FFM psychopathy score, whereas only
LSRP F1 was significantly correlated with NPI nar-
cissism. Overall, the correlations between LSRP F1
and FFM psychopathy and NPI narcissism were sig-
nificantly stronger than those found for F2.
Partial Relations Between LSRP
Scores and Personality Pathology
Finally, we examined the correlations between
the residualized LSRP variables and the PDs (see
Table 2). The significant differences found between
the residualized LSRP factors and the PDs remained
the same as those found for the bivariate correlations.
The residualized LSRP F1 variable was significantly
positively correlated with Paranoid, Antisocial, and
Narcissistic PDs (as measured by the PDQ-4+ and the
NPI variables), and FFM Psychopathy. Alternatively,
the residualized LSRP F2 variable was significantly
positively correlated with Paranoid, Schizotypal,
Borderline, Histrionic, Avoidant, and Dependent PDs.
Discussion
This study examined the relations between the
LSRP factors and both general personality domains
and facets, as measured by the NEO PI-R, and DSM
and non-DSM PD scores. The primary goal of the
study was to test the validity of these psychopathy
factors, particularly LSRP F1. We chose these exter-
nal variables because they have been regularly used
to examine the convergent and discriminant validity
of psychopathy factors from various instruments
including the PCL-R. We interpret our current find-
ings in the context of the known nomological network
that surrounds the psychopathy factors, primarily as
studied using the PCL/PCL-R.
Factor Interrelations
Much like the results for the PCL-R, which finds
that the two factors are strongly correlated at approxi-
mately .50 (e.g., Hare, 1991), the LSRP factors were
significantly interrelated (r = .46). This is consistent
with other studies of the LSRP, which typically report
a correlation between the factors that is close to .50
(e.g., .54; Epstein et al., 2006; .43; Lynam et al., 1999).
We noted earlier that Lynam and colleagues (e.g.,
Lynam, 2002; Lynam & Derefinko, 2006) have argued
that the role of Agreeableness in both psychopathy fac-
tors should explain the significant degree of overlap
between measures of F1 and F2. We found some sup-
port for this in that the semipartial correlation between
the LSRP factors, once Agreeableness was controlled
for, was substantially smaller (i.e., sr = .27) than the
bivariate correlation between the factors. However, there
was still a significant relation between the two factors.
Given Lynam and Derefinko’s meta-analysis, we
expected that the combination of both Agreeableness
456 Assessment
Table 2
Relations Between LSRP Scores and Personality Disorder Symptomatology
Personality Disorder LSRP Total LSRP Factor 1 LSRP Factor 2 LSRP Factor 1 Residual LSRP Factor 2 Residual
Paranoid .45** .40** .36** .27** .19**
Schizoid .22** .19** .18** .13 .10
Schizotypal .26** .21** .24** .11 .16**
Antisocial .45** .38** .43** .21** .28**
Borderline .38** .27** .44** .08 .36**
Histrionic .33** .27** .31** .14 .21**
Narcissistic .46** .47** .27** .39** .06
Avoidant .26** .16** .34** .00 .30**
Dependent .26** .16** .34** .01 .30**
OCPD .04 –.01 –.11 –.07 .13
FFM psychopathy .42** .47** .17** .44** –.05
NPI narcissism .22** .31** –.03 .47** –.03
Note: LSRP = Levenson Self-Report Psychopathy Scale; OCPD = Obsessive Compulsive Personality Disorder; FFM = Five-Factor
Model; NPI = Narcissistic Personality Inventory. Bolded pairs indicate a significant difference (p ≤ .01) exists between the two LSRP
factors.
**p ≤ .01.
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
and Conscientiousness would account for a greater
percentage of the relation between the psychopathy
factors. Again, a semipartial correlation controlling for
these two FFM domains provided partial support as
their inclusion in the analysis reduced the relation
between LSRP F1 and F2 even further (i.e., sr = .20).
However, the relation between the factors remained
significant suggesting that either (a) the LSRP factors
manifest overlap because of item content beyond that
shared with Agreeableness and Conscientiousness or
(b) the shared methodology used to assess the psy-
chopathy factors is responsible for the remaining
interrelation. We suspect that both these factors are at
work in explaining why these two FFM domains
do not entirely account for the relation between LSRP
factors.
General Personality Correlates
Based on Lynam and Derefinko’s (2006) meta-
analysis, we made specific hypotheses for how the
LSRP factors would relate to the FFM domains. For
example, we believed that LSRP F1 would be most
strongly correlated with Agreeableness, followed by
Conscientiousness. Next, we hypothesized that LSRP
F2 would be significantly negatively related to
Agreeableness and Conscientiousness (with similar
effect sizes), as well as Extraversion but to a lesser
degree. We also expected LSRP F2 to manifest a sig-
nificant positive relation with Neuroticism.
In general, the LSRP factors were correlated with
the NEO PI-R in a manner consistent with our hypothe-
ses. LSRP F1 was strongly negatively correlated with
Agreeableness and manifested a moderate negative
correlation with Conscientiousness. Alternatively,
LSRP F2 was most strongly related to Conscientious-
ness (negatively), followed by Agreeableness (nega-
tively) and Neuroticism (positively). The relations
between the LSRP factors and the domains of
Agreeableness and Conscientiousness were quite sim-
ilar to those found in the Lynam and Derefinko (2006)
meta-analysis. For instance, both found substantial
correlations between F1 scores and Agreeableness
(i.e., current study: LSRP F1, r = –.62; meta-analysis:
weighted mean r = –.46) and smaller but significant
correlations with Conscientiousness (i.e., current
study: LSRP F1, r = –.32; meta-analysis: weighted
mean r = –.22). Similarly, F2 scores across studies
were also significantly associated with Agreeableness
(i.e., current study: LSRP F2, r = –.32; meta-analysis:
weighted mean r = –.44) and Conscientiousness (i.e.,
current study: LSRP F2, r = –.66; meta-analysis:
weighted mean r = –.45). Ultimately, the main differ-
ences between the meta-analysis and the current
results were that LSRP F1 was significantly positively
related to Neuroticism (r = .17 vs. a weighted mean r
of .01 in the meta-analysis) and negatively related to
Extraversion (r = –.17 vs. a weighted mean r of .00 in
the meta-analysis). However, the facet-level analyses
are helpful in interpreting the domain-level analyses
and understanding why these findings diverge from
the aforementioned meta-analysis. For instance,
although the Neuroticism domain was significantly
correlated with LSRP F1, this relation was driven pri-
marily by the angry hostility facet as the other five
Neuroticism facets were unrelated to F1 (in fact,
Neuroticism was not significantly related to LSRP F1,
if the angry hostility facet was removed; r = .10). This
is not surprising because experts rated this trait as
being somewhat elevated in prototypically psycho-
pathic individuals (Miller et al., 2001). However, one
does not see evidence of a significant negative relation
between LSRP F1 and the experience of negative
emotions, as one might expect based on Cleckley’s
(1941) conceptualization of psychopathy. The role of
the PCL-R in the derivation of the LSRP may partially
explain this finding. Although the PCL/PCL-R was
largely based on Cleckley’s conceptualization, Hare
did not include low anxiety/distress as one of the cri-
teria. As a result, subsequent measures such as the
LSRP do not include specific content aimed at captur-
ing traits related to fearlessness and low anxiety. In
addition, the correlations between the residualized
LSRP F1 and the Neuroticism domain and facets sug-
gest that the relations between LSRP F1 and negative
affectivity largely work as one might expect (on the
basis of work with the PCL-R), if one controls for the
negative emotionality that is an important component
of LSRP F2.
The findings for LSRP F1 and Extraversion also
diverge from what one might expect on the basis of
the Lynam and Derefinko (2006) meta-analysis.
However, a facet-level analysis suggests that this
finding was due primarily to two significant facets:
warmth and positive emotions. Although negative
relations with these two traits are consistent with pre-
vious conceptualizations of psychopathy (see
Widiger & Lynam, 1998), the overall finding does
not, however, explain the lack of a significant relation
between F1 and Extraversion facets such as assertive-
ness. Although neither the PCL/PCL-R nor Cleckley’s
conceptualization explicitly includes a trait of this
Miller et al. / Levenson Self-Report Psychopathy Scale 457
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
type, one does typically find a relation between dom-
inance and dominance-related psychopathology (i.e.,
narcissism) and F1 scores of psychopathy (e.g.,
Kosson, Steuerwald, Forth, & Kirkhart, 1997). The
lack of a significant negative correlation between
LSRP and self-consciousness (a trait that may help
explain the glib, superficial charm; Widiger & Lynam,
1998) and a positive correlation with assertiveness
suggests that the LSRP F1 is not capturing some of
the “smooth,” charming, and dominant characteristics
thought to be an integral aspect of psychopathy in
general, and F1 specifically (e.g., Kosson et al., 1997;
Zolondek, Lilienfeld, Patrick, & Fowler, 2006).
These specific interpersonal behaviors appear to be
based on the traits of high dominance and low anxi-
ety, which are not well captured by the LSRP F1.
However, it is worth noting that this may be a general
limitation of self-report psychopathy measures; that
is, none of the self-report measures appear to capture
the entirety of the interpersonal traits associated with
F1. For instance, although PPI F1 includes domi-
nance (in the form of a Social Potency scale), it fails
to capture the antagonistic interpersonal approach
that is common to this factor (e.g., Benning, Patrick,
Salekin, & Leistico, 2005; Derefinko & Lynam,
2006).
PD Correlates
The LSRP factors also diverge in their relation to
other PDs. As expected, both LSRP factors were sig-
nificantly correlated with Paranoid PD, along with all
the Cluster B PDs. However, perhaps because of
issues relating to common method variance, both
LSRP factors were correlated with the majority of
DSM-IV PDs (with the exception of OCPD). There
were differences in the strengths of these relations
across the psychopathy factors. For instance, as
expected, F1 was more strongly related to Narcissistic
PD, whereas F2 was more strongly related to
Borderline PD. In addition, F1 was more strongly pos-
itively related to alternative, validated measures of
psychopathy and narcissism (i.e., FFM psychopathy,
and NPI narcissism). Consistent with the findings
based on the NEO PI-R, LSRP F2 appears to capture a
broader construct associated with increased negative
affect and emotional dysregulation, antagonism, and
impulse control problems, whereas F1 demonstrates a
more specific pattern, in which it is primarily linked
with disorders that share an antagonistic core of traits
(e.g., Antisocial, Narcissistic, prototypical psychopa-
thy). These findings are generally consistent with
those found for the PCL-R. As noted earlier, although
both factors are typically correlated with all the
Cluster B PDs, Antisocial and Borderline PD are usu-
ally more strongly related to PCL-R F2, and
Narcissistic PD is usually more strongly related to
PCL-R F1 (e.g., Blackburn & Coid, 1998; Hart et al.,
1991; Hart & Hare, 1989; Hildebrand & de Ruiter,
2004; Rutherford et al., 1997; Salekin et al., 1997;
Soderstrom et al., 2005).
Unlike the PCL-R factors, however, both LSRP fac-
tors demonstrated a wider array of significant corre-
lates with the PDs, including non–Cluster B PDs.
Although this is not surprising for F2 because of the
significant role of neuroticism in LSRP F2 and almost
all the PDs (e.g., see Saulsman & Page, 2004, for a
review), it is more surprising for LSRP F1. This may
be because of the fact that LSRP F1 appears to include
more negative affectivity than one might expect for this
factor. Even a small degree of neuroticism in a con-
struct (e.g., LSRP F1) is likely to result in significant
correlations with PDs such as Avoidant and
Dependent, which are heavily laden with neuroticism
(see Lynam & Widiger, 2001). Regardless of the expla-
nation, these findings are incongruent with previous
research examining the relations between the
PCL/PCL-R and the PDs and are problematic for the
current LSRP F1 scale.
Does the LSRP F1 Measure the
Interpersonal and Affective Aspects of
Psychopathy?
As noted earlier, the LSRP F1 has been criticized
for being more strongly related to “measures of sec-
ondary psychopathy and ASB than to measures of the
core affective and interpersonal features of psychopa-
thy” (Lilienfeld & Fowler, 2006, p. 118). We feel that
the present results are not entirely consistent with this
criticism. These core features, as assessed by the PCL-
R, include items such as glibness/superficial charm,
grandiose sense of self-worth, pathological lying,
conning/manipulative, lack of remorse/guilt, shallow
affect, callousness/lack of empathy, and failure to
accept responsibility for actions. Widiger and Lynam
(1998) translated each of these PCL-R items into their
underlying FFM traits and argued that almost all of
these core traits of PCL-R F1 could be translated into
Agreeableness facets. Given this, the strong correla-
tion between LSRP F1 and Agreeableness (r = –.66)
suggests that the LSRP F1 is capturing a substantial
component of the core interpersonal and affective
458 Assessment
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
components of psychopathy. Consistent with this
notion were the significant relations observed between
LSRP F1 and indices of narcissism and prototypical
psychopathy based on the FFM. It is worth noting,
however, that the LSRP F1 does not appear to assess
certain interpersonal and affective traits as well as one
might expect to see; that is, traits related to being glib
and superficially charming (PCL-R Item 1) and inter-
personally dominant do not appear to be captured by
the LSRP F1 scale. Although the LSRP captures the
callous, cold, immodest, and deceitful style that is
consistent with F1, it appears to miss some of the
other traditional features of this psychopathy factor.
In general, the current relations between the LSRP
factors and general personality traits are quite consis-
tent with results observed in previous studies (e.g.,
Lynam, 2002; Ross et al., 2004). The current findings,
paired with extant research, suggest that the LSRP F1
captures many of the important traits associated with
PCL-R F1 (e.g., grandiosity, callousness, manipula-
tiveness), misses some traits (e.g., glib/superficial
charm, dominance), and includes some that should
not be included (i.e., related to negative emotional-
ity). It is our opinion that some of the problems with
the LSRP F1 that were identified by Lilienfeld and
Fowler (2006) require further scrutiny. Although not
directly addressed here, these issues deserve com-
ment. As noted earlier, the criticism of the LSRP F1
appears to be based on three observations: (a) LSRP
F1 is equally related to both PCL-R factors (Brinkley
et al., 2001), (b) LSRP F1 is correlated more strongly
with F2 scores from certain measures of psychopathy
(e.g., PPI or SRP; Lilienfeld & Hess, 2001; Wilson
et al., 1999), and (c) the LSRP F1 and F2 are
equally correlated with ASB. Although each of these
criticisms has some merit, we do not believe they
invalidate the LSRP scales. We address these points
further below.
The first observation that the LSRP F1 correlates
equally strongly with both PCL-R factors (Brinkley
et al., 2001) appears to be a general shortcoming of
self-report measures of psychopathy. For instance,
Berardino, Meloy, Sherman, and Jacobs (2005) found
that PPI F1 was equally correlated with both PCL-R
F1 (r = .38) and PCL-R F2 (r = .32). Similarly,
Poythress, Edens, and Lilienfeld (1998) presented
data indicating that the eight PPI scales do not corre-
late with the PCL-R factors in a manner that suggests
that the PPI factors are convergent with the PCL-R
factors. Specifically, none of the eight PPI factors
manifested significantly different correlations with
the PCL-R factors. Finally, Zolondek et al. (2006)
found only small-to-moderate convergent correla-
tions between PPI factors and PCL-R factors (i.e., rs
of .25 and .37) and found only partial support for the
discriminant validity of such factors. It appears that
self-report psychopathy measures, while correlating
significantly with the PCL-R total scores, do not
show substantial convergent or discriminant validity
with the PCL-R factors.
The second observation is that the LSRP F1 corre-
lates more strongly with F2 scores than F1 scores
from other self-report psychopathy measures (e.g.,
PPI, SRP, APSD; e.g., Lilienfeld & Hess, 2001;
Wilson et al., 1999). Although this may be true, it
ignores the possibility that these alternative measures
of psychopathy are assessing F1 in a manner that is
not entirely consistent with the traditional F1 con-
struct. For example, although LSRP F1 and PCL-R F1
are significantly related to an antagonistic interper-
sonal orientation (e.g., Skeem, Miller, Mulvey,
Tiemann, & Monahan, 2005), the PPI and SRP assess
much less of this trait. For instance, the F1 scores for
both the PPI and SRP demonstrate much smaller
correlations with Agreeableness (rs = –.13 and –.11,
respectively; Benning, Patrick, Salekin, et al., 2005;
rs = –.19 and –.26, respectively; Derefinko & Lynam,
2006). The F1 scores for the PPI and SRP also demon-
strate a different pattern of relations with regard to
Neuroticism (both are strongly negatively correlated)
and Extraversion (both are significantly positively
correlated) that is not found with the PCL-R (e.g.,
Skeem et al., 2005). Finally, F1 and F2 in the PPI and
SRP are largely uncorrelated, whereas the PCL-R fac-
tors are strongly correlated with one another. The lack
of convergence between LSRP F1 and the F1 scores of
other measures may reflect basic differences in the
salience of traits such as Agreeableness, Neuroticism,
and Extraversion. Overall, one could argue that these
other self-report measures of F1 psychopathy are per-
forming in a manner (e.g., relations to Agreeableness)
that is inconsistent with the traits thought to be core to
this factor such as callousness, grandiosity, pathologi-
cal lying, and engaging in conning/manipulative
behavior.
The final criticism is that both factors of the LSRP
are equally correlated with ASB (e.g., Lynam et al.,
1999; McHoskey et al., 1998). Although both PCL-R
factors are typically correlated with ASB (e.g.,
Patrick, Hicks, Krueger, & Lang, 2005), the PCL-R
F2 is usually the stronger unique correlate (e.g.,
Patrick et al., 2005; Skeem & Mulvey, 2001).
Miller et al. / Levenson Self-Report Psychopathy Scale 459
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
However, this disparity might be due, in part, to over-
lap in predictor-criterion content. That is, the PCL-R
F2 may be the stronger correlate because of the inclu-
sion of items that explicitly reference previous ASB,
which are scored as part of F2. Removing these items,
as Cooke and Michie (2001) suggest, may well result
in PCL-R factors that are more equivalent in their rela-
tions with ASB (see Hall, Benning, & Patrick, 2004).
In addition, an equal correlation between both LSRP
factors and ASB might be preferable to the pattern
found for other self-report psychopathy measures. For
instance, the PPI F1 is unrelated to aggression
(Patrick, Edens, Poythress, Lilienfeld, & Benning,
2006), unrelated or negatively related to substance use
(Benning et al., 2003; Benning, Patrick, Blonigen,
Hicks, & Iacono, 2005; Patrick et al., 2006), and unre-
lated or weakly related to ASB (Benning et al., 2003;
Benning, Patrick, Blonigen, et al., 2005; Patrick et al.,
2006). Given that F1 scores are supposed to assess the
core affective and interpersonal traits of psychopathy
(e.g., grandiosity, manipulativeness, lack of empa-
thy/remorse, failure to accept responsibility), one
would expect these scores to demonstrate some rela-
tion with externalizing problems. Alternative mea-
sures of these interpersonal aspects of psychopathy,
such as the Interpersonal Measure of Psychopathy
(Kosson et al., 1997), do show significant relations
with ASB (Kosson et al., 1997; Zolondek et al., 2006).
As such, we do not believe it is particularly problem-
atic for both the LSRP factors to relate to antisocial,
externalizing behaviors.
Limitations
In the current study, our assessment of psychopa-
thy and personality (general traits and PDs) were
based solely on self-reports. Future research would
do well to examine these issues using other alterna-
tive methodologies such as a semistructured inter-
view for the assessment of the PDs and psychopathy
and/or data derived from objective behavioral tasks.
The reliance on self-report measures may have
inflated the current correlations because of the shared
method variance. In addition, the current data were
entirely cross-sectional. As such, we were not able to
assess the predictive validity of the LSRP factors; this
is an important task for future research because the
PCL-R has proven to be a relatively consistent pre-
dictor of problematic outcomes (Gretton et al., 2004;
cf. Guy, Edens, Anthony & Douglas, 2005).
Another limitation of the current study is that we
tested only one of the primary self-report measures of
psychopathy, the LSRP, rather than including multi-
ple specific measures of psychopathy such as the PPI,
SRP, and/or the APSD. However, the current results
help fill a void in the literature by placing the LSRP
and its factors in context with regard to other fre-
quently used psychopathy instruments. We feel the
current results have helped clarify the manner in
which the LSRP factors relate to important external
criteria and how these patterns converge and diverge
with the patterns demonstrated by the PCL-R and
other measures of psychopathy.
Finally, the current results are derived from under-
graduate students. Although the LSRP was designed
for use in noninstitutionalized samples, our results
would have been stronger if they had included data
from a community sample that may have been more
diverse with regard to demographics and levels of per-
sonality pathology.
Conclusions
Overall, our results are quite consistent with those
put forth by both Benning, Patrick, Salekin, et al.
(2005) and Derefinko and Lynam (2006) in suggest-
ing that the “factor 1” scores on various self-report
psychopathy measures appear to be assessing con-
structs that diverge in important, substantive ways.
We agree with the concerns raised by Derefinko and
Lynam suggesting that “researchers must acknowl-
edge the factors are not isomorphic, that results from
one instrument may not generalize, and to be specific
about which factor 1 they are examining” (p. 278).
However, we believe that criticisms levied against the
LSRP F1 may have been overstated. Although the
LSRP F1 may not capture all traits found in PCL-R
F1 psychopathy (e.g., glib/superficial charm; domi-
nance), it appears to capture successfully a number of
other traits that are central to the construct (e.g.,
grandiosity, callousness, selfishness, and manipula-
tiveness). In fact, we argue that it captures the antag-
onistic aspects of this part of psychopathy as well or
better than other existing self-report measures.
Modifying the scale would be necessary, however, if
one wanted to increase its fidelity with PCL-R F1 or
other conceptualizations of primary psychopathy.
This could be accomplished by increasing its loading
on agentic aspects of extraversion and decreasing its
relation with negative emotionality.
Overall, it is clear that further construct validation
and explication is necessary for all the self-report psy-
chopathy instruments to derive a consensus about the
460 Assessment
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
traits that are considered central to a measure of this
aspect of psychopathy. The lack of empirical conver-
gence between existing measures of F1, as well as the-
oretical divergence with classic descriptions of
psychopathy (e.g., Cleckley), suggests that further
refinement is necessary. As a field, we must determine
which basic traits are central to psychopathy; this
would enable an explicit comparison of the different
psychopathy measures to an a priori personality profile
to see which shows the greatest fidelity to the construct.
Notes
1. More recent factor analyses of the Psychopathy Checklist-
Revised (PCL-R) have resulted in either a three- (Cooke &
Michie, 2001) or four-factor model (e.g., Hare, 2003). Both these
models divide the “old” Factor 1 (F1) into two facets (e.g.,
Interpersonal, Affective). The third factor (e.g., Impulsive,
Irresponsible Lifestyle) in the three-factor model is made up of
the traits and behaviors from the “old” Factor 2 (F2) but does not
include the explicitly antisocial items. Alternatively, the four-
factor model divides the old F2 into two facets (e.g., Lifestyle;
Antisocial), one of which includes antisocial behavior (ASB).
Despite these changes in proposed factor structures for the PCL-
R, most self-report measures of psychopathy still result in a two-
factor solution that is akin to the initial factor structure of the PCL
and PCL-R.
2. It should be noted, however, that the degree of divergence
found for the PCL-R factors with external correlates is often
increased through the use of statistical methods such as partial
correlations or simultaneous regression analyses, in which the
authors control for the substantial overlap between the PCL-R
factors. As a result, the PCL-R factors often demonstrate more
divergent relations than would be found if bivariate correlational
analyses were used (see Lynam, Hoyle, & Newman, 2006).
3. Two residualized variables were created to examine the
unique effects of Levenson Self-Report Psychopathy Scale
(LSRP) F1 and F2. Specifically, each LSRP score was regressed
onto the other and the residuals were saved. Residualized F1 rep-
resents that part of F1 that cannot be predicted from F2; residual-
ized F2 represents that part of F2 that cannot be predicted from
F1. Although correlations with these variables represent semipar-
tial correlations, the procedure is generically referred to as par-
tialling, which we use here.
References
American Psychiatric Association. (1994). Diagnostic and statis-
tical manual of mental disorders (4th ed.). Washington, DC:
Author.
Benning, S. D., Patrick, C. J., Blonigen, D. M., Hicks, B. M., &
Iacono, W. G. (2005). Estimating facets of psychopathy from
normal personality traits: A step toward community epidemi-
ological investigations. Assessment, 12, 3-18.
Benning, S. D., Patrick, C. J., Hicks, B. M., Blonigen, D. M., &
Krueger, R. F. (2003). Factor Structure of the Psychopathic
Personality Inventory: Validity and implications for clinical
assessment. Psychological Assessment, 15, 340-350.
Benning, S. D., Patrick, C. J., Salekin, R. T., & Leistico, A. R.
(2005). Convergent and discriminant validity of psychopathy
factors assessed via self-report: A comparison of three instru-
ments. Assessment, 12, 270-289.
Berardino, S. D., Meloy, J. R., Sherman, M., & Jacobs, D. (2005).
Validation of the Psychopathic Personality Inventory on a
female inmate sample. Behavioral Sciences and the Law, 23,
819-836.
Blackburn, R., & Coid, J. (1998). Psychopathy and the dimen-
sions of personality disorder in violent offenders. Personality
and Individual Differences, 25, 129-145.
Brinkley, C. A., Schmitt, W. A., Smith, S. S., & Newman, J. P.
(2001). Construct validation of a self-report psychopathy
scale: Does Levenson’s self-report psychopathy scale measure
the same constructs as Hare’s psychopathy checklist-revised.
Personality and Individual Differences, 31, 1021-1038.
Cleckley, H. (1941). The mask of sanity. St. Louis, MO: Mosby.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/cor-
relation analysis for the behavioral sciences (2nd ed., pp. 53-
55). Hillsdale, NJ: Lawrence Erlbaum.
Cooke, D. J., & Michie, C. (2001). Refining the construct of psy-
chopathy: Towards a hierarchical model. Psychological
Assessment, 13, 171-188.
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality
Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-
FFI) professional manual. Odessa, FL: Psychological
Assessment Resources.
Derefinko, K. J., & Lynam, D. R. (2006). Convergence and diver-
gence among self-report psychopathy measures: A personality-
based approach. Journal of Personality Disorders, 20, 261-280.
Elwood, C., Poythress, N. G., & Douglas, K. S. (2004).
Evaluation of the Hare P-SCAN in a non-clinical population.
Personality and Individual Differences, 36, 833-843.
Epstein, M. K., Poythress, N. G., & Brandon, K. O. (2006). The
self-report psychopathy scale and passive avoidance learning:
A validation study of race and gender effects. Assessment, 13,
197-207.
Frick, P. J., & Hare, R. D. (2001). The Antisocial Process Screening
Device (APSD). Toronto, Canada: Multi-Health Systems.
Gretton, H. M., Hare, R. D., & Catchpole, R. E. (2004).
Psychopathy and offending from adolescence to adulthood: A
ten year follow-up. Journal of Clinical and Consulting
Psychology, 72, 636-645.
Guy, L. S., Edens, J. F., Anthony, C., & Douglas, K. S. (2005).
Does psychopathy predict institutional misconduct among
adults? A meta-analytic investigation. Journal of Consulting
and Clinical Psychology, 73, 1056-1064.
Hall, J., Benning, S. D., & Patrick, C. J. (2004). Criterion-related
validity of the three-factor model of psychopathy: Personality,
behavior, and adaptive functioning. Assessment, 11, 4-16.
Hare, R. D. (1985). A comparison of procedures for the assessment
of psychopathy. Journal of Consulting and Clinical Psychology,
53, 7-16.
Hare, R. D. (1991). Manual for the Psychopathy Checklist–
Revised. Toronto, Canada: Multi-Health Systems.
Hare, R. D. (2003). Manual for the Psychopathy Checklist–
Revised (2nd ed.). Toronto, Canada: Multi-Health Systems.
Harpur, T. J., Hare, R. D., & Hakstian, A. R. (1989). Two-factor
conceptualization of psychopathy: Construct validity and
assessment implications. Psychological Assessment: A
Journal of Consulting and Clinical Psychology, 1, 6-17.
Miller et al. / Levenson Self-Report Psychopathy Scale 461
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
Hart, S. D., Forth, A. E., & Hare, R. D. (1991). The MCMI-II and
psychopathy. Journal of Personality Disorders, 5, 318-327.
Hart, S. D., & Hare, R. D. (1989). Discriminant validity of the
psychopathy checklist in a forensic psychiatric population.
Psychological Assessment, 1, 211-218.
Hemphill, J. F., Hare, R. D., & Wong, S. (1998). Psychopathy and
recidivism: A review. Legal and Criminological Psychology, 3,
141-172.
Hicklin, J., & Widiger, T. A. (2005). Similarities and differences
among antisocial and psychopathic self-report inventories
from the perspective of general personality functioning.
European Journal of Personality, 19, 325-342.
Hildebrand, M., & de Ruiter, C. (2004). PCL-R psychopathy and
its relation to DSM-IV Axis I and II disorders in a sample of
male forensic psychiatric patients in the Netherlands.
International Journal of Law and Psychiatry, 27, 233-248.
Hyler, S. E. (1994). Personality Disorder Questionnaire–4.
Unpublished test, NYSPI.
Karpman, B. (1941). On the need of separating psychopathy into
two distinct clinical types: The symptomatic and the idiopathic.
Journal of Criminology and Psychopathology, 3, 112-137.
Kosson, D. S., Steuerwald, B. L., Forth, A. E., & Kirkhart, K. J.
(1997). A new method for assessing the interpersonal behav-
ior of psychopathic individuals. Preliminary validation stud-
ies. Psychological Assessment, 9, 89-101.
Levenson, M. R., Kiehl, K. A., & Fitzpatrick, C. M. (1995).
Assessing psychopathic attributes in a noninstitutionalized
population. Journal of Personality and Social Psychology, 68,
151-158.
Lilienfeld, S. O. (1990). Development and preliminary validation
of a self-report measure of psychopathic personality. Doctoral
dissertation, University of Minnesota, MN.
Lilienfeld, S. O., & Fowler, K. A. (2006). The self-report assess-
ment of psychopathy: Problems, pitfalls, and promises. In C. J.
Patrick (Ed.), Handbook of Psychopathy (pp. 107-132). New
York: Guilford.
Lilienfeld, S. O., & Hess, T. H. (2001). Psychopathic personality
traits and somatization: Sex differences and the mediating role
of negative emotionality. Journal of Psychopathology and
Behavioral Assessment, 23, 11-24.
Lilienfeld, S. O., & Windows, M. R. (2005). Psychopathic
Personality Inventory–Revised: Professional manual. Lutz,
FL: Psychological Assessment Resources.
Lynam, D. R. (2002). Psychopathy from the perspective of the
five-factor model. In P. T. Costa & T. A. Widiger (Eds.),
Personality disorders and the five-factor model of personality
(2nd ed., pp. 325-350). Washington, DC: American
Psychological Association.
Lynam, D. R., & Derefinko, K. (2006). Psychopathy and person-
ality. In C. J. Patrick (Ed.), Handbook of psychopathy
(pp. 133-155). New York: Guilford.
Lynam, D. R., Hoyle, R. H., & Newman, J. P. (2006). The perils
of partialling: Cautionary tales from aggression and psychopathy.
Assessment, 13, 328-341.
Lynam, D. R., Whiteside, S., & Jones, S. (1999). Self-reported
psychopathy: A validation study. Journal of Personality
Assessment, 73, 110-132.
Lynam, D. R., & Widiger, T. A. (2001). Using the five-factor
model to represent the DSM-IV personality disorders: An
expert consensus approach. Journal of Abnormal Psychology,
110, 401-412.
McHoskey, J. W., Worzel, W., & Szyarto, C. (1998).
Machiavellianism and psychopathy. Journal of Personality
and Social Psychology, 74, 192-210.
Miller, J. D., & Lynam, D. R. (2003). Psychopathy and the five-
factor model of personality: A replication and extension.
Journal of Personality Assessment, 81, 168-178.
Miller, J. D., & Lynam, D. R. (2006). Reactive and proactive
aggression: Similarities and differences. Personality and
Individual Differences, 41, 1469-1480.
Miller, J. D., Lynam, D., Widiger, T., & Leukefeld, C. (2001).
Personality disorders as an extreme variant of common per-
sonality dimensions: Can the Five Factor Model represent
psychopathy
. Journal of Personality, 69, 253-276.
Patrick, C. J., Bradley, M., & Lang, P. J. (1993). Emotion in the
criminal psychopath: Startle reflex modulation. Journal of
Abnormal Psychology, 102, 82-92.
Patrick, C. J., Edens, J. F., Poythress, N. G., Lilienfeld, S. O., &
Benning, S. D. (2006). Construct validity of the Psychopathic
Personality Inventory two-factor model with offenders.
Psychological Assessment, 18, 204-208.
Patrick, C. J., Hicks, B. M., Krueger, R. F., & Lang, A. R. (2005).
Relations between psychopathy facets and externalizing in a
criminal offender sample. Journal of Personality Disorders,
19, 339-356.
Porter, S., Birt, A. R., & Boer, D. P. (2001). Investigation of the crim-
inal and conditional release histories of Canadian federal offend-
ers as a function of psychopathy and age. Law and Human
Behavior, 25, 647-661.
Poythress, N. G., Edens, J. F., & Lilienfeld, S. O. (1998).
Criterion-related validity of the Psychopathic Personality
Inventory in a prison sample. Psychological Assessment, 10,
426-430.
Raskin, R., & Terry, H. (1988). A principle-components analysis of
the Narcissistic Personality Inventory and further evidence of its
construct validity. Journal of Personality and Social Psychology,
54, 890-902.
Ross, S. R., Lutz, C. J., & Bailley, S. E. (2004). Psychopathy and
the five-factor model in a noninstitutionalized sample: A
domain and facet level analysis. Journal of Psychopathology
and Behavioral Assessment, 26, 213-223.
Rutherford, M., Alterman, A. I., Cacciola, J. S., & McKay, J. R.
(1997). Validity of the psychopathy checklist–revised in male
methadone patients. Drug and Alcohol Dependence, 44, 143-149.
Salekin, R. T., Rogers, R., & Sewell, K. W. (1997). Construct
validity of psychopathy in a female offender sample: A multi-
trait multi-method evaluation. Journal of Abnormal
Psychology, 106, 576-585.
Saulsman, L. M., & Page, A. C. (2004). The five-factor model
and personality disorder empirical literature: A meta-analytic
review. Clinical Psychology Review, 23, 1055-1085.
Schmitt, W. A., & Newman, J. P. (1999). Are all psychopathic
individuals low-anxious? Journal of Abnormal Psychology,
108, 353-358.
Skeem, J. L., Johansson, P., Andershed, H., Kerr, M., & Eno
Louden, J. (2007). Two subtypes of psychopathic violent
offenders that parallel primary and secondary variants.
Journal of Abnormal Psychology, 116, 395-409.
Skeem, J. L., Miller, J. D., Mulvey, E., Tiemann, J., & Monahan,
J. (2005). Using a five factor lens to explore the relation
between personality traits and violence in psychiatric patients.
Journal of Consulting and Clinical Psychology, 73, 454-465.
462 Assessment
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
Skeem, J. L., & Mulvey, E. P. (2001). Psychopathy and commu-
nity violence among civil psychiatric patients: Results from
the MacArthur Violence Risk Assessment Study. Journal of
Consulting and Clinical Psychology, 69, 358-374.
Soderstrom, H., Nilsson, T., Sjodin, A. K., Carlstedt, A., &
Forsman, A. (2005). The childhood-onset neuropsychiatric
background to adulthood psychopathic traits and personality
disorders. Comprehensive Psychiatry, 46, 111-116.
Taylor, J., & Lang, A. (2005). Psychopathy and substance use
disorders. In C. J. Patrick (Ed.), Handbook of psychopathy
(pp. 495-511). New York: Guilford.
Verona, E., Patrick, C. J., & Joiner, T. T. (2001). Psychopathy,
antisocial personality, and suicide risk. Journal of Abnormal
Psychology, 104, 462-470.
Widiger, T. A. (2006). Psychopathy and DSM-IV psychopathology.
In C. J. Patrick (Ed.), Handbook of psychopathy (pp. 156-171).
New York: Guilford.
Widiger, T. A., & Coker, L. A. (2001). Assessing personality disor-
ders. In J. N. Butcher (Ed.), Clinical personality assessment:
Practical approaches (2nd ed., pp. 407-434). New York: Oxford
University Press.
Widiger, T., & Lynam, D. (1998). Psychopathy and the five-
factor model of personality. In T. Millon, E. Simonsen, M.
Birket-Smith, & R. D. Davis (Eds.), Psychopathy: Antisocial,
criminal, and violent behaviors (pp. 171-187). New York:
Guilford.
Wilson, D. L., Frick, P. J., & Clements, C. B. (1999).
Gender, somatization, and psychopathic traits in a col-
lege sample. Journal of Psychopathology and Behavioral
Assessment, 21, 221-235.
Zolondek, S., Lilienfeld, S. O., Patrick, C. J., & Fowler, K. A.
(2006). The interpersonal measure of psychopathy: Construct
and incremental validity in male prisoners. Assessment, 13,
470-482.
Miller et al. / Levenson Self-Report Psychopathy Scale 463
at UNIV OF GEORGIA LIBRARIES on November 20, 2008 http://asm.sagepub.comDownloaded from
A preview of this full-text is provided by SAGE Publications Inc.
Content available from Assessment
This content is subject to copyright.