Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Psychometric and biometric evaluation of psychopathic personality traits in 9 and 10 year old twins
(USC Thesis Other)
Psychometric and biometric evaluation of psychopathic personality traits in 9 and 10 year old twins
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
PSYCHOMETRIC AND BIOMETRIC EVALUATION OF PSYCHOPATHIC
PERSONALITY TRAITS IN 9 AND 10 YEAR OLD TWINS
by
Serena Bezdjian
_________________________________________________
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PSYCHOLOGY)
August 2008
Copyright 2008 Serena Bezdjian
ii
Acknowledgments
I would like to express my sincerest gratitude to my advisor Laura A. Baker
for all of her encouragement, guidance and support through the years. I would also
like to thank my guidance and dissertation committees for all of their invaluable
guidance, help, and support: Thank you Drs. Adrian Raine, David Schwartz,
Carol Prescott, Penelope Trickett, and John J. McArdle.
I especially want to thank all of my family and friends for all of their endless
encouragement, love and support. Thank you is not enough; I could not have done
this without your support!!
Thank you to everyone at the Southern California Twin Project-you all have
made my time at the lab a great and memorable experience! I not only take away
wonderful memories of you all, but also everlasting friendships!
Thank you to Catherine Tuvblad, Kristen Jacobson, and Dora Isabel Lozano.
I have learned a tremendous amount from all of you!
A very special thank you to all of the twins and their families who
participated in our study and made this thesis possible!!
iii
Table of Contents
Acknowledgments ii
List of Tables v
List of Figures vii
Abbreviations viii
Abstract x
Chapter One: Introduction and Background 1
Psychopathy in Adults 2
Psychopathic-like traits in children 5
Operationalizing and Measuring Psychopathic-like Traits in
Children 7
Psychopathic Traits and Externalizing Behavioral Problems 8
Conduct Disorder and Oppositional Defiant Disorder 8
Psychopathy and Aggression 11
Psychopathic Traits and Internalizing Behavioral Problems 13
Psychopathic Traits and Attention Deficit-
Hyperactivity Disorder and Conduct Disorder 14
Genetic Studies of Psychopathic Personality Traits 16
Aims of the Present Study 18
Chapter Two: Methods and Design 21
Study Participants 21
Inclusion Criteria 21
Sample Recruitment and Characteristics 21
Procedures 23
Zygosity Determination 23
Measures 24
Psychopathic Personality Traits 24
Aggression 26
Externalizing and Internalizing Behavioral Problems 26
Statistical Analyses 27
Structural Equation Modeling 29
Multivariate Models 31
Exploratory (EFA) and Confirmatory
Factor (CFA) Analyses 34
iv
Validation of Psychopathic Personality Trait
in children 35
Genetic and Environmental Etiology
of Psychopathic-like traits in children 37
Genetic and Environmental Covariation between
Psychopathic Traits and Externalizing Behavioral
Problems 37
Chapter Three: Results 39
Descriptive Statistics 39
Exploratory and Confirmatory Factor Analyses 41
Psychopathic Traits and Externalizing Behavior Problems
(Convergent and Discriminant Validity) 46
Genetic Analyses of Psychopathic Personality Traits 60
Genetic and Environmental Covariation between
Psychopathic Personality Traits
and Externalizing Behavioral Problems 65
Chapter Four: Discussion 72
Factor Structure of Psychopathic Traits 72
Psychopathic Personality Traits and Externalizing,
Internalizing Behavior Problems 75
Genetic and Environmental Influences on
Psychopathic Personality Traits 80
Genetic and Environmental Covariation between
Psychopathic Traits and Externalizing
Behavioral Problems 82
Strengths and Limitations 84
Assumptions in Twin Studies 86
Future Directions 87
Conclusions 88
References 90
Appendix: Supplementary Analyses, Tables, and Figures 96
v
List of Tables
Table 1 Means and SDs for CPS-RE scales by sex 40
Table 2 Factor Structure and EFA and CFA loadings 42
Table 3 Model Fits for Confirmatory Factor Analyses 45
Table 4 Means and Standard Deviations for the Two
Psychopathic Personality factors by Sex 46
Table 5 Phenotypic Correlations between Psychopathic Traits
and Externalizing Behavior Problems 49
Table 6 Phenotypic Correlations between Psychopathic Traits
and DISC-IV symptoms of generalized anxiety (GA) 51
Table 7 Caregiver item endorsements for psychopathic traits
CD, ADHD, ODD for three example cases 54
Table 8 Means and SDs for CD, ODD, ADHD symptoms counts
And Reactive and Proactive aggression by sex 56
Table 9 Mean scores and symptoms counts of Reactive and
Proactive Aggression, CD, ADHD, ODD for high and
low psychopathy groups 57
Table 10 Twin correlations for the CPS-RE Factors 60
Table 11 Model fit Comparisons for the CPS-RE factors 62
Table 12 Genetic and environmental correlations for CPS-RE
Factors 1 & 2 for boys and girls 65
Table 13 Intra-class twin correlations and cross-twin cross-trait
Correlations by zygosity (MZ, DZ) and sex (male, female) 69
Table 14 Parameter Estimates from Best-Fitting Models-
Genetic and Environmental Correlations Between
Psychopathic Traits and other behavioral problems 71
vi
Table A1 ACE Estimates For CPS-RE Factors 96
Table A2 Item endorsements from CPS-RE caregiver
Reports for each factor by sex 97
Table A3 Correlations for Psychopathic Traits and
Other behavioral problems by sex 99
Table A4 Multiple Regression Analyses of Psychopathic
Traits with Internalizing and Externalizing
Behavioral Problems (males, females) 100
Table A5 ODD Symptoms (DISC-IV) 104
Table A6 CD Symptoms (DISC-IV) 105
Table A7 ADHD Symptoms (DISC-IV) 106
Table A8 GAD Symptoms (DISC-IV) 107
Table A9 Best fitting models from Bivariate
Cholesky Analyses of Covariation among
Psychopathic Traits and CD, ADHD, ODD, and
Reactive and Proactive Aggression 108
vii
List of Figures
Figure 1 Univariate AC/DE path diagram in twins 30
Figure 2 Path Diagram of an ACE Cholesky Decomposition
For one twin 32
Figure 3 Independent Pathway model in one twin 32
Figure 4 Common Pathway Model in one twin 33
Figure 5 Illustration of the Factor Structure of the CPS-RE 43
Figure 6 Standardized Symptom Counts and Means of CD, ADHD
ODD, GA, and Reactive and proactive aggression for low
and high psychopathic trait groups 59
Figure 7 Best-fitting ACE estimates of
Psychopathic Personality Traits 64
Figure A1 Illustration of Genetic and Environmental
Correlations 101
Figure A2 Illustration of alternative factors models
of psychopathic traits 102
viii
Abbreviations
A Additive genetic effect
ADHD Attention-deficit hyperactivity disorder
AIC Akaike Information Criterion
BIC Bayesian Information Criterion
C Shared environmental effect
CD Conduct Disorder
CFI Comparative fit index
CI Confidence Interval
CPS-RE Child Psychopathy Scale-Revised Extended
df Degrees of freedom
DISC-IV Diagnostic Interview Schedule for Children-Version IV
DZ Dizygotic
E Unique environmental effect
Factor 1 Callous/Disinhibited factor
Factor 2 Charming/Manipulative factor
MZ Monozygotic
ODD Oppositional Defiant Disorder
OS Opposite sex twins (opp. sex)
r
g
Genetic correlation
r
c
Shared environmental correlation
ix
r
e
Unique environmental correlation
RMSEA Root mean squared error approximation
SD Standard deviation
η
2
Eta squared
β Standardized Beta coefficient
χ
2
Chi-square
-2LL -2(log-likelihood)
x
Abstract
Psychopathic personality traits and their genetic and environmental etiology
were investigated in 9-10 year old twins. Specifically, the psychometric properties
of psychopathic personality traits were examined by exploring the factor structure of
these traits, investigating their relationship to other problem behaviors that occur in
childhood, and investigating the extent of the genetic and environmental influences
on these traits. A community sample of 1219 children (born between 1990 and
1995) from the Southern California Twin Project was employed. Psychopathic
personality traits were assessed through caregiver reports of the twins psychopathic
behaviors using the Child Psychopathy Scale-Revised Extended (CPS-RE;
Lynam, 1997). Confirmatory factor analyses revealed an optimal two-factor solution
(Callous/Disinhibited and Charming/Manipulative) to the CPS-RE subscales.
Multivariate genetic modeling revealed significant genetic as well as unique
environmental influences on psychopathic personality traits in both boys and girls.
The heritability estimates for the two factors were 0.64 and 0.46, respectively in
boys; and 0.49 and 0.58, respectively in girls. Unique environmental effects were
0.36 and 0.53, for the two factors in boys; and 0.44 and 0.37, respectively in girls.
No shared environmental influences were found. Moderate to strong phenotypic
relationships between psychopathic personality traits with other measures of problem
behaviors also emerged. Specifically, psychopathic traits were significantly
correlated with symptoms of conduct disorder (CD), oppositional defiant disorder
xi
(ODD), attention deficit hyperactivity disorder (ADHD), and reactive and proactive
aggression. Bivariate genetic analyses examining the genetic and environmental
covariation between psychopathic traits and other behavioral problems also revealed
significant genetic and environmental relationships between psychopathy and CD,
ODD, ADHD, and aggression. In summary, significant genetic and unique
environmental influences on psychopathic traits were found in 9 and 10 year old twin
boys and girls. Children who displayed higher mean psychopathy scores exhibited
higher levels of symptoms for CD, ODD, aggression, and ADHD. Findings provide
further evidence that psychopathic traits may develop as early as age 9-10, and that
these traits are strongly related to other early behavioral problems, which may stem
from common genetic predispositions. This knowledge may aid in the future
prediction and therefore amelioration of aggressive and criminal behaviors in
adolescents and adults.
1
Chapter One: Introduction and Background
The construct of psychopathy has been around for well over a century. It
was once used to represent all mental irregularities or a lack of adherence to
morals and norms, however; within the past several decades, it has adopted a
more refined definition (Lynam & Gudonis, 2006). Hervey Cleckley offered
one of the most profound and insightful (1941) descriptions of psychopathy.
In his work, Cleckley (1941) described 15 males and females that he thought
were psychopaths, meaning these individuals seemed or appeared sane,
intelligent and competent, but were visibly (noticeably) disturbed. Cleckley
said these individuals wore masks of sanity (Lynam & Gudonis, 2006).
Since Cleckleys work, other clinicians and researchers have also offered their
descriptions of psychopathy and psychopaths (Buss 1996, Hare 2003). At
present, psychopathy is described as a form of personality disorder.
Behaviorally, the psychopath is an impulsive risk-taker involved in a variety of
criminal activities. Interpersonally, the psychopath has been described as
grandiose, manipulative, glib, and callous. Affectively, the psychopath
displays shallow affect and emotion, lacks empathy, has no compunction for
wrongdoings, and is unable to maintain close relationships
(Lynam & Gudonis, 2006; Blair, 2001).
The psychopathic offender is versatile, prolific, manipulative, impulsive
and the most violent of offenders (Hare, 2003). Psychopaths show little or no
2
guilt or repentance for their actions and acts of antisocial behavior (Gacono,
2000; Hare, 1991), and they are the most violent of criminal offenders
(Kosson et al., 1990). Psychopathic individuals are also involved in other
forms of risky behaviors including, higher or more elevated rates of drug and
alcohol use, abuse and dependence (Tourian et al., 1997;
Lynam & Gudonis, 2006). In addition, psychopaths are mostly male and make
up approximately 1-2% of the greater population (Hare, 1991) however; they
are involved in over 50% of the crimes committed (Hare, 1993, pg. 87).
Thus, it is no surprise that the construct has received intensive
investigation. Most of the research on psychopathy has been concerned with
documenting the reliability and utility of the construct of psychopathy, or
identifying the underlying structure or mechanisms that embody the
psychopath (Cruise et al., 2000; Lynam & Gudonis, 2006). It is crucial to
focus on the development and genetic and environmental etiology of
psychopathic personality traits.
Psychopathy in Adults
There are currently several different instruments used to measure and
assess the construct of psychopathy in adults including the Self-Report
Psychopathy-II scale (SRP-II; Hare, 1991), the Psychopathy Checklist-
Screening Version (PCL:SV; Hart, Cox & Hare, 1995), the Minnesota
Multiphasic Personality Inventory (MMPI); the Psychopathic Personality
Inventory (PPI; Lilienfeld & Andrews, 1996), and the Self-Report Psychopathy
3
Scale (SPPS; Levenson, Kiehl, & Fitzpatrick, 1995). These measures vary in
design and structure, particularly in terms of reporters. For instance,
assessments are based on self-reports of psychopathy and some also employ
collateral information gathered from criminal records as well as from family
members and/or teachers. Most of assessments of psychopathy are based on
Cleckleys (1976) description of the personality phenomenon, which attempted
to dissociate the disorder from common criminality.
At present, however, the instrument of choice among researchers and
clinicians assessing psychopathy in adults is the Hare Psychopathy Checklist-
Revised (PCL-R; Hare, 1980, 1991, 2003). The PCL-R has been the standard
for operationalizing the construct of psychopathy. It is widely used by
psychiatrists, psychologists, researchers, and criminal justice personnel (Hare,
1996). The PCL-R is a clinical rating scale comprised of 20 items based on
Cleckleys criteria (1976) that describes psychopathy as a constellation of
interpersonal, affective and behavioral traits. Reliability and construct validity
for the PCL have been well established (Harpur, Hakstian, & Hare, 1988; Hare
et al., 1990; Ogloff, Wong, & Greenwood, 1990; Kosson, Smith, & Newman,
1990).
There is ongoing debate as to whether psychopathic traits are under laid
by two, three, or four factors (see Cooke & Michie, 2001; Vitacco, Rogers,
Neumann, Harrison, & Vincent, 2005), but at least two broad correlated factors
have been identified in the adult psychopathy literature using the Hare
4
Psychopathy Checklist (PCL-R; Hare, 1991, 2003). The two factors are
comprised of affective-interpersonal features (e.g. glibness, superficial
charm, lack of empathy; Factor 1), and socially deviant lifestyles and
behaviors (e.g. impulsivity, sensation seeking, irresponsibility, poor
behavioral control; Factor 2) (Harpur, Hakstian, Hare, 1988; Hare, 1991, 2003;
Lynam, 1997). Recently Hare has also proposed a two-factor-four facet
model of psychopathic traits that is comprised of interpersonal, affective,
lifestyle and antisocial facets (Hare, 2003).
Hares Factor 1 is positively correlated with narcissism and histrionic
personality disorder, along with increased risk for antisocial behavior and
decreased scores on measures of dependence and anxiety, while Factor 2 is
highly correlated with substance abuse and general antisocial behaviors, and
poor academic performance (Kotler & McMahon, 2005).
Offenders with high scores on both factors of the PCL-R display violent
and aggressive behaviors disproportionately more often than other adult
criminals (Hare, 1991). Criminal offenders scoring high on measures of
psychopathy usually have extensive and versatile illicit histories (Hare, Forth,
& Strachan, 1992), and are more likely than low-scoring offenders to commit
crimes upon release or discharge (Serin, 1991, 1996). Crimes committed by
psychopaths are different (in nature) from those committed by non-
psychopaths: they are more goal-oriented and are often more violent and often
involve non-relatives (Cornell, 1996; Serin, 1996).
5
In an effort to refine the construct of psychopathy even further, Cooke
and Michie (2001) conducted exploratory and confirmatory factor analyses
using the PCL-R on a sample of over 2,000 participants to determine whether
the two-factor model of psychopathy prevailed over their suggested three-
factor model. Cooke and Michie developed a 3-factor hierarchical model in
which a sound super-ordinate factor, Psychopathy, is under laid by three
factors: Arrogant and Deceitful Interpersonal Style, Deficient Affective
Experience, and Impulsive and Irresponsible Behavioral Style. Their three-
factor results were validated on several diverse samples including, a North
American sample and on Scottish PCL-R data (Cooke & Michie, 2001;
Hare, 2003).
Recent attempts to delineate the factor structure of psychopathy have
also replicated the three-factor structure of psychopathy that includes the
grandiose-manipulative, the callous-unemotional, and the impulsive-
irresponsible dimensions of the construct (Andershed et al., 2002;
Larsson et al., 2006).
Psychopathic-like traits in Children
The notion of identifying and treating psychopathic personality traits in
children and adolescence is hardly new or recent (Vincent & Hart, 2002; McCord &
McCord, 1956). Presumably, the traits of a personality disorder do not have a
sudden onset at the moment an individual becomes an adult. Some evidence for the
existence of psychopathy in adolescence can be seen in the adult literature
6
(Vincent & Hart, 2002). Moreover, as early as the 1940s Cleckley acknowledged
that the disorder might have its roots in childhood in The Mask of Sanity
(Salekin & Frick, 2005).
Accordingly, several researchers have argued that a select group of youth
with serious conduct problems will mature into adults with psychopathic personality
disorder (Frick, 1998; Lynam, 1996, 1997, 1998). Several factors have led to the
increasing interest in the topic of child and adolescent psychopathy. First, the strong
associations between psychopathy factors and antisocial behavior and violence in the
adult literature have prompted interest in whether conduct problems, aggression, and
violence in some youth might be explained by similar personality correlates
(Moffitt, Caspi et al., 1996). Second, Lynam (1996, 1997) has proposed that to
prevent serious negative outcomes associated with psychopathy, the early
identification of psychopathic traits is crucial because attempts to alleviate and treat
psychopathy in adulthood have been quite unsuccessful (Hart et al., 1988) and
evidence suggests that psychopathic individuals have antisocial and criminal
histories that begin prior to adulthood (Hart & Hare, 1997). Lynams proposal is
based on the assertion that psychopathic traits in childhood and adolescence are
related to psychopathy in adulthood. This view is shared by several other
researchers including, Blair (2006). Thus, interest in the concept of child
psychopathy has increased considerably within the last decade
(Kotler & McMahon, 2006).
7
Furthermore, Lynams research has suggested that children who manifested
symptoms of hyperactivity-impulsivity-attention problems and conduct problems
were afflicted with a virulent strain of Conduct Disorder best described as fledgling
psychopathy (Lynam, 1996; p. 209). In support of this theory, children who were
diagnosed with both conduct disorder (CD) and attention deficit-hyperactivity
disorder (ADHD) demonstrated that they embodied more severe and aggressive
patterns of antisocial behavior and neurological deficits that were similar to deficits
exhibited by adults with psychopathy (Lynam, 1997).
However, there is a concern regarding the examination of this construct in
children. It might not be developmentally appropriate to diagnose this syndrome in
children and adolescents. Some developmental researchers have argued that the
disorder is inapplicable for youth (Kotler & McMahon, 2005; Vincent & Hart, 2002).
However, considerable evidence also suggests that the roots of this disorder may
begin in early developmental periods. Furthermore, researchers have made the
compelling argument that the traits of a personality disorder presumably do not have
a sudden onset at the moment an individual becomes an adult at 18 years of age
(Vincent & Hart, 2002; Frick & Hare, 2001; Salekin et al., 2004; Forth et al., 2003).
Operationalizing and Measuring Psychopathic Traits in Children
As interest in examining psychopathic-like traits in children grew, it became
apparent that the measurement tools used to assess psychopathy in adults were not
appropriate to use with children (Edens, Skeem, Cruise, Cauffman, 2001). For
instance, a number of items commonly used in the Hare Psychopathy Checklist were
8
unsuitable for children (e.g. many short-term marital relationships). Some of the
items measured behaviors that might very well be applicable to children and
adolescents (e.g. impulsivity, failure to accept responsibility, etc.) however; they
might not accurately distinguish youths with psychopathic traits from their peers
(Kotler & McMahon, 2005). With these issues in mind, a number of researchers
have made an effort to accurately assess the construct of child psychopathy by
creating new measurement tools, or adapting existing adult assessment tools. Thus,
most of the instruments that measure psychopathy in youth were derived from the
PCL-R (Hare, 1991a, 2003). Some of these comprehensive and self-report measures
include, the Antisocial Process Screening Device (APSD; Frick & Hare, 2001), the
Child Psychopathy Scale, (CPS; Lynam, 1997), the Youth Psychopathic Traits
Inventory (YPI; Andershed, Kerr et al., 2002), and the Psychopathy Checklist: Youth
Version (PCL:YV, Forth, Hart & Hare, 1990).
Psychopathic Personality Traits and Externalizing Behavioral Problems
Psychopathic Traits, Conduct Disorder (CD), and Oppositional Defiant Disorder
(ODD)
The relationship between behavior problems and measures of child and
adolescent psychopathy has been examined in several different samples including,
juvenile offender populations, clinic-referred samples, and in community samples
9
(Lynam, 1997; Frick et al., 1994; Forth, 1995; Kosson, et al., 2002). Even with a
wide range of measures and methodology, findings have been consistent
(Frick, 1998). For instance, findings have indicated that the interpersonal/affective
factor of psychopathy (also referred to as the callous/unemotional factor) is strongly
related to severe and persistent conduct problems (Kotler and McMahon, 2005). The
associations between the youth psychopathy construct and conduct problems may be
very significant because they may point to a unique etiological pathway for the
development of severe problem behavior (Lynam, 1996, 1997;
Frick and Ellis, 1999).
A study on a sample of 95 clinic-referred 6-13 year old children using parent
and teacher ratings of the Antisocial Process Screening Device (APSD) found that
the impulse control/conduct problem (I/CP) factor of the APSD was strongly related
to DSM-III-R diagnoses of Oppositional Defiant Disorder (ODD) and Conduct
Disorder (CD) and was also significantly correlated with other conduct problem
rating scales including the aggression and delinquent behavior subscales from
Achenbachs Child Behavior Checklist (CBCL) (Barry et al., 2000;
Frick et al., 2003). However, the callous/unemotional (C/U) factor was only
moderately correlated with CD diagnoses. According to Frick and his colleagues,
their findings indicated that the I/CP and C/U factors were measuring different but
correlated constructs. Furthermore, children with elevated scores on the C/U factor
and a diagnosis of ODD or CD had a family history of higher rates of parental arrest
(Barry et al., 2000; Frick et al., 2003).
10
More recently, interest in the question of whether similar associations
between psychopathy and behavior problems could be identified in a community
sample has developed. Frick and his colleagues examined a sample of 98 third,
fourth, sixth and seventh graders from the community (Frick et al., 2000). APSD
parent and teacher report data were collected and conduct problems were assessed
through parent, teacher, and self-report measures. After a one-year follow-up,
children with conduct problems and high scores on the C/U factor (at baseline) had
more conduct problems and were more likely to have CD compared to children with
conduct problems alone. These children were also reportedly more aggressive and
had more self-reported delinquency during follow-up assessments (Frick et al., 2000;
Kolter & McMahon, 2005).
The studies conducted by Frick and his colleagues made several important
contributions. First, the psychopathy factors were found to have unique relations
with outcome measures. Second, results suggested that the C/U traits seemed to
influence some of the associations between conduct problems and individual
characteristics usually found in the conduct problem literature. Third, it extended
finding from a clinic-referred sample to a community sample, and it also provided
evidence that the presence of C/U traits alone might be enough to increase risk for
later conduct problems. Finally, the combination of serious conduct problems
(diagnoses of ODD and CD) and C/U traits was related to a family history of
antisocial behavior (Kotler & McMahon, 2005).
11
Empirical evidence has also established that at the adult level, the
relationship between psychopathy and antisocial personality disorder (APD) is
asymmetric (Forth & Burke, 1998; Forth & Mailloux, 2000). Specifically, 90% of
adult psychopathic offenders meet APD criteria, but only 25% of those diagnosed
with APD are psychopathic (Hare, 1991, 2003). A similar relation has emerged
when comparing CD and psychopathy as measured by Psychopathy Checklist:Youth
Version (PCL:YV; Forth, Kosson, & Hare, 2003). Forth and Burke (1998) reported
that nearly 97-100% of adolescent offenders qualify for a CD diagnosis whereas less
than 30% of the young offenders with CD met diagnostic criteria for psychopathy on
the PCL:YV (Salekin, 2006).
Psychopathy and Aggression
A strong link between psychopathic personality traits and aggressive
behaviors has been established in adults, children, adolescents, and psychiatric
patients (Porter & Woodworth, 2006; Flight & Forth, 2007). A large number of
studies have illustrated that the existence of psychopathic traits (or psychopathic like
traits) is associated with a propensity for violent and aggressive behavior (Hare &
Jutai, 1983; Harpur & Hare, 1994). Adult psychopathic offenders have been charged
with violent crimes twice as often as non-psychopathic offenders
(Hare & Jutai, 1983). Within a large number of federal offenders, researchers found
that psychopaths had been convicted on average one in a half times more often on
violent crimes than non-psychopathic offenders (Porter, Birt, & Boer, 2001).
12
Even though most of the research on psychopathy is conducted on adults,
there is growing evidence that suggests that psychopathic traits are related to
aggression much earlier in life. Precursors to psychopathy have been said to emerge
in early childhood in the form of callous unemotional traits (Lynam, 2002; 1997),
which closely resemble adult psychopathic like traits (particularly PCL-R Factor 1;
Hare 1991). These characteristics or traits are usually associated with aggressive and
violent behavior (Frick, 1998; Dodge, 1991; Lynam 2002; 1997). Furthermore,
numerous research findings have indicated that the interpersonal/affective factor of
psychopathy (also referred to as the callous/unemotional factor) is strongly related to
severe and persistent conduct problems (Kotler and McMahon, 2005). The
associations between the youth psychopathy construct and conduct problems may be
very significant because they may point to a unique etiological pathway for the
development of severe problem behavior (Lynam, 1996, 1997;
Frick and Ellis, 1999).
Proactive aggression has been characterized as organized, goal-
directed, instrumental and cold-blooded, with little evidence of automatic
arousal (Dodge, 1991), while reactive aggression has been defined as being an
unplanned, impulsive type of aggression (Flight & Forth, 2007). Psychopaths
who are characterized as parasitic, manipulative, stimulation seeking,
emotionally blunted and automatically under-aroused (Hare et al., 1999) have
higher rates of violent crimes than do non-psychopaths
13
(Hare & McPherson, 1984). Criminals identified as committing proactive,
violent offenses have higher scores on the Psychopathy Checklist than those
with a history or background of reactive aggression or violence
(Cornell et al., 1996). Some have argued that true psychopathy is more
associated with proactive rather than reactive aggression (Patrick, 2001). Frick
and colleagues (2003) have found that psychopathic-like children with callous-
unemotional traits have higher proactive aggression scores (Raine et al., 2006).
Thus far no study has examined the underlying genetic and environmental
etiologic covariation between psychopathic personality traits with reactive and
proactive aggression in children. Based on previous literature, a stronger link
between psychopathic traits and proactive aggression would be expected.
However, since psychopaths are also impulsive it would also be reasonable to
expect a relationship or link between psychopathic traits and reactive
aggression. Specifically, the behavioral dyscontrol and impulsive features of
psychopathic traits would be expected to correlate strongly with reactive
aggression.
Psychopathic Traits and Internalizing Behavior Problems
There is a well-established relationship between anxiety and antisocial
behavior in children (Russo, Beidel, 1993), adolescents (Krueger et al., 1994), and
adults (Robins, Tipp, & Pryzbeck, 1991). However, the relationship between
anxiety and antisocial behaviors has been somewhat contentious. Some theories
14
have centered on a lack of inhibition (Cloninger, 1987; Lykken, 1982; 1995), while
other theories have focused more on an absence or deficiency in anxiety for
psychopaths or antisocial individuals (Cleckley 1941; 1976; 1982). When the two
psychopathic dimensions as described by Hare are separated in children
(Frick, 1998; Lynam, 1997), and adults (Hare, Hart, & Harpur, 1991), they show
evidence of different relationships to anxiety. For instance, the first dimension
(impulsive and antisocial lifestyle) is positively correlated with anxiety, while the
second dimension (the callous/unemotional interpersonal style) is negatively
associated with anxiety (Harpur et al., 1989). However, many of the instruments that
tap into psychopathic traits include a combination of impulsive and callous and
unemotional items (within the two dimensions), it is often difficult to discern the
relationship between psychopathy and trait anxiety (Lilienfeld, 1994). Evidence also
suggests that the interpersonal/callous traits of psychopathy should be inversely
related to depression (Blonigen et al., 2005). A weak relationship has been
previously found between psychopathic traits and internalizing behaviors in young
children (Lynam, 1997). Thus, psychopathic traits are expected to exhibit a weak
relationship with internalizing behaviors.
Psychopathic Traits and Attention Deficit Hyperactivity Disorder (ADHD) and
Conduct Disorder (CD)
Conduct problems in children have consistently demonstrated that these
problems co-occur with a large number of other disorders and behavioral adjustment
15
problems (Frick, 1998). One of the most common overlapping disorders in children
with conduct problems is attention deficit hyperactivity disorder (ADHD)
(Waschbush, 2002). Children with both ADHD and conduct problems exhibit a
more severe pattern of aggressive and antisocial behavior, have poorer outcomes,
have more contact with the police (OBrien & Frick, 1996), higher rates of
delinquency and theft (Salekin, et al., 2003; Loeber et al., 1990; Robbins, 1966), and
more arrests and convictions in adulthood (Lynam, Caspi, Moffitt, Loeber, &
Stouthamer-Loeber, 2007; Robbins, 1966).
Theories have suggested that the combination of ADHD with conduct
problems is a distinct disorder and qualitatively different from either disorder alone
(Lynam, 1996, 1997, 1998). Studies have demonstrated that children with attention
deficit hyperactivity disorder (ADHD) and conduct disorder (CD) had particularly
higher rates of antisocial behavior making them appear to be similar to adults with
psychopathy (Lynam, 1996, 1997, 1998). Specifically, Lynam examined a
community sample of 508 12 to 13-year old boys and identified four groups within
his sample: (1) the first group was high on both ADHD and CD symptoms; (2) high
on ADHD symptoms only; (3) high on CD symptoms only; and (4) the control group
with no significant ADHD or CD symptoms. His results demonstrated that the group
high on both ADHD and CD symptoms also scored the highest on psychopathic
personality traits (Lynam, 1997; 1998). He concluded that the ADHD plus CD
group most closely resemble the psychopathic adult (Lynam, 1998 pg. 571). This
model could help explain why children with childhood conduct problems are more
16
likely to exhibit the cold and callous features associated with psychopathy.
However, Frick and his colleagues demonstrated that children with a comorbid
diagnosis of CD and ADHD do not display the hallmark traits of psychopathy and
that this comorbidity may not be the best marker for childhood precursors to
psychopathy (Frick, Bodin and Barry, 2000 pg. 391).
Genetic Studies of Psychopathic Personality Traits
There has been very little research on the etiology of psychopathic traits
especially in youth. To date, there has been only one study that has explored the
heritability of psychopathic traits in young children. Viding et al. (2005) explored
the heritability of extreme callous/unemotional traits in seven-year-old twin boys.
They found a 67% heritability estimate for callous/unemotional traits in children,
although they did not explore any sex differences of these psychopathic traits
(Viding et al., 2005).
There have been three additional studies that have investigated the etiology of
psychopathic personality traits, however they have been conducted on adolescents
and young adults. The first study examined psychopathic personality traits using the
Multidimensional Personality Questionnaire (MPQ) in a longitudinal study of 17 and
24-year-old same-sex twins (Blonigen, Hicks, Krueger, Patrick and Iacono 2006).
Their results indicated that there were significant genetic as well as unique
environmental influences in the fearless dominance (FD) and impulsive antisociality
(IA) factors in their total combined sample with equal effects in males and females.
17
Approximately half of the variance of both FD and IA was due to additive genetic
effects, while the other half was due to unique environmental effects; no significant
shared environmental influences were found (Blonigen et al., 2006). The second
study examined the etiology of psychopathic personality in 16-17 year old
adolescents using the Youth Psychopathic Traits Inventory (YPI;
Andershed et al., 2002) (Larsson et al., 2006). The study found that 33-53% of the
variance in psychopathic personality was due to genetic effects, while no significant
common environmental influences in psychopathic personality traits were found
(Larsson et al, 2006). Additionally, they did not find any quantitative sex differences
in the three psychopathic personality dimensions in their adolescent to young adult
sample (Larsson et al., 2006). Finally, Taylor et al., (2003) examined psychopathic
like traits in adolescent twin boys 16-18 years old using the Minnesota Temperament
Inventory (MTI). Their study also found significant additive genetic and unique
environmental influences on psychopathic-like traits, but no shared environmental
influences. There is still paucity in the literature, however; and a need to gain a
better understanding of the etiology of psychopathic traits particularly, in children.
Since the present study utilizes a sample of both males and females, it also aims to
bridge the gap in the literature and contribute to a better understanding of sex
differences in the etiology of psychopathic personality traits.
Children who show antisocial behavior from early childhood are at great risk
for showing antisocial and criminal behavior in adulthood. Despite an impressive
start to the scientific study of child and adolescent psychopathy, empirical work on
18
this topic has been lacking. As such, many questions still remain regarding the
etiology of psychopathy in children and adolescence and in particular, questions
regarding the genetic and environmental etiology of any sex differences in the
construct of child psychopathy. The proposed study aims to investigate the genetic
and environmental etiology of psychopathy in children and examine any sex
differences using a community sample of male and female twins. Additionally since
the current study employs a community-based sample of preadolescent twin children,
it may help provide valuable information regarding the genetic and environmental
etiology of psychopathic traits at an early age, which cannot be garnered from
previous studies that have employed clinical or incarcerated adult male samples
(Frick et al., 2000; Taylor et al., 2003).
The motivation and purpose for examining psychopathic personality traits in
children stems from the lack of clear psychometric understanding of the construct in
children due to a lack of a clear factor structure and a lack of understanding the
genetic and environmental etiology of psychopathic personality traits. Additionally,
behavior genetic designs help inform us about the structure and nosology of
behavioral constructs including, psychopathy. Twin designs provide the ability to
get to the heart of what the constructs are and aid in the thorough understanding of
psychopathic traits in childhood.
Aims of the Present Study
The overall aim of this study was to investigate psychopathic personality
traits in children. Specifically, the psychometric properties of psychopathic
19
personality traits were examined by exploring the factor structure of these traits,
investigating their relationship to externalizing behavior problems in childhood, and
investigating the extent of the genetic and environmental influences on these traits.
More specifically the present study aimed to:
(1) Identify the factor structure of psychopathic-like traits in children using the
Child Psychopathy Scale-Revised Extended (CPS-RE; Lynam, 1997, 2002)
during wave 1 of the Southern California Twin Study (when the twins are 9
and 10 years old). Based on the literature, if at least two underlying factors
have been found in the adult psychopathy literature, then a similar pattern
should emerge in children, since psychopathic personality traits have been
shown to be stable.
(2) Investigate the convergent and discriminant validity of the emerging
psychopathic personality dimensions and total psychopathy by examining
their relationship to other validated measures of problem behavior including
aggression, conduct disorder and other externalizing as well as internalizing
behavioral problems. Based on previous findings in the adult psychopathy
literature, psychopathic traits should be strongly associated with externalizing
behavioral problems, and should be distinct from and have weaker
relationships with internalizing behaviors including anxiety.
(3) Examine the extent of genetic and environmental influences on psychopathic
traits in both boys and girls. Both the adult and child literature demonstrate a
significant genetic influence on psychopathic traits; with little to no evidence
20
of shared environmental influences on psychopathic personality traits
(Viding, 2005; Blonigen et al, 2003, 2005, 2006). Therefore, a significant
genetic influence on psychopathic traits is expected in children as young as 9
and 10 years old.
(4) Examine the genetic and environmental covariation between psychopathic
traits with other externalizing behavioral problems in children including
aggression, conduct disorder, oppositional defiant disorder, and attention-
deficit-hyperactivity disorder. In adults, psychopathic traits display a specific
pattern of relation with other forms of psychopathology. In adults,
psychopathy is negatively or uncorrelated with internalizing disorders
including anxiety and is positively correlated with antisocial personality.
Evidence has also shown that in young adults there is a significant genetic
and unique environmental overlap between psychopathic traits and antisocial
behaviors (Larsson et al., 2006). Thus, the current study hypothesizes that a
similar pattern will emerge in children.
21
Chapter Two: Methods and Design
Study Participants
The current study is based on data from the Southern California Twin Study,
an ongoing longitudinal twin study assessing risk factors for aggressive and
antisocial behaviors in children and adolescents. Additionally, the Twin Study aims
to examine the relationship among genetic, environmental, social, and biological
factors on the development of antisocial behavior from childhood to adolescence to
young adulthood through a multi-rater multi-measure approach. The sample was
derived from the Los Angeles community and includes twins, triplets and their
caregivers. The initial sample includes 1,219 twins and triplets born between 1990
and 1995. The first wave of data assessment occurred from January 2001 to June of
2004.
Inclusion Criteria
Study participation required that the twins be proficient in English and 9 or
10 years old at 1
st
assessment (see Baker, Barton, Lozano, Raine & Fowler, 2006). In
addition, either English or Spanish proficiency was required for the twins primary
caregiver. A detailed description of the inclusion criteria and process may be found in
Baker et al., 2007.
Sample Recruitment and Characteristics
The participants were recruited from the Los Angeles community, and
consisted of 9 and 10 year old twin pairs and their primary caregivers (N=605 sets of
twins and triplets). They were ascertained primarily through local schools, both public
22
and private, as well as advertisements in local newspapers (Baker et al., 2002). The
first wave of data collection was completed in June of 2004. The ethnic composition
of the twin pairs is as follows: 44% Hispanic, 25% Caucasian, 16% African
American, 3% Asian, and 12% mixed or other ethnicities. The sample is comparable
to that of the greater Los Angeles area both ethnically and socio-economically.
Families identified as having twins were sent letters describing the study in
extensive detail. Each letter included a prepaid envelope that contains an
information form for interested families to provide contact information and return to
the twin study staff. Basic demographic information was also obtained. Interested
families were contacted and those agreeing to participate were scheduled for a full
days visit to our laboratory, for assessments of the twins and their primary
caregiver. The twins participated in an interview process as well cognitive and
neuropsychological testing and psychophysiological assessment. The caregivers
were asked to participate in an extensive interview process including questions about
parental antisocial behavior, substance use, as well as detailed information about the
twins antisocial behaviors. The twins teachers were also contacted via mail with
parental permission. They were asked to fill out questionnaires about the behaviors
and conduct problems the twins display in school and specifically in the classroom.
During the first wave of data collection, the twins and their primary
caregivers participated in a 6-8 hour laboratory assessment at USC. The mean ages
were 9.60 years (SD=0.60) for the total sample of children and 40.14 years (SD=6.61)
for their caregivers. Caregiver participation was primarily the biological mothers
23
(> 90 %). Among the 1219 child participants, there was approximately equal gender
distribution with 48.7% boys (n=594) and 51.3% girls (n=625), while the caregivers
were primarily female (94.2%).
Procedures
Participants and their caregivers (generally the biological mother) were
invited to the Twin Project laboratory at USC for a full day of assessments
(approximately 7 hours). The caregivers were administered self-report
questionnaires and interviewed about their twins behavior at home and in school.
The current study is part of a larger longitudinal twin study on risk factors for
aggressive and antisocial behaviors (see Baker, Barton & Raine, 2002).
Testers and examiners consisted of full time staff members as well as USC
graduate and undergraduate students. All examiners were extensively trained on all
testing procedures. Training included inter-examiner reliability checks, videotaped
monitoring to ensure strict adherence to standardized testing protocols, and
supervised training sessions for all aspects of testing.
Zygosity Determination
In order to extract DNA for zygosity testing, the participating families were
asked to provide cheek swap samples. The twins zygosity was predominantly
determined through DNA microsatellite analyses. Questionnaires designed to
determine zygosity were also employed to help establish the zygosity of ambiguous
pairs of twins. When both assessment methods were available, there was a 90%
agreement between the DNA results and the questionnaire (Baker et al., 2007). The
24
following is a breakdown of the zygosity groups (by sex) used in the present study:
MZ males (N=138 pairs), MZ females (N= 139 pairs), DZ males (N=84 pairs), DZ
females (N=97 pairs), and DZ opposite sex pairs (N=147 pairs).
Measures
The main focus of the current study is to examine psychopathic-like
personality traits in children. Below is a description of how psychopathic personality
traits are measured along with other various measures of externalizing behaviors
used in this study to understand the relationship between psychopathic personality
traits and other problem behaviors are presented.
Psychopathic Personality Traits
Psychopathic personality traits were measured using a slightly extended
version of the Child Psychopathy Scale - Revised (CPS-RE; Lynam, 1997; 2001;
2002). The CPS-RE is a well-validated instrument measuring psychopathic
personality traits in children and adolescence (see Lynam, 1997). The CPS-RE is
comprised of 14 subscales (based on 58 yes or no items-scored 1=No; 2=Yes), which
include assessments of glibness, untruthfulness, lack of guilt, callousness,
impulsiveness, boredom susceptibility, manipulation, poverty of affect, parasitic
lifestyle, behavioral dyscontrol, lack of planning, unreliability, failure to accept
responsibility, and grandiosity. The CPS-RE was administered to the twins and their
caregivers in interview form. Internal reliabilities or consistencies of the scales were
moderate to high with Cronbachs alpha ranging from α =.33 for failure to accept
responsibility to α = .75 for behavioral dyscontrol. The CPS-RE total score (mean of
25
all 58 items) demonstrates moderate to strong correlations with caregiver reports of
childrens aggressive and antisocial behaviors (see Baker et al., 2007). In addition,
test-retest correlations within a 3-6 month period were calculated for the CPS-RE.
Test-retest correlations were all significant and ranged from r = .48 (lack of guilt) to
r = .81 (for boredom susceptibility), and r = .44 for total score.
The CPS-RE was primarily developed to operationalize, in childhood and
adolescence, the personality traits found in the Revised Psychopathy Checklist (PCL;
Hare, 2003). Two PCL-R items, criminal versatility and juvenile delinquency, were
not included so that the CPS-RE might serve as a pure measure of personality
uncontaminated by antisocial behavior. Further validation of the CPS-RE scales and
total psychopathy score is provided in Lynam 1997. In brief, Donald Lynam
provides detailed analyses to examining the relation between childhood psychopathy
and concomitant psychopathology. In his analyses validating the CPS-RE, Lynam
demonstrated that the means for stable, delinquent children were significantly higher
than the means for stable, non-delinquent children, and higher than children who
engaged in various delinquent acts on all CPS-RE subscales. Furthermore, Lynams
work also demonstrated that the CPS-RE predicted childhood psychopathy above
and beyond other well-known predictors including impulsivity, SES, previous
delinquency, and IQ (Lynam, 1997). Thus, the CPS-RE provide[s] a unique
proportion of variance in the prediction of delinquency above other well-known
predictors (Lynam, 1997; pg. 432).
26
Aggression
Each twins aggressive behavior was assessed through the Reactive and
Proactive Aggression Questionnaire (RPQ). The RPQ is a 23-item validated
questionnaire designed to measure reactive and proactive aggression in young
children and adolescents (Raine, 2006). The RPQ includes 11 reactive items (e.g.,
he/she damages things when he/she is mad; he/she gets mad or hit others when
they tease him/her; he/she gets angry when frustrated; he/she has temper
tantrums; he/she feels better after hitting or yelling at someone) and 12 proactive
items (e.g., he/she damages or breaks things for fun; he/she uses force to get
money or things from others; he/she threatens and bullies other kids). The RPQ
is scored on a three-point response scale: 0 = never, 1 = sometimes, 2 = often. Mean
scores for the item responses were computed to form reactive and proactive
aggression subscales. Caregiver reports of the twins aggressive behaviors were
used in the current study.
Externalizing and Internalizing Behavior Problems
The Diagnostic Interview Schedule for Children - Version IV (DISC-IV)
(Schaffer, Fisher, Lucas, Dulcan, & Schwab-Stone, 2000) was used to assess both
externalizing and internalizing behavior problems in the twins. The DISC-IV is a
highly structured interview designed to assess psychiatric disorders (adapted from
the DSM-III-R) and symptoms in children and adolescents aged 6 to 17 years. The
DISC-IV was designed to be administered by well-trained lay interviewers for
epidemiological research. The twins caregivers were administered several modules
27
including, the Conduct Disorder (CD), Generalized Anxiety (GA), Oppositional
Defiant Disorder (ODD), and Attention Deficit Hyperactivity Disorder (ADHD),
which were used for the current study. Both diagnoses and symptom counts are
provided through computerized scoring of the DISC-IV modules.
Statistical Analyses
The classical twin design capitalizes on the difference in genetic relatedness
between monozygotic (MZ) and dizygotic (DZ) pairs of twins and estimates the
relative genetic and environmental contributions to a phenotype. The traditional
approach divides the phenotypic variance (V
p
) into genetic (V
G
) and environmental
(V
E
) influences [V
p
=V
G
+V
E
]. Environmental factors may be divided further into
common or shared variances (V
C
) and unique or non-shared influences (V
U
). The
traditional quantitative genetic model is based on the notion that the observed
variance in a trait is a linear function of genetic (A), shared (C), and non-shared (E)
environmental variances. Genetic variance is said to reflect the dissimilarity in the
genotypes, which are transmitted directly from parents to offspring. Shared
environmental variance reflects the difference in environments that affects all
children growing up within a family to the same degree and differentiates among
families. Non-shared environmental variance reflects variation in environments that
has different effects on individual family members. Moreover, error or residual
variance is also included in the estimate of non-shared environment (Johnson &
Krueger, 2005; Plomin et al., 2001). MZ twins reared together share 100% of their
segregating genes as well as part of their environment. Similarity and resemblance
28
between them is thus due to genetic and shared environmental effects. Non-shared
environmental factors are the degree to which MZ twins do not resemble one
another. Similarity or resemblance between DZ twins reared together is due to
shared environment and shared genes. DZ twins on average share 50% of their
segregating genes. The extent to which DZ twins do not resemble each other is due
to non-shared environmental factors and to non-shared genetic effects. Evidence for
genetic influence is suggested when MZ twins show greater resemblance than DZ
twins in measures that are of interest.
Similarity or resemblance among twins may therefore be calculated through
Pearson intraclass correlations (Neale & Cardon, 1992). To obtain an initial estimate
of the relative variance or contribution of additive genetic effects or influences (A)
for a specific phenotype or trait, twice the difference between MZ and DZ
correlations may be calculated (A= 2 [r
MZ
- r
DZ
]). Shared environmental (C)
influences or effects are calculated by subtracting the correlation between MZ twins
from twice the DZ correlation (C= 2 r
DZ
r
MZ
). Non-shared environmental effects
(E) can be obtained by subtracting the MZ correlation from one (E=1-r
MZ
). Finally,
the contribution of dominant (D) or non-additive genetic effects may also be
determined by subtracting four times the DZ correlation from twice the MZ
correlation (D= 2rMZ- 4 rDZ) (Posthuma et al., 2003). It is impossible to estimate
(C)
and (D)
simultaneously or test the ACDE model with data from only twin pairs
reared together because the estimation of (C)
and (D)
both
rely on the same
information (i.e. the difference between MZ and DZ intraclass correlations). If the
29
DZ correlation is greater than half the MZ correlations, the ACE model is assumed to
be the correct model and the estimate of (D) in the ADE model is always estimated
as zero. If the DZ correlation is less than half of the MZ correlation, the ADE model
is assumed to be correct and the estimate of (C)
in the ACE model should
become zero.
Heritability is a population parameter that encompasses the proportion of
total phenotypic variation in a given characteristic that can be attributed to genetic
effects. Heritability may include additive as well as non-additive genetic variance.
There are two types of heritability; broad-sense and narrow-sense heritability:
broad-sense heritability includes all additive and non-additive genetic variability,
while narrow-sense heritability encompasses only additive genetic variability
(Plomin et al., 2001).
Structural Equation Modeling
Structural equation modeling was used to perform the genetic model-fitting
analyses in Mx (Neale et al., 2002). A basic twin model estimates the relative
influence of genetic and environmental components of variance for a specific
measure. Circles represent latent unmeasured variables, and squares denote the
observed phenotypes (measured variables) for each twin (please see Figure 1). As it
is assumed that MZ twins share 100% of their segregating genes, and DZ twins share
50% of their segregating genes, the correlation between the genetic factors is thus
fixed accordingly (MZ=1.0, DZ=0.5). However, for dominance effects the
correlation between DZ twins is reduced to .25. Other assumptions present within
30
this twin model are that there are shared environmental effects that influence both
MZ and DZ twins equally, therefore the correlation between the latent common
environmental factors is fixed at 1.0 for both sets of twins and that there is an
absence of assortative mating. Yet another assumption is that there is an absence of
gene-environment correlations. This model also assumes that the total variance can
be explained by the additive combination of separate genetic and environmental
influences, which sum to 100% or 1.0.
Figure 1. Univariate AC/DE path diagram in twins.
a
A E C/D
Tw in 1
A E C/D
Tw in 2
ac (d)
e
c (d) e
ra = MZ = 1.0 / DZ = 0.5 rc = MZ=1.0 (1.0) / DZ=1.0 (0.25)
Note: A = additive genetic effects; C=shared environmental effects; E= non-shared
environmental effects; D= non-additive genetic effects.
31
Multivariate Models
Models fit included saturated covariance models to estimate the phenotypic
means and variance-covariance matrices within each of the five zygosity groups; and
more constrained models to determine the magnitude and significance of genetic (A),
shared twin environment (C) and non-shared environment (E) variance for each
factor, as well as to understand the nature of the covariance between each factor i.e.,
the extent to which genetic and environmental aspects may influence the shared
genetic and environmental influences between the factors. Specifically, (1) a
Cholesky decomposition model (Figure 2) was employed to examine the genetic and
environmental effects of psychopathic traits; (2) an independent pathway (Figure 3)
model in which genetic and environmental influences were uncorrelated for the two
psychopathic personality factors; and (3) a more parsimonious common pathway
model (Figure 4) in which unitary latent factors for A, C and E were specified to
influence a common factor underlying the two psychopathic personality factors, in
addition to genetic and environmental influences specific to each factor.
32
Figure 2. Path diagram of an ACE Cholesky Decomposition for one twin.
Note: In this model, the variance is decomposed into additive genetic (A
1
, A
2
), and
shared (C
1
, C
2
) and non-shared (E
1
, E
2
) environmental effects. a
11,
c
11
and e
11
represent the additive genetic, shared and non-shared environmental contribution to
phenotype 1, respectively; a
21,
c
21
and e
21
paths represent additive genetic, shared
and non-shared environmental effects and contributions to phenotype 2, respectively;
a
22,
c
22
, and e
22 .
Figure 3. Independent Pathway Model in one twin.
Phenotype1
Cs As Es
A C E
Phenotype 2
Cs As Es
Phenotype 3
Cs As Es
Note: There are two types of genetic and environmental effects in this model (1)
general or common (A, C, E) and (2) specific (As, Cs, Es).
A
1
C
1
E
1
A
2
C
2
E
2
a
11
e
11
a
21
e
21
e
22
a
22
c
22
c
21
c
11
Phenotype 1 Phenotype 2
33
Figure 4. Common pathway model in one twin.
Phenotype1
Cs As Es
Latent
Factor
A C E
Phenotype2
Cs As Es
Phenotype3
Cs As Es
Note: In the common factor or pathway model, the general effects are mediated
through an underlying latent factor that represents the variance shared among the
measures. This model estimates fewer parameters than an independent pathway
model making it a more parsimonious model. In addition to the genetic and
environmental effects of the latent underlying factor, specific effects for the ACE
parameters are also estimated (As, Cs, Es).
The fit of the genetic models was assessed with χ
2
difference test, which is
the difference between -2 log likelihood (-2LL). This difference follows a χ
2
distribution. Degrees of freedom (df) for this test are equal to the difference between
the numbers of estimated parameters in two tested models. In addition to the
likelihood-ratio χ
2
test, AIC criterion was also computed. The AIC represents the
balance between model fit and the number of parameters (parsimony), with lower
values of AIC indicating the most suitable model (Akaike, 1987).
34
Exploratory (EFA) and Confirmatory (CFA) factor analyses
To investigate the first aim of the study, which is to identify the underlying
factor structure of psychopathic like traits in children using all fourteen of the CPS
scales, exploratory (EFA) and confirmatory factor analyses were undertaken. In
addition to phenotypic exploratory and confirmatory factor analyses, multivariate
genetic analyses were also performed to confirm the factor structure of the CPS
across the two twins.
The total sample of 1210 was split into 2 independent groups: group 1 and
group 2. Exploratory factor analysis (EFA) was conducted on the first group to
investigate the underlying factor structure of the CPS-RE in a child sample using the
CPS-RE subscales. The EFA revealed an underlying two-factor structure. Next,
confirmatory factor analyses (CFA) were conducted on group 2 using the EQS 6
software (Bentler, 2002) on the same underlying factor structure obtained in group
1s EFA to determine the adequacy of fit for the EFA model obtained. Mean scores
for items in the resulting factors were thus created and used for all subsequent
(genetic) analyses.
Additional multivariate genetic models were also fit to the CPS-RE
subscales, in order to confirm whether the factor structure in the EFA and CFA
appeared valid in both twins simultaneously.
35
Several goodness of fit indices were used to assess the fit of the factor
models: chi-square statistic χ
2
, Akaike Information Criterion (AIC= χ
2
2df;
Akaike, 1987), the root mean squared error approximation (RMSEA) index
(McDonald, 1989), and the comparative fit index (CFI) (Bentler, 1990). Direct
comparisons were made between the 1, 2 and 3 factor models using the difference in
the chi-square statistic (∆χ
2
). Hu and Bentler (1998) have provided evidence that
the CFI and RMSEA are among the most sensitive indicators of model
misspecification, particularly in cases of mis-specified factor loadings, and
consequently particular weight was given to these fit indices in evaluating model fit.
Evidence suggests that an RMSEA of less than 0.05 is indicative of a close-fitting
model (Browne & Cudeck, 1993; MacCallum, Browne, & Sugawara, 1996), as is a
CFI above 0.90 (Bentler & Bonett, 1980).
Validation of Psychopathic Traits in Children
To investigate the convergent and discriminant validity of psychopathic
personality traits in children (Aim 2), the relationship between psychopathic traits
with other validated measures of both externalizing and internalizing behaviors was
examined using phenotypic correlation coefficients, i.e. the Pearson product moment
(r). Convergent and discriminant validity are two subtypes of construct validity.
Convergent validity is the agreement among ratings or measures that should
theoretically be related, whereas discriminant validity is the lack of agreement
among measures that should not be related. To establish convergent validity of the
emergent psychopathic personality dimensions, psychopathic traits were examined
36
with other externalizing behaviors to which they should in theory be related. To
demonstrate discriminant validity of the emergent psychopathic factors as well as
total psychopathy in childhood, psychopathic traits were examined with other
behavioral measures (internalizing behaviors) to which psychopathic traits should
not, in theory, be as strongly related. In addition to examining the full range of total
psychopathy and factor scores, mean differences between groups defined by the high
and low ends of the total psychopathy score (CPS-RE total) were also examined.
Multivariate analyses of variances (MANOVA) were employed to examine
differences between the high and low psychopathy groups on various measures of
externalizing behaviors including CD, ODD, ADHD, and reactive and proactive
aggression. Partial eta squared was the effect size reported for all multivariate
analyses of variance, and it describes the proportion of variance accounted for.
Specifically, a 2(high and low psychopathy groups) x 2(gender) x 5(externalizing
behavior problems) MANOVA with symptom counts for CD, ADHD, ODD and
mean reactive and proactive aggression as dependent variables. Additionally, due to
the nature of the data, mixed model analyses were also employed to examine the
dyadic twin data. In general, type-1 error may be inflated in research involving the
observations that are not independent (such as in nested twin data); therefore it may
be necessary to treat individual scores as repeated measures in a pair using a mixed
model analysis of variance (ANOVA) (Singer, 1998). Degrees of freedom were
determined by Satterthwaite approximation, and can be a fractional number
37
(Campbell & Kashy, 2002). The aforementioned analyses were conducted in SPSS
15.0.
Genetic and Environmental Etiology of Psychopathic Traits in Children
To investigate the genetic and environmental influences of the emerging
psychopathic dimensions (Aim 3), multivariate genetic models were fit using the
structural equation-modeling program Mx (Neale, 2003) to examine the etiology of
psychopathic personality traits in children. In addition, it is also valuable to examine
whether sex specific influences are important for psychopathic like traits in children.
To assess and examine the role of sex differences in the genetic and environmental
etiology of psychopathic like traits, sex-limitation models will also be fitted
(Aim 3). A constrained model, with no specific sex effects will be fitted to
determine if the genetic and environmental components in males and females may be
constrained to be equal.
Genetic and Environmental Covariation between Psychopathic Traits and
Externalizing Behavioral Problems
A traditional twin design was also employed to examine the bivariate
covariation between the psychopathic personality factors and other behavior
problems including an investigation of the genetic and environmental correlations
between psychopathic traits and other behavioral problems. Twin correlations and
genetic model-fitting analyses were used to investigate the genetic and
environmental etiology of psychopathic personality traits with aggression, conduct
problems (CD and ODD), and ADHD (Aim 4). To get a first indication of the
38
underlying sources of variance among psychopathic personality traits and other
behavioral problems, intraclass twin correlations were examined (i.e., Twin-1 and
Twin-2 correlations within each measure). For instance, a (DZ) intraclass correlation
which is approximately half the value of the monozygotic (MZ) intraclass correlation
would signify the presence of genetic effects for a given trait, while a DZ intraclass
correlation more than half the MZ intraclass correlation signifies the presence of
both genetic and shared environmental effects. Cross-twin cross-trait correlations
were also examined to investigate the genetic and environmental correlations
between traits. However, formal model fitting analyses is needed to determine the
extent of genetic and environmental influences on specific traits. Data were
transformed using PROC RANK, a commonly used method in behavior genetic
research to transform data in order to approximate a normal distribution
(Burt et al., 2001).
39
Chapter Three: Results
Descriptive Statistics
Means and standard deviations for the CPS-RE scales are provided in
Table 1 for boys and girls separately. Scale means ranged from 1.11
(callousness) to 1.70 (grandiosity) for boys and 1.08 (callousness) and 1.69
(grandiosity) for girls. Mean sex differences were found for several of the
subscales with boys significantly higher than girls on glibness (t=2.64,
df=1214, p<.01); untruthfulness (t=2.06, df=1214; p<.05); boredom
susceptibility (t=5.89, df=1214, p<.001); callousness (t=4.13, df=1214,
p<.001); parasitic lifestyle (t=2.53, df=1214, p<.001); behavioral dyscontrol
(t=2.66, df=1214, p<.01); lack of planning (t=5.37, df=1214, p<.001),
impulsivity (t=5.32, df=1214, p<.001), unreliability (t=3.93, df=1214, p<.001),
failure to accept responsibility (t=2.82, df=1214, p<.01). There were no sex
differences found for manipulativeness (t=1.85, df=1214, p=.06), lack of guilt
(t=-1.07, df=1214, p=.29), poverty of affect (t=.94, df=1214, p=.34), and
grandiosity (t=.47, df=1214, p=.64). Caregiver endorsements of the items are
provided in Table A2 of the Appendix.
40
Table 1
Means and Standard Deviations for CPS-RE Scales by Sex
Boys Girls
CPS-RE Subscales Skewness Kurtosis Cronbachs α Mean SD Mean SD
Glibness* .43 -.61 α = .61 1.41 .28 1.37 .26
Untruthful* 1.38 1.56 α = .56 1.18 .22 1.15 .20
Boredom
Susceptibility*
.68 -.33 α = .46 1.29 .24 1.21 .22
Manipulative .72 -.69 α = .53 1.28 .30 1.25 .30
Lack of Guilt 1.00 1.18 α = .39 1.25 .26 1.26 .24
Poverty of Affect .82 .07 α = .38 1.18 .16 1.17 .15
Callousness* 2.10 5.06 α = .47 1.11 .18 1.08 .14
Parasitic Lifestyle* 1.32 1.97 α = .37 1.18 .20 1.15 .20
Behavioral
Dyscontrol*
.52 -1.27 α = .75 1.39 .40 1.34 .38
Lack of Planning* .95 -.28 α = .59 1.31 .36 1.21 .29
Impulsivity* .81 -.39 α = .52 1.34 .33 1.24 .29
Unreliability* 2.05 3.89 α = .43 1.27 .23 1.08 .18
Failure to Accept
Responsibility*
.20 -.87 α = .33 1.46 .31 1.41 .32
Grandiosity -.82 -.18 α = .54 1.70 .31 1.69 .31
* Indicates a significant sex difference in means.
Note: N=592 for males and 624 for females. CPS-RE items scored from 1-2 (1=No;
2=Yes).
41
Exploratory and Confirmatory Factor Analyses
To investigate the phenotypic factor structure of psychopathic traits in children
(Aim 1), exploratory and confirmatory factor analyses were undertaken. Due to the
dyadic nature of the twin data, two independent groups were formed (using the total
N=1210 sample) using a random number generator in SAS [random= ranuni (-1)].
Exploratory factor analysis was conducted on the first independent half (twin 1 or
group 1) of the total sample and revealed an optimal two-factor solution for the CPS.
Specifically, a principal axis factor analysis was conducted, with factors rotated
using an Oblimin oblique rotation. Kaisers criterion indicated three factors, but this
solution lacked simple structure and in addition one factor was defined by only one
scale. Furthermore, the last factor was uninterpretable and the only factor
intercorrelation that was higher than 0.2 was between factors 1 and 2 (r = 0.43;
p<.01). Accordingly, analyses were repeated to explore a two-factor solution. The
factor loadings are reported in Table 2. Inspection of the loadings suggested a factor
of callous/disinhibited traits (with loadings ranging from .41 to .62) and a
charming/manipulative factor (with loadings ranging from .43 to .81). Two
subscales (grandiosity and lack of guilt) were dropped from further analyses because
they failed to load consistently with either factor.
42
Table 2
Factor Structure and Loadings (EFA and CFA)
Factor 1: Callous/Disinhibited
(α = .79)
EFA CFA Factor 2: Charming/
Manipulative (α = .77)
EFA CFA
Unreliable .62 .57 Manipulative .81 .54
Poverty of Affect .53 .59 Failure to accept
Responsibility
.46 .61
Lack of Planning .56 .41 Glibness .55 .46
Boredom Susceptibility .49 .58 Parasitic Lifestyle .45 .68
Impulsivity .62 .71 Untruthful .43 .66
Behavioral Dyscontrol .45 .58
Callousness .41 .54
Note: Exploratory Factor Analyses (EFA) were conducted on the first independent
half of the sample (group 1) and Confirmatory Factor Analyses were conducted on
the second independent half of the sample (group 2). The subjects in the groups
were randomly selected from the larger pool of participants.
43
Figure 5. Illustration of the CPS-RE Factor Structure
The Callous/Disinhibited factor (Factor 1), which may be described as traits
that encompass behavioral disinhibition, while also embodying a sense of
callousness and dis-regulation of intense emotions includes the following seven
subscales: unreliability, poverty of affect, boredom susceptibility, impulsivity, lack of
planning, callousness, and behavioral dyscontrol, while the Charming/Manipulative
factor (factor 2) consists of five subscales including glibness, manipulative, parasitic
lifestyle, untruthfulness, and lack of responsibility (see Table 2; Figure 1), and may
be described as embodying the core traits of psychopathy and having the ability to
verbally manipulate and control situations and people through the use of charm.
Callousness
Behavioral Dyscontrol
Impulsivity
Unreliable
Poverty of Affect
Lack of Planning
Boredom Susceptibility
Fails to Accept Responsibility
Glibness
Parasitic Lifestyle
Untruthfulness
Manipulative
Callous/Disinhibited
Charming/Manipulative
44
Next, CFA was conducted on the second, independent half of our sample
(group 2) to test whether this two-factor model could be substantiated. CFA was
also conducted to test alternative, two, and three-factor models based on Hares
classic design, as well as a one-factor model. Subscale items were picked based on
previous psychopathy literature and theories. Specifically, the Hare 2 factors in
Table 3 refers to Hares Factor 1 (Interpersonal-Affective) and Factor 2 (Lifestyle-
Antisocial); the 3 Factors refers to the theoretical model proposed by Cooke and
Michie (2001); and 3 Factors b refers to the theoretical 3 Factor model excluding
the poor loading grandiosity and lack of guilt subscales (illustration of alternative
factor models can be found in Appendix Figure A2). The comparison between the
one, two and three factor models revealed a better fit for the two-factor solution of
the CPS-RE (∆χ
2
= 211.64, ∆ df=24, p<0.01 - see Table 3). The two-factor model fit
better than the one- and three-factor models on all three indices (see Table 3). In
addition, the two-factor structure was chosen not only because it was the more
parsimonious and better fitting model (with the lowest AIC), but also because it fell
in line with the psychopathy literature (PCL-R; Hare, 1991, 2003; Lynam, 1997).
The internal consistencies (based on item scores) for Factors 1 and 2 were adequate
with Cronbachs alpha= 0.79 and 0.77, respectively.
Several goodness of fit indices were reported to assess the fit of the models:
chi-square statistic χ
2
, Akiake Information Criterion (AIC; Akiake, 1987), the root
mean squared error approximation (RMSEA) index (McDonald, 1989), the
comparative fit index (CFI) (Bentler, 1990), and the nonnormed fit index (NNFI;
45
Garver & Mentzer, 1999). Direct comparisons were made between the 1, 2 and 3
factor models using the difference in the chi-square statistic (∆χ
2
). Hu and Bentler
(1998) have provided evidence that the CFI and RMSEA are among the most
sensitive indicators of model misspecification, particularly in cases of misspecified
factor loadings, and consequently particular weight was given to these fit indices in
evaluating model fit. Evidence suggests that an RMSEA of less than 0.08 is
indicative of a close-fitting model (Hoe, 2008; Browne & Cudeck, 1993;
MacCallum, Browne, & Sugawara, 1996), as is a CFI and NNFI above 0.90 (Hoe,
2008; Bentler & Bonett, 1980). The AIC represents the balance between model fit
and the number of parameters (parsimony), with lower values of AIC indicating the
most suitable model (Akaike, 1987).
Table 3
Model Fits for Confirmatory Factor Analyses
Model χ
2
df RMSEA CFI NNFI ∆χ
2
∆ df AIC p
1 Factor 525.45 77 .10 .78 .74 -- -- 371.45 <.01
2 Factors
(Present
Study)
313.81 53 .09 .87 .83 211.64 24 207.81 <.01
Hare 2-
Factors
521.15 76 .10 .79 .74 4.30 1 369.15 <.01
3 Factors 433.85 62 .10 .79 .74 91.60 15 309.85 <.01
3 Factors b 348.45 41 .11 .82 .76 177.00 36 266.35 <.01
* Model comparisons were made to the 1-factor model. However, all subsequent
models also fit worse when compared to the 2-factor model. Model descriptions are
provided in text.
Means and standard deviations for total score of psychopathic traits are
provided in Table 4. Means and standard deviations for the emergent two-factors
46
(Callous/Disinhibited and Charming/Manipulative) are also provided in Table 4.
Means for psychopathic traits in boys range from 1.25 to 1.30, whereas means for
girls range from 1.19 to 1.27. Mean sex differences were found for both
psychopathic dimensions and for the total psychopathy score with boys significantly
higher than girls on all factors: Charming/Manipulative (t=3.45, df=1214, p<.01),
Callous/Disinhibited (t=6.46, df=1214, p<.01), total psychopathy (t=5.58,
df=1214, p<.01).
Table 4
Means and Standard Deviations for the Two Psychopathic Trait Factors by Sex
Callous/Disinhibited Charming/Manipulative Total Psychopathy
Mean S.D. N Mean S.D. N Mean S.D. N
Boys 1.25 .18 592 1.30 .18 592 1.26 .13 592
Girls 1.19 .15 624 1.27 .18 624 1.22 .12 624
Multivariate genetic models were also fit to all fourteen of the CPS-RE
subscales in both twins simultaneously to confirm the validity of the same two-factor
solution found in the EFA (in one half of the sample) and the CFA (in the other half).
The two-factor solution was superior to a one-factor solution, and provided the most
interpretable solution compared to a three-factor model.
Psychopathic Traits and Externalizing Behavioral Problems (Convergent and
Discriminant Validity)
To investigate Aim 2, that is to examine the phenotypic relationships between
psychopathic traits and other behavior problems and to establish convergent validity
47
of the two psychopathic trait factors, Pearson correlations were calculated between
psychopathic traits and other behavioral problems. Correlations among the measures
are provided in Table 5 for the entire sample (correlations are presented for the entire
sample to facilitate the test of difference of association). A table of correlations
presented by sex can be found in Table A3 of the appendix. Significant moderate to
strong relationships between psychopathic traits and other externalizing behavioral
problems including aggression, CD, ODD and ADHD symptoms emerged.
Relationships ranged from a moderate r=.42, p<.01 (for the relationship between
Factor 2 and ADHD) to a stronger r=.59, p<.01 (for the relationship between total
psychopathy and Factor 2 with ADHD symptoms) (Table 5). To establish
discriminant validity for psychopathic traits, the relationship between generalized
anxiety and psychopathic traits was examined. Consistent with literature,
psychopathic traits were weakly correlated with symptoms of anxiety, which
supports the discriminant validity of the construct. To further establish discriminant
validity, the relationship between depressive symptoms and psychopathic traits was
also examined. A low but significant relationship between depressive symptoms and
psychopathic traits also emerged (Table 6).
Multiple regression analyses were performed on symptoms of ADHD, CD,
and the interaction between ADHD*CD using total psychopathy as the dependent
variable to determine if the combined ADHD*CD symptoms contributed
significantly beyond the contributions of ADHD, and CD. Main effects for sex and
the interaction of sex with ADHD and CD (and their product) were also examined in
48
a final model (variables entered sequentially). Results illustrated that R
2
change for
the model including the interaction terms was not significant (R
2
= .46,
F(3,1194) = 2.0, p=.11). The ADHD*CD interaction term did not significantly
contribute to predicting total psychopathy, beyond the significant contributions from
ADHD and CD alone (β=-.05 standardized; t=-.97, p=.33). ADHD significantly
predicted total psychopathy (β=.48 standardized; t=17.53, p<.01); CD also
contributed significantly in predicting total psychopathy (β=.37 standardized;
t=8.90, p<.01). There was no significant interaction effect for sex*ADHD
(β=.06 standardized; t=.86, p=.39), or for sex*CD (β=.12 standardized; t=1.65,
p=.10). These analyses were also repeated for Factors 1 and 2. Similarly, for Factors
1 and 2 R
2
was not statistically different from zero for the models including the
interaction effects [R
2
= .42, F(3, 1194) = .17, p=.91 for Factor 1 and R
2
= .31,
F(1, 1193) = .25, p=.26 for Factor 2]. Thus, results for Factor 1 demonstrated
significant main effects for sex (β= -.04 standardized; t= -2.0, p=.045), ADHD
(β=.53 standardized; t=21.60, p<.01), and CD only (β=.19 standardized; t=8.0,
p<.01). For Factor 2, there were significant main effects for ADHD (β=.36
standardized; t=13.87, p<.01), and CD (β=.31 standardized; t=3.2, p<.01). There
were no significant interaction effect for ADHD*CD (β= -.11 standardized; t= -2.0,
p=.05), or interaction effects with sex. Therefore, analyses pertaining to the
interaction of ADHD*CD (ADHD with CD symptoms) were not pursued any further
and CD and ADHD were examined separately.
49
Table 5
Phenotypic Correlations between Psychopathic Traits and Externalizing Behavior
Problems
Reactive
Aggression
Proactive
Aggression
CD ODD ADHD
CPS total score .56**
.52**
.46**
.45**
.59**
Callous/Disinhibited
(Factor 1)
.51**
.43**
.36**
.39**
.59**
Charming/Manipulative
(Factor 2)
.50**
.47**
.46**
.45**
.42**
** p<.01 Note: sample sizes range from N=1201 to 1216.
A test of difference between two correlation coefficients was conducted to
determine if the difference in correlations between the two psychopathic personality
dimensions with CD, ODD, ADHD, and reactive and proactive aggression was
significant. The twopcor function (Wilcox, 2003) in the R program was used, which
implements a boot-strapping method or calculate 95% confidence intervals of the
difference in the relationships. The difference is said to be significant if the CI does
not contain zero. The tests revealed that the relationship between Factor 1
(callous/disinhibited) and CD was significantly weaker than the relationship between
Factor 2 (charming/manipulative) and CD (95% CI for difference in
correlations= .02 to .15). Similarly, previous studies have also found moderate
relationships between psychopathic traits and CD (Frick et al., 2003) in community
samples of children. In contrast, the relationship between Factor 1 and ADHD was
significantly stronger than the association between Factor 2 and ADHD (95% CI for
50
difference in correlations = .11 to .23). Additionally, to help support the claim that
psychopathic traits should demonstrate stronger relationships with externalizing
behavior problems than with internalizing behavior problems, the relationships
between both Factors 1 and 2 with the externalizing behavior problems were each
found to be significantly greater than the relationships between Factors 1 and 2 with
symptoms of generalized anxiety (Tables 4 and 5). For instance, the association
between Factor 1 and symptoms of GA was significantly greater than the
relationship between Factor 2 and GA (95% CI for difference in
correlations = .03 to .19), which is falls in line with the literature. This gives further
support to the convergent and discriminant validity of the two psychopathic
personality dimensions identified in this study. Additionally, the association
between Factor 1 and CD was significantly greater than the relationship between
Factor 1 and GA (95% CI for difference in correlations = .03 to .19). Similarly, the
association between Factor 1 and ADHD was significantly greater than the
association between Factor 1 and GA (95% CI for difference in
correlations = .28 to .41). The relationship between Factor 1 and ODD was also
significantly greater than the association between Factor 1 and GA (95% CI for
difference in correlations = .06 to .20). Likewise, the association between Factor 1
and reactive aggression was significantly greater than the relationship between
Factor 1 and GA (95% CI for difference in correlations = .18 to .31). In addition,
the relationship between Factor 2 and ADHD was also significantly greater than the
association between Factor 2 and GA (95% CI for difference in
51
correlations = .21 to .35), which further supports the notion that psychopathic traits
are more strongly associated with externalizing behavior problems than with
internalizing behavior problems.
Table 6
Phenotypic correlations between psychopathic
Traits and DISC-IV symptoms
of generalized anxiety (GA)
GA
CPS total score .21**
Callous/Disinhibited .26**
Charming/Manipulative .15**
** p<.01. Sample sizes N=1202 for GA
To help illustrate some of the traits and features of a psychopathic-like
child, Table 7 provides example cases of children high on psychopathic traits
according to caregiver endorsements of the CPS-RE items. The three example cases
were chosen because they illustrated kids high on psychopathic traits, and high on
other externalizing behavior problems according to caregiver endorsement. The first
case falls within the 100
th
percentile for Factor 1 and the 98
th
percentile for Factor 2,
while the second and third cases fall on the 99
th
percentile for both Factors 1 and 2.
Table 7 also provides number of symptom endorsements for CD, ADHD, and ODD
to help demonstrate the relationship between psychopathic traits and externalizing
behaviors. In example case #1, the caregivers report suggests that this child has a
52
difficult time waiting for her turn, is mean to others and teases others is unreliable
and breaks promises, and is easily frustrated and loses her temper. Thus, this child
appears to be impulsive, overactive, and callous (traits found in Factor 1). The
caregiver report also indicated that this child is capable of charming others to get his
or her way, takes advantage of others, is capable of talking his or her way out of
trouble, and therefore appears embody traits that resemble glibness,
manipulativeness, and has a parasitic orientation (some of the hallmark traits of
psychopathy). Caregiver reports of case #1 also suggest that this psychopathic-
like child show signs of CD, ADHD, and ODD, which help support the notion that
psychopathic traits are related to other externalizing behaviors (at a young age).
However, CPS-RE item endorsements indicate that this child exhibits signs of
manipulativeness, parasitic orientation, glibness at an early age. To help support the
utility and necessity of a two-factor structure of psychopathic traits in children, item
endorsements for cases 2 and 3 indicate that these MZ male twins are not only high
on Factor 1 (Callous/Disinhibited), but also high on Factor 2
(Charming/Manipulative). In addition, for cases 2 and 3 caregiver reports revealed
that these children had very few or no symptoms of anxiety. Furthermore, the mean
for caregiver endorsements of factor 2 (Charming/Manipulative) items was
significantly greater than that of Factor 1 (Callous/Disinhibited) (t= -12.95, df=1215,
p<.01), suggesting that caregivers more highly endorsed the items that pertain to
manipulating others, blaming others, and taking advantage of others (Table A2 in
appendix provides item endorsements). Furthermore, caregiver endorsements of
53
CD, and ODD are lower in cases 2 and 3 compared to case 1, suggesting that kids
high on psychopathic traits may be qualitatively different from kids who exhibit
symptoms of CD or ODD (Lynam, 1997; Blair; Frick et al., 1994). On the opposite
end of the spectrum, an example case (female) of a child low on psychopathic traits
revealed that the caregiver endorsed zero items on CPS-RE, and did not endorse any
symptoms of CD, ADHD, ODD, or reactive or proactive aggression. Thus, children
who do not exhibit psychopathic-like traits seem not to exhibit symptoms of CD,
ADHD, ODD or aggression.
54
Table 7 (continued on next page) Caregiver item endorsements for psychopathic personality traits, CD, ADHD,
ODD for three example cases
Case Callous/Disinhibited Charming/Manipulative CD symptoms ADHD
symptoms
ODD
symptoms
#1
(female)
-Hard time waiting for things
-Easily bored; needs thrills
-Does risky things for fun
-Doesnt plan ahead; no goals
-Feelings seems fake
-Is mean & teases others
-Breaks promises
-Not warm and kind
-Easily frustrated; loses temper
-Doesnt set or reach goals
-Cant be counted on
-Acts charming to get his/her way
-Tells stories to look good
-Is talkative
-Can talk his way out of trouble
-Blames others for wrongdoings
-Takes advantage of others
-Not shy; Needs to be the center
of attention; Shows off
14/16 items endorsed
-Shoplifted
-Stolen things
-Lied to get money
-Damaged things
-Bullies others; fights
16/21 items
endorsed
-trouble
concentrating
-overactive;
fidgety
-interrupts
others
-blurts out
answers
-cant wait turn
12/12 items
endorsed
-loses temper
-argues, talk
back
-blames others
for things;
Note: In addition, the caregiver for Case #1 endorsed 11/11 items on the RPQ-Reactive Aggression and 9/12 items
on the RPQ-Proactive Aggression scale.
55
Table 7 (Table 7, continued) Caregiver item endorsements for psychopathic personality traits, CD, ADHD, ODD for
three example cases
Case Callous/Disinhibited Charming/Manipulative CD
symptoms
ADHD
symptoms
ODD
symptoms
#2 male -Breaks promises
-Cant be counted on
-Is mean & teases others
-Easily frustrated; loses temper
-Hard time waiting for things
-Doesnt think about actions and
behavior
-Unpredictable Moods
-Feelings change often and quickly
-Cant tell how others feel
-Doesnt plan ahead; no goals
-Acts charming to get his way
-Takes advantage of others
-Good at lying; Not open and honest
-Cant be trusted
-Blames others for wrongdoings
-Can talk his way out of trouble
-Does not share with others
-Takes but doesnt give in return
-Tries to see how much he can get away with
6/16 items
endorsed
8/21 items
endorsed
5/12 items
endorsed
#3 MZ male
twin of case
#2
-Is mean & teases others
-Easily frustrated; loses temper
-Hard time waiting for things
-Doesnt think about actions and
behavior
-Unpredictable Moods
-Feelings change often and quickly
-Cant tell how others feel
-Doesnt plan ahead; no goals
-Acts charming to get his way
-Takes advantage of others
-Good at lying; Not open and honest
-Cant be trusted
-Blames others for wrongdoings
-Can talk his way out of trouble
-Does not share with others
- Takes but doesnt give in return
-Tries to see how much he can get away with
5/16 items
endorsed
14/21 items
endorsed
8/12 items
endorsed
Note: In addition, the caregiver for Case #1 endorsed 11/11 items on the RPQ-Reactive Aggression and 9/12 items on the
RPQ-Proactive Aggression scale.
56
Mean symptom counts for CD, ODD, ADHD, and mean reactive and proactive
aggression for boys and girls (on the total sample) are provided in Table 8.
Complete descriptions of these measures are provided in Baker et al. 2006. Mean
differences were found for psychopathic dimensions (means provided in Table 4),
reactive and proactive aggression, CD, ODD, and ADHD with CD with boys
significantly higher than girls on all factors: Charming/Manipulative (t=3.45,
df=1214, p<.01), Callous/Disinhibited (t=6.46, df=1214, p<.01), total psychopathy
(t=5.58, df=1214, p<.01), reactive aggression (t=4.18, df=1214, p<.01), proactive
aggression (t=2.65, df=1214, p<.01), CD (t=5.25, df=1203; p<.01) ; and ODD
(t=2.30, df=1200, p<.05). No significant mean differences across sex were found for
generalized anxiety symptoms (t=-.24, df=1200, p=.81).
Table 8
Means and SDs for CD, ODD, ADHD symptom counts and Reactive and Proactive
Aggression by Sex
CD ODD ADHD Reactive Aggression Proactive Aggression
Mean S.D. Mean S.D. Mean S.D. Mean S.D. Mean S.D.
Boys 1.43 2.02 4.8 3.3 6.39 5.32 .71 .33 .11 .17
Girls .89 1.54 4.36 3.18 4.35 4.51 .63 .33 .08 .14
S.D. = Standard Deviation
Sample size ranges from N= 589-592 for boys and N= 616 to 624 for girls.
Reactive Aggression (mean) ranges from 0 to 1.91; and Proactive aggression ranges
from 0 to 1.7.
To examine further the relationship between psychopathic personality traits
and externalizing behaviors in children, high and low psychopathy groups were
57
created using the total psychopathy score. As in previous studies, taking one
standard deviation above the mean (which corresponded to the 85
th
percentile of the
sample) for the high on psychopathy group, and taking one standard deviation below
the mean (15
th
percentile of the sample) for the low on psychopathy group created
groups high and low on psychopathy (Glenn et al., 2007). The mean and standard
deviation for total psychopathy (on the total sample) was 1.24 and .13, respectively.
The low psychopathy group had a (mean=1.08 and SD=.02); and the high
psychopathy group had a (mean=1.47 and SD=.09).
Table 9
Mean scores and symptoms counts of Reactive and Proactive aggression, CD,
ADHD, ODD for high and low psychopathy groups
Reactive Proactive CD ADHD ODD
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Low on Total
Psychopathy (N=184)
.41 (.23) .02 (.05) .36 (.76) 1.86 (2.36) 2.59 (2.27)
High on Total
Psychopathy (N=175)
1.01 (.33)* .26 (.26)* 2.92 (2.9)* 11.38 (4.82)* 7.59 (3.01)*
*indicates significant difference in groups for presented externalizing behaviors.
Standard deviations (SD) are presented in parentheses.
Multivariate analyses of variance revealed significant effects for psychopathy
group F (5, 342) = 168.79, p<.001, η
2
= .71 . There were no significant main effects
for sex F (5, 342) = 1.81, p=.11, η
2
= .03 or for sex*psychopathy group
F (5, 342) = .49, p=.78, η
2
= .01. Tests of between-subjects effects revealed that the
high on psychopathy group also had higher symptom counts of CD
58
F (1, 346) = 126.04, p<.001, η
2
= .27; significantly higher symptoms counts of ODD
F (1, 346) = 314.29, p<.001, η
2
=.48; ADHD F(1, 346) = 509.61, p<.001, η
2
= .60;
greater mean reactive aggression F(1, 346) = 352.98, p<.001, η
2
= .51; and greater
mean proactive aggression F(1, 346) = 138.68, p<.001, η
2
= .29 (Table 9). For
illustrative purposes, Figure 6 depicts group differences in standardized (z-score)
mean levels for CD, ADHD, ODD, GA, and reactive and proactive aggression.
59
Figure 6. Standardized Symptom Counts and Means for CD, ADHD, ODD, GA and
Reactive and Proactive aggression for the low and high psychopathic trait groups.
In addition to the multivariate analyses of variance, mixed model ANOVAs
were also employed (due to the nature of the twin data) to examine group differences
between the high and low psychopathy groups with other externalizing behaviors.
Similar to the results obtained from the multivariate analyses, mixed model
ANOVAs revealed that the group high on psychopathy displayed significantly higher
symptom counts of CD (t=11.14, df= 308.24, p<.001), higher symptom counts of
ADHD (t=23.33, df= 282.71, p<.001), ODD (t=16.72, df=343.43, p<.001), higher
Psychopathy Groups
High on Psychpopathic Traits Low on Psychopathic Traits
12
10
8
6
4
2
0
-2
Standardized Mean Level (z-score)
G A Symptoms
ODD S ymptoms
ADHD Symptoms
CD Symp tom s
Pr oa ctive Aggression
Reactive A ggression
60
mean reactive aggression (t=18.18, df= 334.91, p<.001), and higher means for
proactive aggression (t=11.58, df= 305.51, p<.001).
Genetic Analyses of Psychopathic Traits
To examine the genetic and environmental influences on the two
psychopathic personality dimensions (Aim 3), multivariate genetic analyses were
employed. Twin correlations for the two CPS-RE factors within each of the five-
zygosity groups are reported in Table 10. The twin correlations are consistently
higher for MZ pairs (both boys and girls) than for DZ pairs, suggesting the presence
of a genetic influence (A) in the CPS-RE factors. The twin correlations are mostly
moderate; except for the DZ male correlations which are anomalously low and non-
significant. MZ correlations exceed the DZ correlations in both psychopathic
personality factors in both males and females, suggesting the presence of genetic
influence. There is also a suggestion of dominance effects in the males where the
MZ correlations are greater than twice the DZ correlations, although this effect is not
observed in the females. These patterns were confirmed in multivariate genetic
models, as described next.
Table 10
Twin Correlations for the CPS-RE Factors
CPS Factor 1
Callous/disinhibited
CPS Factor 2
Charming/manipulative
MZ males 0.63** 0.50**
DZ males 0.11 0.02
MZ females 0.53** 0.58**
DZ females 0.30** 0.35**
DZ opp. sex 0.26** 0.27**
**p < 0.01.
61
A summary of the various multivariate genetic models fit to the two CPS-RE
factors is provided in Table 11 along with the fit statistics for each model and
comparisons to the fully saturated covariance model. Among the three types of
models fit in which variance components were allowed to differ across sexes (m≠f),
i.e., the bivariate Cholesky (Model 1), the common pathways (CP; Model 10), and
the independent pathways (IP; Model 12), the Cholesky model for the two CPS-RE
factors demonstrated the best fit to the data (see Table 11). Constraining parameters
to be equal for males and females (m=f) provided significant loss of fit in all cases
(Models 2, 11, and 13). Further analyses based on the better fitting Cholesky model
also revealed that shared environment (C) could be dropped from the model for both
boys and girls without a significant loss in fit as demonstrated by the lower Akaike
(AIC) index (see Table 11).
62
Table 11
Model fit comparisons for the two CPS-RE Factors
Model -2ll df AIC χ
2
∆ df P
1. Saturated covariance model* -2269.51 2362
- - - -
2. Bivariate Cholesky m≠f -2223.43 2394 -17.92 46.08 32 .051
3. Bivariate Cholesky m=f -2192.16 2403 -4.65 77.35 41 .001
4. Cholesky m≠f; No C in boys -2223.14 2397 -23.63 46.37 35 .095
5. Cholesky m≠f; No C in girls -2222.95 2397 -23.44 46.56 35 .092
6. Cholesky m≠f; No C in boys or
girls
-2222.95 2400 -29.44 46.56 38 .16
7. Cholesky m≠f; No A in boys -2186.93 2397 12.58 82.58 35 .001
8. Cholesky m≠f; No A in girls -2201.26 2397 -1.748 68.25 35 .001
9. Cholesky m≠f; No A in boys or
girls
-2165.79 2400 27.72 103.72 38 .001
10. 1-factor CP m≠f -2089.49 2392 120.02 180.02 30 .001
11. 1-factor CP m=f -2088.29 2395 115.22 181.22 33 .001
12. 1-factor IP m≠f -2223.54 2388 -6.035 45.96 26 .009
13. 1-factor IP m=f -2217.36 2394 -11.85 52.15 32 .014
14. 1-factor IP m=f; no common C -2217.36 2396 -15.86 52.15 34 .024
*All subsequent models are compared to this fully saturated model. IP= common
factor independent pathways model. CP= common factor common pathways model.
Better-fitting models are presented in bold.
Due to the pattern of correlations in males, whereby rMZ>2rDZ, we also fit a
bivariate Cholesky model with dominance effects instead of common environment in
both males and females. Although this model fit better (AIC = -23.56) than Model 2
(AIC =-17.92) in Table 11, it fit less well than the best-fitting Model 6
(AIC = -29.44). The parameter estimates from this alternative model did suggest the
presence of dominance effects in males, but not females, whereby the latter showed
only A and E effects as estimated in Model 2.
These effects were confirmed in univariate analyses of same sex pairs for
CPS-RE Factors 1 and 2. As expected, univariate analyses revealed a DE (genetic
dominance and non-shared environment) model in boys for both Factor 1
63
(callous/disinhibited) and Factor 2 (charming/manipulative), which fit well (p=.61
for Factor 1; and p=.65 for Factor 2). Results also revealed an AE (additive genetic
and non-shared environment) model in the girls as best fitting for both Factor 1
(p=.34) and Factor 2 (p=.68). Although a model with genetic dominance effects in
males and additive genetic effects in females might be considered based on the same-
sex twin correlations, this would suggest that negligible correlation for opposite sex
pairs should be found. Given the moderately high DZ male-female correlation (see
Table 10), however, the bivariate model with additive genetic influences instead of
dominance was deemed more plausible for the CPS-RE factors. Although we are not
aware of any biological model that would allow for different genetic etiologies in
each sex, with considerable opposite sex DZ twin resemblance at the same time, the
possibility of dominance genetic effects for psychopathy could be considered in
other studies.
The full model ACE estimates for boys and girls from Model 2 are provided
in Table A1 of the appendix. Significant genetic as well as unique environmental
effects for Factors 1 and 2 in both boys and girls emerged. Shared environmental
effects were small and non-significant (ranging from 0 to 0.06). Heritability
estimates for factors 1 and 2 were (h
2
= 0.64 and h
2
= 0.46, respectively in boys; and
h
2
= 0.49 and h
2
= 0.58, respectively in girls). The unique environmental effects for
Factors 1 and 2 were (e
2
= .36 and e
2
= .53, respectively in boys; and e
2
= .44 and
e
2
= .37, respectively in girls). Best fitting estimates are illustrated in Figure 7.
64
Figure 7. Best-fitting ACE estimates of the psychopathic personality traits
Boys Girls
Bivariate analyses tell us what the contributions of genetic and
environmental factors to the covariance between two traits are. Therefore, two traits
(callous/disinhibited and charming/manipulative factors) may be related because
they share common genetic or common environmental influences. Thus, the same
genetic/environmental factors may be influencing the traits. Furthermore, genetic
analyses conducted on the two CPS-RE factors revealed significant correlations
between genetic (r
g
= .48 for boys; r
g
= .57 for girls) and non-shared (r
e
= .41 for boys;
r
e
= .30 for girls)
environmental factors (see Table 12). The common or shared
environmental (r
c
= .96 in boys; r
c
= .83 in girls) correlation was not significant in
either boys or girls, which is not surprising given the non-significant C values for
either factor illustrated in Figure 7.
00.5 1
Callous/Dis
Charm/Manip
A-Genetic
C-Shared
Environment
E- Unique
Environment
00.5 1
Callous/Dis
Charm/Manip
A-Genetic
C-Shared
Environment
E-Unnique
Environment
65
Table 12
Genetic and Environmental Correlations
for CPS-RE Factors 1 & 2 (boys, girls)
CPS-RE Factors 1 & 2
Boys Girls
r
g
.48* (.28-.72) .57* (.29-.89)
r
c
.96
ns
(-1.0 1.0) .83
ns
(-1.0 -1.0)
r
e
.41* (.26-.54) .30* (.15-.43)
*Parameter is significant at the .05 level. 95% confidence intervals are in
parentheses.
Note: r
g
= genetic correlation between factors 1 & 2; r
c
= common
environmental correlations between factors 1 & 2; and r
e
= unique (or
unshared) environmental correlations.
Genetic and Environmental Covariation between Psychopathic Personality
Traits and Externalizing Behavior Problems
Twin correlations for the two psychopathic personality factors and reactive
and proactive aggression, CD, ODD, within each of the five-zygosity groups are
reported in Table 13. The twin correlations are consistently higher for MZ pairs
(both boys and girls) than for DZ pairs, suggesting the presence of a genetic
influence (A) in all measures. The twin correlations are mostly moderate; except for
DZ males where the correlations are lower, and in some instances non-significant
(particularly for psychopathic traits).
Bivariate genetic analyses using Cholesky decompositions were implemented
to examine the relationship between psychopathic traits and other behavioral
problems. Bivariate analyses provide information on what the contributions of
genetic and environmental factors to the covariance between two traits are.
66
Therefore, two traits or phenotypes (in this case either psychopathic factor with
either reactive or proactive aggression) may be related because they share common
genetic or common environmental influences. Thus, the same genetic/environmental
factors may be influencing the traits. Specifically, the genetic and environmental
relationships between psychopathic personality traits and CD, ODD, ADHD, and
reactive and proactive aggression were examined separately.
The genetic and environmental correlation estimates from the best-fitting
bivariate Cholesky decompositions are provided in Table 14. Sexes could be
constrained to be equal without a significant loss in fit for all relationships except for
psychopathy and CD, even though there were significant mean differences in sex for
all measures. Analyses revealed significant genetic as well as unique environmental
correlations between the two psychopathic personality factors with CD, ODD, and
reactive aggression see Table 14. Specifically, genetic and unique environmental
correlations between the Callous/Disinhibited factor with reactive aggression were
(r
g
= .49, r
e
= .47, respectively) indicating that there is a significant genetic overlap
between these two traits. The genetic and environmental correlations between
Callous/Disinhibited and proactive aggression were (r
g
= .41, r
e
= .37, respectively),
however the genetic correlation between these two traits was not significant.
Similarly, the genetic and environmental correlations between the
Charming/Manipulative factor with reactive aggression were significant (r
g
= .57,
r
e
= .34, respectively), while the genetic correlation between Charming/Manipulative
with proactive aggression (r
g
= .35, r
e
= .41) was not significant. The common or
67
shared environmental correlations were non significant between both psychopathic
personality dimensions and all other externalizing behaviors examined including CD,
ODD, and reactive and proactive aggression (Table 14).
Significant genetic overlap was also found between both psychopathic
personality dimensions with CD, ADHD, and ODD (Table 14). The genetic and
environmental correlation between Callous/Disinhibited and CD were (r
g
= .52,
r
e
= .11 in boys; and r
g
= .44, r
e
= .14 in girls), while the genetic correlation between
Charming/Manipulative and CD were (r
g
= .51, r
e
= .36 in boys; and r
g
= .62, r
e
= .16 in
girls). Similarly, the genetic and environmental correlations between (Factor 1)
Callous/Disinhibited and ADHD, and with ODD were also significant (r
g
= .69,
r
e
= .43 with ADHD; r
g
= .72, r
e
= .35 with ODD) indicating a strong, significant
etiological relationship between Factor 1 and other externalizing behaviors.
Likewise, the genetic and environmental correlations between Factor 2
(Charming/Manipulative) and ADHD, and with ODD were also significantly
correlated (r
g
= .52, r
e
= .27 with ADHD; r
g
= .65, r
e
= .32). Thus, significant genetic
covariation was found between psychopathic traits and other externalizing behaviors;
however, significant unique environmental correlations between psychopathic traits
and symptoms counts of CD, ADHD, and ODD were also found suggesting that the
etiological covariation or relationship between psychopathic traits and other
externalizing behaviors are not entirely mediated by genes or under genetic
influences. Best fitting bivariate Cholesky decompositions are provided in Tables 2
68
and 3 of the Appendix. Illustration of a bivariate correlation model is provided in
Figure 1 of the Appendix.
69
Table 13 (continued on next page)
Intra-class correlations (on diagonals) and cross-twin cross-trait correlations (off diagonals) by zygosity
(MZ , DZ) and sex (male, female)
Call/Disinhib Charm/Manip CD ODD Reactive Proactive
MZ male
Callous/Disinhibited .63 .28 .31 .20 .20 .22
Charming/Manipulative .50 .29 .28 .31 .24
CD .68 .46 .42 .39
ODD .61 .42 .23
Reactive Aggression .48 .33
Proactive Aggression .61
DZ male
Callous/Disinhibited .11 -.13 .11 .08 .13 .10
Charming/Manipulative .02 .20 .06 .07 .08
CD .56 .34 .15 15
ODD .38 .23 .03
Reactive Aggression .35 .17
Proactive Aggression .34
Significant correlations are in bold. Intra-class twin correlations are in italics and bold on the diagonals.
70
Table 13 (Table 13, continued from previous page)
Intra-class correlations (on diagonals) and cross-twin cross-trait correlations (off diagonals) by zygosity
(MZ , DZ) and sex (male, female)
Call/Disinhib Charm/Manip CD ODD Reactive Proactive
MZ female
Callous/Disinhibited .53 .18 .25 .13 .24 .20
Charming/Manipulative .58 .43 .31 .35 .30
CD .68 .27 .28 .27
ODD .64 .48 .36
Reactive Aggression .60 .41
Proactive Aggression .57
DZ female
Callous/Disinhibited .30 .25 -.00 -.13 .26 .18
Charming/Manipulative .35 .26 .09 .37 .32
CD .49 .31 .39 .40
ODD .51 .25 .31
Reactive Aggression .46 .38
Proactive Aggression .48
DZ OS
Callous/Disinhibited .26 .25 .15 .16 .26 .25
Charming/Manipulative .27 .17 .16 .23 .24
CD .32 .17 .12 .11
ODD .38 .26 .23
Reactive Aggression .50 .35
Proactive Aggression .55
Significant correlations are in bold. Intra-class twin correlations are in italics and bold on the diagonals.
71
Table 14
Parameter Estimates from Best-Fitting Models -Genetic and Environmental Correlations between psychopathic traits and
other behavioral problems
Genetic and Environmental Correlations
Boys Girls
Callous/Disinhibited Charming/Manipulative Callous/Disinhibited Charming/Manipulative
Genetic
Correlation
r g
Shared
Env.
r c
Unique
Env.
r e
Genetic
Correlation
r g
Shared
Env.
r c
Unique
Env.
r e
Genetic
Correlation
r g
Shared
Env.
r c
Unique
Env.
r e
Genetic
Correlation
r g
Shared
Env.
r c
Unique
Env.
r e
CD .52*
(.38-.64)
0
.11*
(-.04-.26)
.51*
(.37-.64)
0
.36
(.22-.49)
.44*
(.31-.57)
0 .14
(-.02-.34)
.62*
(.51-.73)
0 .16*
(.10-.31)
ADHD .69*
(.61-.75)
0
.43*
(.34-.52)
.52*
(.41-.61)
0
.27
(.16-.37)
SAME AS BOYS SAME AS BOYS
ODD .72*
(.45- 1.0)
0
.35*
(.24-.45)
.65*
(.36-.96)
0
.32
(.21-.42)
SAME AS BOYS SAME AS BOYS
Reactive
Aggression
.49*
(.06 - .81)
0
.47*
(.38 - .55)
.57*
(.17 1.0)
0
.34
(.23 - .43)
SAME AS BOYS SAME AS BOYS
Proactive
Aggression
.41
(-.00- .72)
0
.37
(.27 - .47)
.35
(-.20.62)
0
.41
(.31 - .50)
SAME AS BOYS SAME AS BOYS
*Parameter is significant at the .05 level. 95% confidence intervals are in parentheses.
Note: r
g
= genetic correlation r
c
= common environmental correlations and r
e
= unique (or unshared) environmental
correlations.
72
Chapter Four: Discussion
The present study demonstrated that (1) a two factor structure for
psychopathic personality traits in children was the most optimal solution to the data,
similar to the structure found in adult psychopaths (Hare, 1991, Lynam, 1997);
(2) moderately strong and significant relationships emerged as expected between
psychopathic traits and other behavioral problems including CD, ODD, ADHD, and
aggression, while weak but significant relationships emerged between psychopathic
traits with anxiety and depression; results also demonstrated that individuals
comparatively higher on total psychopathy also exhibited higher mean symptom
counts of CD, ODD, ADHD, and higher means for reactive and proactive aggression
when compared to individuals low on total psychopathic traits; (3) genetic and
unique environmental influences were significant for both psychopathic personality
factors in both boys and girls, with no shared environmental influences on
psychopathic traits; and (4) significant genetic and unique environmental
relationships appeared between psychopathic traits and other behavioral problems
including CD, ADHD, ODD, and reactive aggression.
Factor Structure of Psychopathic Personality Traits
Exploratory and confirmatory factor analyses demonstrated an underlying
two-factor structure to be the best fitting for psychopathic personality traits within
this preadolescent sample. This two-factor structure was confirmed using twelve of
the fourteen CPS-RE subscales. Two subscales, grandiosity and lack of guilt, were
dropped from the analyses due to lack of clear factor loading estimates.
73
The first factor, Callous/Disinhibited, is comprised of the unreliable, poverty
of affect, lack of planning, boredom susceptibility, impulsivity, behavioral
dyscontrol, and callousness subscales, while factor two, the Charming/Manipulative
factor, consists of the manipulativeness, failure to accept responsibility, glibness,
parasitic lifestyle, and untruthfulness subscales. The two factors were moderately
correlated with each other and displayed moderate to strong reliabilities with a
Cronbachs alpha = 0.73 for Factor 1 (Callous/Disinhibited) and α=0.71 for Factor 2
(Charming/Manipulative).
These results fall in line with previous studies that have also found a two-
factor solution to account for psychopathic traits in community samples of children
and adolescents (Taylor et al., 2003; Frick et al., 1994; Salekin et al., 2001; Brandt et
al, 1997; Lynam, 1997). A study conducted on 6-13 year olds found a two-factor
correlated solution to be optimal, much like in the adult literature (Frick et al., 1994).
Other studies have also explored the factor structure of psychopathic traits on
adolescent samples (Brandt et al., 1995), and also found a two-factor structure for
psychopathy, replicating to some extent what has been found with adult psychopaths
(Hare, 1991). The two factors that emerged in the present study, however, closely
resemble but are slightly different from the two-factor structure found in the adult
psychopathy literature (Hare, 1991). Hares (1991) two factors are comprised of
affective-interpersonal features (e.g. glibness, superficial charm, lack of empathy,
callousness; Factor 1), and socially deviant lifestyles and behaviors (e.g.
impulsivity, sensation seeking, irresponsibility, poor behavioral control; Factor 2),
74
whereas the first factor identified in the present study, the Callous/Disinhibited
factor, includes characteristics and traits such as behavioral dyscontrol, callousness,
and lack of empathy, and impulsivity while the second factor, the
Charming/Manipulative factor, includes characteristics that demonstrate more of a
verbal ability such as glibness, manipulativeness, and untruthfulness. Slight
differences in the factor structures may be due to the fact that we have a young
preadolescent sample. One might not necessarily expect to find identical factor
structures in adults and in children due to developmental reasons. Slight differences
in the factor structures may be due to the fact that the sample is comprised of a
young, preadolescent community sample that has yet to experience the influences of
others including peers. Differences may also arise due to changes in development.
One might not necessarily expect to find identical factor structures in adults and in
children at such a young age. Adult psychopathy may be confounded by the fact that
adult psychopaths have experienced years of drug abuse, multiple incarcerations, and
physical fighting (Hare, 1984), whereas in children we can observe the development
of the disorder before it has had an opportunity to destroy its host (Lynam, 1997,
pg. 434).
Multivariate genetic models were also fit to all fourteen of the CPS-RE
subscales in both twins simultaneously to confirm the validity of the same two-factor
solution found in the EFA (in one half of the sample) and the CFA (in the other half).
Specifically, correlated and uncorrelated genetic pathway models were fit to the
CPS-RE subscales to confirm the underlying factor structure of psychopathic traits.
75
In common pathway models, general effects may be mediated through one or more
underlying latent factors that represent the variance shared among all measures.
Analyses revealed that the two-factor solution was superior to a one-factor solution,
and provided the most interpretable solution compared to a three-factor model, but
was not better fitting than the three-factor model. Thus based on phenotypic CFA
results, a two-factor solution was accepted to be the most parsimonious and best fit
to the data.
Psychopathic Personality Traits and Externalizing, Internalizing Behavioral
Problems
Callous/Disinhibited and Charming/Manipulative traits as well as total
psychopathy were significantly associated with externalizing behavioral problems.
As expected, total psychopathy was significantly correlated to reactive and proactive
aggression CD, ODD, and to ADHD, and weakly correlated with symptoms of
anxiety, which respectively aids to establish convergent and discriminant validity of
the construct.
In further support of convergent validity of the emergent psychopathic
personality dimensions, significant relationships also emerged between the two
psychopathic personality dimensions and other externalizing behaviors. Specifically,
the Callous/Disinhibited factor was significantly related to symptoms of ADHD,
reactive and proactive aggression, CD and ODD. The Callous/Disinhibited factor
embodies traits that consist of a general overactivity and behavioral dyscontrol as
76
well as traits that pertain to impulsivity and sensation seeking, therefore it is not
surprising that this dimension correlates strongly with symptoms of ADHD, ODD,
and reactive aggression. Previous studies have also found moderate to strong
relationships between psychopathic traits and CD, ODD, in children (Lynam, 1997,
1998; Krueger et al., 2003; Blonigen et al., 2005, 2006). Moreover, reactive
aggression has been previously defined and characterized as being impulsive, and
unplanned (Flight & Forth, 2007), which coincide with traits that pertain to Factor 1
(Callous/Disinhibited). Previous research has also demonstrated a significant, but
moderate relationship between CD and psychopathic traits (Frick et al., 1994; 2000;
2003). Although present results indicated that the relationships between
psychopathic traits and CD, ODD, and ADHD were significant, they were only
moderate, which may indicate that psychopathic traits are fundamentally different
from other behavioral disorders including CD, ODD, and ADHD (Frick et al., 1994;
Lynam, 1997; Hare, Hart & Harpur, 1991). Moreover, the Callous/Disinhibited
factor was only weakly correlated to symptoms of anxiety, which helps demonstrate
the discriminant validity of the factor.
Additionally, tests of difference in correlations revealed that the relationship
between both Factors 1 and 2 with externalizing behaviors was significantly greater
than the relationships between Factor 1 and 2 with internalizing behaviors, which
suggest that psychopathic traits are correlated more strongly with externalizing
behaviors than with internalizing behaviors. Although the relationship between
Factors 1 and 2 with internalizing behaviors was significantly weaker than the
77
association between Factors 1 and 2 with externalizing problems, they were still
significantly related. This may be an effect of caregiver report bias, which may lead
to a general overestimation and failure to distinguish externalizing and internalizing
behavioral problems in twins with more behavior problems of some kind. It may
also be due to the fact that many of the measures that assess psychopathic personality
traits including, the CPS-RE are comprised of a combination of impulsive, callous,
unemotional and manipulative items (within the two dimensions or factors),
therefore making it difficult to distinguish the relationship between psychopathic
traits and anxiety (Lilienfeld, 1994). Still, the relationship between Factor 2
(Charming/Manipulative) traits was significantly weaker than the relationship
between Factor 1 (Callous/Disinhibited) and anxiety, which is in line with previous
literature that has demonstrated that anxiety is positively associated with traits
related to impulsivity and behavioral dyscontrol traits (Blonigen et al., 2005; 2006;
Harpur et al., 1989). Weak relationships between psychopathic traits and
internalizing behaviors have also been previously demonstrated in children (Lynam,
1997, 1998; Krueger et al., 2003; Frick et al., 1999; Blonigen et al., 2006).
Although the two psychopathic personality factors identified are moderately
correlated (r=.43, p<.01), they each demonstrated a distinct pattern of relationships
with other measures of externalizing and internalizing behaviors. Furthermore,
example cases suggested that the comparatively more psychopathic-like children
exhibit traits that encompass features from both Factors 1 (Callous/Disinhibited) and
factor 2 (Charming/Manipulative). The caregiver endorsements also suggest that
78
children as young as 9 and 10 display some of the core features of psychopathy
including, parasitic orientation, manipulativeness, glibness, taking advantage or
charming others for personal gain, and little or no regard for others. Previous studies
have suggested that these core features of psychopathy remain stable over time,
while the more antisocial, impulsive features of psychopathy decline with age (Hare,
1991; Blonigen et al., 2006), which resembles the successful psychopath
(Cleckley, 1941; Hare, 1993). Caregiver reports also suggest that psychopathic-
like children also show symptoms of CD, ADHD, and ODD, which help support the
notion that psychopathic traits are related to other externalizing behaviors (at a
young age). To lend further support to the necessity of a two-factor structure of
psychopathic traits in children, caregiver reports also indicate higher endorsements
for the factor 2 (Charming/Manipulative) items than for factor 1
(Callous/Disinhibited) (see Table A2 of item endorsements provided in Appendix).
Furthermore, caregiver endorsements of the CPS-RE items also suggested that in
some instances (cases 2 and 3 of Table 7), children who display comparatively
higher psychopathic-like traits may necessarily exhibit extremely high symptom
counts of other externalizing problems including CD, and ODD, which may suggest
that kids high on psychopathic traits may be somewhat different from those who
exhibit high symptoms of CD or ODD alone (Lynam, 1997; Frick et al., 1994).
Comprehensive analyses on psychopathic traits and symptoms of ADHD and
CD revealed that there were significant main effects for ADHD and CD on
psychopathic traits, but no significant interaction effect of ADHD*CD on
79
psychopathy (for total psychopathy score as well as for Factor 1 and 2). Thus,
ADHD and CD have an additive effect on psychopathic traits. Moreover, there were
no significant sex*ADHD or sex*CD effects, suggesting that the effects of ADHD
and CD on psychopathic traits were the same across gender. There was however, a
significant main effect for sex, suggesting that psychopathic traits vary across
gender.
In line with the correlation analyses between CPS-RE and other behavior
problems, mean comparisons between groups of children from the high and low
extremes of the CPS distribution indicated that the group comparatively higher on
psychopathic traits displayed significantly higher mean symptoms for all behavioral
measures including CD, ODD, ADHD, and reactive and proactive aggression than
the group comparatively lower on psychopathic traits. Consistent with present
findings, previous studies have also demonstrated that psychopathy prone
individuals have higher mean symptom counts of ADHD (Fung et al., 2005).
Although most research on psychopathy has focused thus far on adults,
evidence maintains that psychopathic traits are related to aggression and other
behavioral problems much earlier in life (Lynam, 2002; Frick et al., 2000; OBrien &
Frick, 1996). Current results provide further evidence that these traits may be
associated early in life. Psychopathic-like traits are seen in children as young as 9
and 10 years old, and these psychopathic traits are significantly related to other
significant behavioral problems that are prevalent in children including CD, ODD,
ADHD, and aggression (Lynam, 1997, 1998; Frick et al., 1999, 2003; Barry et al.,
80
2000). This knowledge may in turn help in the future prediction of aggressive and
antisocial behaviors in adolescents and adults.
There are several benefits to studying psychopathic traits in children.
Examining and understanding the construct and the genetic and environmental
etiology of psychopathy in children may help provide future treatment plans and
prevention of adult psychopathy. Adult psychopaths have been demonstrated to be
intractable to efforts at rehabilitation (Hart et al., 1988; Serin et al., 1990). Moffitt
suggests that one reason for this may be due to the fact that high recidivism rates in
career criminals may be that the entanglement of the individuals to their previously
poor actions and decisions (such as, previous incarcerations, drug and alcohol
addiction) has hindered access to more legitimate opportunities (Moffitt, 1993).
However, with children there is more of an opportunity to intervene early in multiple
areas including in school, with peers, and within the family as well as on an
individual basis before the amassing of negative consequences. In children where
psychopathic personality traits are seen, there is an opportunity to observe the
development of the disorder and intervene in a timely manner (Lynam, 1997).
Genetic and Environmental Influences on Psychopathic Personality Traits
Genetic analyses on the two psychopathic personality dimensions revealed
significant heritable as well as unique environmental effects of psychopathic
personality traits, with significant differences between boys and girls. Specifically, a
heritability estimate of .64 was found for factor 1, (Callous/Disinhibited) in boys,
and .49 in girls, whereas for factor 2 (Charming/Manipulative), .46 of the variance
81
was due to genetic effects in boys and .58 in girls. Consistent with previous
literature, no shared environmental influences were found in psychopathic
personality traits within our child sample (Blonigen et al., 2006; Larsson et al., 2006;
Taylor, 2003).
Previous studies have also found a strong heritability for psychopathic-like
traits in children, which is consistent with the present findings (Viding et al., 2005;
Blonigen et al., 2006; Larsson et al., 2006; Taylor, 2003). However, these previous
studies either did not explore sex differences or did not find any sex differences
within their samples. The present study examined psychopathic personality traits in
young preadolescent girls and boys) and found significant genetic influences on both
Callous/Disinhibited and Charming/Manipulative traits for both sexes. Mean scores
for both factors 1 and 2 as well as for total psychopathy were significantly higher in
boys than in girls. Even though the genetic and environmental effects for factors 1
and 2 appear to be similar in both boys and girls, sexes could not be constrained to
be equal in the model without a significant loss in fit. Thus, boys and girls may
exhibit somewhat distinct etiologies for psychopathic personality traits.
Genetic and environmental correlations between the two psychopathic
personality dimensions were also investigated in this sample. Bivariate genetic
analyses revealed significant genetic as well as unique environmental correlations
between the two CPS-RE factors in both boys and girls. The genetic correlation
between Factors 1 and 2 was r
g
= .48 in boys and r
g
= .57 in girls. The unique
environmental correlations between factors 1 and 2 were r
e
= .41 in boys and r
e
= .30
82
in girls. Two traits may be related because they share common genetic or common
environmental influences. The current study demonstrated that the relationship
between the Callous/Disinhibited and Charming/Manipulative traits in both boys and
girls were mediated by both genetic and environmental factors common to both
traits. Nonetheless, both the genetic correlation and the unique environmental
correlation were significantly less than 1.0, indicating less than complete overlap in
the etiologies of Callous/Disinhibited and Charming/Manipulative traits. These
results give further support to the notion of a two-factor structure underlying
psychopathic traits in children. These results imply that the construct of psychopathy
is in children is underlaid by at least two related but distinct factors. Present findings
provide support for the function of having at least two distinct factors underlying
psychopathic traits in children.
Genetic and Environmental Covariation Between Psychopathic Personality
Traits and Externalizing Behavioral Problems
To examine the genetic and environmental co-variation between
psychopathic traits and other behavioral problems, bivariate Cholesky decomposition
models were employed. Bivariate analyses revealed significant genetic and unique
environmental correlations between both psychopathic personality factors and total
psychopathy with aggression, CD, ODD, and ADHD indicating that there is genetic
overlap between psychopathic traits and other behavioral problems in early
childhood. A significant genetic overlap was not found for the relationship between
psychopathic traits (either Factor 1 or 2) with proactive aggression. This may be due
83
to the fact that reactive and proactive aggressions are moderate to strongly correlated
and caregivers may have difficulty in deciphering the questions from one another.
Additionally, when the sexes were constrained to be equal in our model, there was
no significant loss in fit to the model, except for the relationship between
psychopathic traits and CD. Thus, we did not find any quantitative sex differences in
the covariation among psychopathic personality traits and aggression, ODD, ADHD.
Even though the means for boys were significantly greater than girls for all
externalizing behavioral measures including, ADHD, ODD, and reactive and
proactive aggression, the genetic and environmental etiologies illustrate a similar
pattern for boys and girls. Previous studies that have examined antisocial behavior
have found that the genetic and environmental influences on adolescent antisocial
behavior are similar for boys and girls (Baker et al., 1989; Jacobson et al., 2002).
There were no significant shared environmental influences found for any of the
relationships between psychopathic traits and other externalizing behavioral
problems (Larsson et al., 2006; Blonigen et al., 2006). However, previous studies
have found shared environmental influences on aggressive and antisocial behaviors
(see Baker et al., 2008), but not for personality traits (Finkel, 1997). These results
fall in line with previous literature that has examined the covariation between
psychopathic traits and antisocial behaviors and have found significant genetic and
unique environmental overlap between psychopathic traits and externalizing
behaviors (Blonigen et al., 2005; Larsson et al., 2006). In addition, previous studies
have also demonstrated that shared environmental influences on psychopathic
84
personality traits have been insignificant (Blonigen et al., 2005; Larsson et al., 2006).
Knowledge and understanding of the link between psychopathic traits with
other behavioral problems including, aggression, CD, ODD, and ADHD, may greatly
aid in the prediction of future antisocial and violent behavior in adult offenders
(Salekin et al., 1996). Research has demonstrated that psychopathic offenders were
approximately five times more likely than non-psychopathic offenders to engage in
violent recidivistic behaviors (Serin and Amos, 1995). It is important to examine
and understand the co-variation among behavior problems because it will aid in
developing specific treatment/prevention plans. Individuals who exhibit one type of
behavioral problem are likely to have different treatment responses and prognoses
compared to individuals with multiple behavior problems.
Strengths and Limitations
Several factors must be considered when interpreting these results. First, the
fact that these analyses were based on data obtained in a general population may be
considered a possible limitation. The use of school-based community samples and
the requirement that families be willing to come to the laboratory for a full
assessment battery may have led to an under sampling of the most severely affected
children who display psychopathic like traits. Second, the narrow age range between
9 and 10 could also be considered a limitation. However, targeting pre-adolescent
children may be beneficial when studying early behavior problems. Third, the study
used only caregiver reports of the twins psychopathic-like behavior, whereas the use
of other raters including self-reports might yield different results. Child self-reports
85
or possibly teacher reports of the twins behaviors at school might also be used in a
multi-rater approach in future studies to help corroborate the observed and reported
psychopathic personality behaviors. However, studies that employ childrens self-
reports could be affected by social desirability, poor reading skills and a general lack
of comprehension, which may occur in children (Lyons, 1995). Self-report measures
have also been found to be highly influenced by question content and wording, and
thus might not provide the best assessments (Schwarz, 1999). Caregivers may be
able to comprehend difficult and complex behaviors and constructs such as
psychopathy better than children (Lyons, 1995; Lynam, 1997). In addition, slight
overlap in the CPS-RE items with the items from the DISC-IV (e.g. ODD) should
also be taken into consideration when interpreting the results from the current
analyses.
Fourth, the examination of psychopathic traits in children could also be
questioned. It has been argued that it might not be developmentally appropriate to
diagnose this syndrome in children and adolescents (Vincent & Hart, 2002), however
there is evidence that psychopathic personality traits are developmental and can be
found in child samples (Lynam, 1997; Viding et al., 2005). Prospective longitudinal
data are needed to establish the predictive utility of psychopathic traits from
childhood to later ages, and the stability and change in these traits over time.
Finally, one may also question the occurrence of low DZ male correlations in
our sample. This might be due to some bias in the mother reports of the CPS-RE or
perhaps due to non-additive or dominant genetic effects. However, comprehensive
86
model fitting analyses of the data revealed that an AE model fit the data better than a
model taking dominance effects into account. Even though the low DZ male
correlation might suggest dominant genetic effects at work in the boys, the moderate
resemblance among DZ opposite sex pairs suggests otherwise.
Assumptions in Twin Studies
The results stemming from this study need to be taken within the context of
basic assumptions underlying twin studies. First, the equal environments assumption
(EEA) assumes that the environment is the same for both identical (MZ) and
fraternal (DZ) twins. If this assumption is violated, higher correlations among MZ
twins may be due to environmental effects rather than genetic influences, which may
lead to a higher estimate of heritability (Plomin, 1991). In addition, MZ twins have
been perceived or expected to be more similar by peers, teachers and parents, which
may effect or influence their trait similarities. Previous research that has examined
EEA has demonstrated that the assumption is justified and does not affect childhood
psychopathology including CD, ODD, and ADHD and aggression (Jacobson et al.,
2002; Cronk et al., 2002).
Second, twins and singletons are generally thought to be similar on most
traits, including antisocial and aggressive behaviors. The twins in the current sample
are no more or less disordered than any other population in terms of conduct or
oppositional defiant disorders (Baker, Barton, Raine, Fowler, 2006). Twins are
however, different from singletons with regard to delayed language development
(Rutter et al., 1991) and higher rates of birth complications including low birth
87
weight (Plomin et al., 2001); however, evidence suggests that birth complications
may lead to later antisocial and aggressive behaviors (Raine, 2002a), but they require
the presence of adverse environmental factors including poor parenting and maternal
rejection (Raine et al., 1997). Finally, it is assumed that the parent generation
randomly mates. Assortative mating may lead to higher DZ twin correlations, which
may underestimate heritability and overestimate or inflate shared environmental
influences; however, assortative mating has not been found to influence personality
traits (Maes, 2002).
Future Directions
The present study examined caregiver reports of psychopathic traits with
other behavioral problems. Future studies stemming from this sample could
investigate the underlying factor structure and the genetic and environmental
influences of psychopathic traits as well as the relationship between psychopathic
traits and other behavioral problems using the twins self-report measures of their
externalizing and psychopathic-like behaviors. The present study primarily
examined the bivariate genetic and environmental co-variation between psychopathic
traits and other behavioral adjustment problems. An extension of investigating the
co-variation at a bivariate level might be to further explore the relationship between
psychopathic traits and other behavioral problems employing multivariate genetic
models and analyses to investigate whether the co-variation among psychopathic
personality traits and other behavioral problems including CD, ODD, ADHD and
aggression are mediated by common genetic or common environmental influences.
88
Future studies from this sample might also explore the phenotypic and etiologic
stability of psychopathic like traits in youths on a longitudinal level in order to
examine the stability of genetic and environmental influences over development.
There is little research on the temporal continuity/stability and change of child and
adolescent psychopathic traits over time. Additionally, an investigation of the factor
structure over time to examine continuity and/or developmental change may also be
a significant addition to the existing literature. A final future direction might be to
investigate the predictive utility of early measures in childhood to later antisocial
behaviors.
Conclusions
This study provides one of the first reports from a large twin study of
childhood psychopathic personality traits examined in 9 and 10 year old twin boys
and girls. Results demonstrated that an underlying two-factor solution for
psychopathic personality traits was the most optimal solution within this
preadolescent community sample. Findings also demonstrated that these
psychopathic traits were under genetic influence in both boys and girls. Consistent
with the literature, no significant shared environmental influences were found for
either factor.
Results also demonstrated that two psychopathic personality factors were
more strongly correlated with other externalizing behavioral problems than
internalizing behavioral problems. Significant genetic overlap was also found
between psychopathic traits and other externalizing behaviors including, CD, ODD,
89
ADHD, and reactive aggression. However, these relationships are not entirely under
genetic influences, unique environmental effects were also found to contribute
significantly to the covariation between psychopathic traits and externalizing
behaviors. These results provide further evidence that antisocial behaviors develop
early in childhood and intervention methods should be tailored accordingly and
possibly start during the early school years.
90
References
Akaike, A. C. (1987). Factor analysis and AIC. Psychometrika, 52, 317-332.
Baker LA, Barton M, Lozano DI, Raine A, Fowler JH. (2006). The Southern
California twin register at the University of Southern California: II. Twin
Research and Human Genetics 9, 933-40.
Bentler, P.M. & Bonett, D.G. (1980). Significance tests and goodness of fit in the
analysis of covariance structures. Psychological Bulletin, 88, 588-606.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological
Bulletin, 107, 238-246.
Bentler, P. M. (2002). EQS 6 Program Manual. Encino, CA: Multivariate
Software, Inc.
Blonigen DM, Hicks BM, Krueger RF, Patrick CJ, Iacono WG. 2005. Psychopathic
personality traits: heritability and genetic overlap with internalizing and
externalizing psychopathology. Psychological Medicine 35: 637-48.
Blonigen, D. M., Hicks, B. M., Krueger, R. F., Patrick, C. J., & Iacono, W. G.
(2006). Continuity and change in psychopathic traits as measured via
normal-range personality: A longitudinal-biometric study. Journal of
Abnormal Psychology, 115, 85-95.
Brandt, J. R., Kennedy, W. A., Patrick, C. J., & Curtin, J. J. (1997). Assessment of
psychopathy in a population of incarcerated adolescent offenders.
Psychological Assessment, 9, 429-435.
Blair, R. J. R. (2001). Neurocognitive models of aggression, the antisocial
personality disorders, and psychopathy. Journal of Neurology,
Neurosurgery, and Psychiatry, 71, 727-731.
Burt, A. S., Krueger, R. F., McGue, M., & Iacono, W. (2001). Sources of
covariation among attention-deficit/hyperactivity disorder, oppositional
defiant disorder, and conduct disorder: The importance of shared
environment. Journal of Abnormal Psychology, 110, 516-525.
Campbell L, Kashy D. 2002. Estimating actor, partner, and interaction effects for
dyadic data using PROC MIXED and HLM: a user-friendly guide. Personal
Relationships 9: 327-42.
91
Cleckley H. 1941/1976. The mask of sanity. St. Louis, MO: Mosby
Cronk, N., Slutske, W. S., Madden, P. A., Bucholz, K. K., Reich, W. & Heath, A.
(2002). Emotional and behavioral problems among female twins: an
evaluation of the equal environments assumption. Journal of American
Academy of Child and Adolescent Psychiatry, 41, 829-837.
Dawson M, Schell A, Filion D. 2000. The electrodermal system. In Handbook of
Psychophysiology, ed. J Cacioppo, L Tassinary, G Berntson: Cambridge
University Press
Cooke, D. J., & Michie, C. (2001). Refining the construct of psychopathy: Towards
a hierarchical model. Psychological Assessment, 13, 171-188.
Cruise, K. R., Colwell, L. H., Lyons Jr., P. M. & Baker, M. D. (2003). Prototypical
analysis of adolescent psychopathy: Investigating the juvenile justice
perspective. Behavioral Sciences and the Law, 21, 829-846.
Finkel, D., & McGue, M. (1997). Sex differences and nonadditivity in heritability
of the multidimensional personality questionnaire scales. Journal of
Personality and Social Psychology, 72, 929-938.
Flight, J .I. & Forth, A. E. (2007). Instrumentally violent youths: The roles of
psychopathic traits, empathy, and attachment. Criminal Justice and
Behavior, 34, 739-751.
Frick, P. J., Kimonis, E. R., Dandreaux, D. M. & Farell, J. M. (2003). The 4 year
stability of psychopathic traits in non-referred youth. Behavioral Science and
the Law, 21, 713-36.
Frick, P. J., Lahey, B. B., Loeber, R., Tannenbaum, L., Van Horn, Y., Christ, M. A.
G., Hart, E. A. & Hanson, K. (1993). Oppositional defiant disorder and
conduct disorder: A meta-analytic review of factor analyses and cross-
validation in a clinic sample. Clinical Psychology Review, 13, 319-340.
Gacono, C. (2000). The clinical and forensic assessment of psychopathy: a
practitioners guide. Mahwah, NJ: Lawrence Erlbaum.
Gorsuch, R.L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum
Associates.
Hare, R. D. (1991a). The Hare Psychopathy Checklist-Revised. Toronto, Canada:
Multi-Health Systems.
92
Hare, R. D. (1991). The revised psychopathy checklist. Toronto: Multi-Health
Systems.
Hare, R. D., Forth, A. E. & Strachan, K. E. (1992). Psychopathy and crime across
the life-span. In: R. Peters, R. J. McMahon, & V. L. Quinsey (Eds.),
Aggression and violence throughout the life span (pp. 285-300). Newbury
Park, CA: Sage.
Hare, R. D. (1993). Without conscience. New York: The Guilford Press
Harpur, T. J., Hakstian, A. R., & Hare, R. D. (1988). Factor structure of the
Psychopathy Checklist. Journal of Consulting and Clinical Psychology, 56,
741-747.
Harpur T, Hakstian A, Hare R. (1989). Two-factor conceptualization of
psychopathy: construct validity and assessment implications. Psychological
Assessment 1: 6-17
Heath, A. C., Madden, P. A. & Martin, N. G. (1998). Assessing the
effects of cooperation bias and attrition in behavioral genetic
research using data-weighting. Behavior Genetics 28(6), 415-27.
Hollingshead, A. G. (1979). Hollingshead Four-Factor Index of Socio-
Economic Status New Haven: Yale University, CT.
Jacobson, K. C., Prescott, C. A. & Kendler, K. (2002). Sex differences in the genetic
and environmental influences on the development of antisocial behavior.
Developmental Psychology, 14, 395-416.
Kotler, J. S., & McMahon, R. J. (2005). Child psychopathy: Theories,
measurement, and relations with the development and persistence of conduct
problems. Clinical Child and Family Psychology Review, 8, 291-325.
Krueger, R. F., Moffitt, T. E., Caspi, A., Bleske, A. & Silva, P. A. (1998).
Assortative mating for antisocial behavior: Developmental and
methodological implications. Behavior Genetics, 28, 173-186.
Larsson, H., Andershed, H., & Lichtenstein, P. (2006). A genetic factor explains
most of the variation in the psychopathic personality. Journal of Abnormal
Psychology, 115, 221-230.
93
Larsson, H., Tuvblad, C., Rijsdijk, F. V., Andershed, H., Grann, M., & Lichtenstein,
P. (2006). A common genetic factor explains the association between
psychopathic personality and antisocial behavior. Psychological Medicine,
1-12.
Lilienfeld, S. O. (1994). Conceptual problems in the assessment of psychopathy.
Clinical Psychology Review, 14, 17-38.
Lilienfeld, S. O. & Andrews, B. P. (1996). Development and preliminary
validation of a self-report measure of psychopathic personality traits in
noncriminal populations. Journal of Personality Assessment, 66, 488-524.
Lynam, D. R. (1996). The early identification of chronic offenders: Who is the
fledgling psychopath? Psychological Bulletin, 120, 209-234.
Lynam, D. R. (1997). Pursuing the psychopath: Capturing the fledgling psychopath
in a nomological net. Journal of Abnormal Psychology, 106, 425-438.
Lynam, D. R. (1998). Early identification of the fledgling psychopath: Locating the
psychopathic child in the current nomenclature. Journal of Abnormal
Psychology, 107, 566-575.
Lynam, D. R., & Gudonis, L. (2005). The development of psychopathy. Annual
Review of Clinical Psychology, 381-407.
Lyons, M. J., True, W. R., Eisen, S. A., & Goldberg, J. (1995). Differential
heritability of adult and juvenile antisocial traits. Archives of General
Psychiatry, 52, 906-915.
Maes, H. H. M., Neale, M. C., Kendler, K., Hewitt, J. K., Silberg, J. L., Foley, D. L.,
Meyer, J. M., Rutter, M., Simonoff. E., Pickles, A. & Eaves, L. J. (1998).
Assortative mating for major psychiatric diagnosis in two population-based
samples. Psychological Medicine, 28, 1389-1401.
Moffitt, T. E. (1993a). Adolescence-limited and life-course-persistent antisocial
behavior: a developmental taxonomy. Psychological Review, 100, 674-701.
Moffitt, T. E. (1993b). The neuropsychology of conduct disorder. Developmental
Psychology, 5, 135-151.
Neale, M. C. (1997). Mx: Statistical modeling. Richmond, VA: Department of
Psychiatry, Medical College of Virginia.
94
Neale, M. C. & Cardon, L. R. (1992). Methodology for genetic studies of twins and
families. Dordrecht, The Netherlands: Kluwer Academic Publications.
Plomin, R., DeFries, J. C., McClearn, G. E. & McGuffin, P. (2001). Behavioral
Genetics. United States of America: Worth Publisher.
Porter, S., & Woodworth, M. (2006). Psychopathy and aggression. In Christopher
Patrick (Ed.), Handbook of Psychopathy (481-494). New York: The Guilford
Press.
Posthuma, D., Beem, A. L., de Geus, E. J. C., van Baal, C. M., von Hjelmborg, J. B.,
Iachine, I. & Boomsma, D. I. (2003). Theory and practice in quantitative
genetics. Twin Research and Human Genetics, 6, 361-376.
Raine, A. (2002a). The role of prefrontal deficits, low autonomic arousal, and early
health factors in the development of antisocial and aggressive behavior in
children. Journal of Child Psychology and Psychiatry, 43, 417-434.
Raine, A. (2002b). The role of prefrontal deficits, low autonomic arousal, and early
health factors in the development of antisocial and aggressive behavior in
children. Journal of Child Psychology and Psychiatry, 43, 417-434.
Raine, A., Brennan, P. A. & Mednick, S. A. (1997). Interaction between birth
complications and early maternal rejection in predisposing individuals to
adult violence: specificity to serious, early-onset violence. American Journal
of Psychiatry, 154, 1265-1271.
Rutter, M. & Redshaw, J. (1991). Growing up as a twin: twin-singleton differences
in psychological development. Journal of Child Psychology and Psychiatry,
32, 885-895.
Salekin, R. T. (2006). Psychopathy in children and adolescents: Key issues in
conceptualization and assessment. In Christopher Patrick (Ed.), Handbook of
Psychopathy. (pgs. 389-414). New York: The Guilford Press.
Salekin, R. T. & Frick, P. J. (2005). Psychopathy in children and adolescents: The
need for a developmental perspective. Journal of Abnormal Child
Psychology, 33, 403-409.
Salekin, R. T., Ziegler, T. A., Larrea, M. A., Anthony, V. L., & Bennett, A. D.
(2003). Predicting psychopathy with two Millon Adolescent psychopathy
scales. The importance of egocentric and callous traits. Journal of
Personality Assessment, 80, 154-163.
95
Salekin, R. T., Rogers, R. & Machin, D. (2001). Psychopathy in youth: Pursuing
diagnostic clarity. Journal of Youth and Adolescence, 30, 173-195.
Schwarz, N. (1999). Self reports: How the questions shape the answers. American
Psychologist, 54, 93-105.
Singer, J. (1998). Using SAS PROC MIXED to fit multilevel models, hierarchical
models, and individual growth models. Journal of Educational and
Behavioral Statistics, 24(4): 323-355
Stevens, J. (1996). Applied multivariate statistics for the social sciences (3rd ed.).
Mahwah, NJ: Lawrence Erlbaum Associates.
Taylor, J., & Lang, A. (2006). Psychopathy and substance use disorders. In
Christopher Patrick (Ed.), Handbook of Psychopathy (495-511). New York:
The Guilford Press.
Viding, E., Blair, R. J. R., Moffitt, T. E. & Plomin, R. (2005). Evidence for
substantial genetic risk for psychopathy in 7-year-olds. Journal of Child
Psychology and Psychiatry, 46, 592-597.
Vincent, G. M., & Hart, S. D. (2002). Psychopathy in children and adolescents:
Implications for the assessment and management of multi-problem youths.
In R. R. Corrado, R. Roesch, S. D. Hart & J. K. Gierowski (Eds.), Multi-
problem violent youth: A foundation for comparative research on needs,
interventions, and outcomes (pp. 150-163). Amsterdam, IOS Press.
Wilcox, R. R. (2003). Applying contemporary statistical techniques. Los Angeles,
Academic Press.
96
Appendix:
Supplementary Analyses, Tables, and Figures
Table A1
ACE Estimates for CPS-RE Factors (for both boys and girls separately)
ACE Effects for CPS-RE Factors 1 & 2
Boys Girls
Callous/disinhibited Charming/manipulative Callous/disinhibited Charming/manipulative
A .64* (.49-.72) .46* (.22-.58) .49* (.21-.65) .58* (.25-.70)
C 0
ns
(0 .11) 0.01
ns
(0-.195) .06
ns
(0 -.30) .04
ns
(0-.34)
E .36* (.28-.47) .53* (.41-.66) .44* (.35-.56) .37* (.29-.48)
* parameter is significant at the .05 level. 95 % confidence intervals are in parentheses.
97
Table A2.
Item endorsements from CPS-RE caregiver reports for each factor by sex
Item % endorsement
Factor 1-Callous/Disinhibited
Males Females
Unreliable (α = 43)
1. When he/she starts working on something does he stick with it?
(R)
20.7 13.5
2. Does he/she often have to break his promises? 14.4 9.5
3. Can people count on him/her? (R) 3.4 1.6
Poverty of Affect (α = .38)
1. Do his/her feelings change often and quickly? 25.4 26.5
2. Are his/her moods unpredictable? 21.0 19.7
3. Do his/her feelings come and go quickly? 52.3 49.4
4. Do his/her feelings seem fake? 9.0 13.7
5. Does he/she make close friends with other people? (R) 13.4 7.9
6. Are his/her feelings intense? 64.2 67.3
7. Is he/she a warm and kind person? (R) 2.5 .8
8. Does he feel things very strongly? (R) 26.4 27.7
9. Is he/she open with his/her feelings? (R) 11.2 8.0
10. Is he protective of people who are close to him? (R) 4.2 4.2
Lack of Planning (α = .59)
1. Does he/she set goals for himself and try to reach them? (R) 29.9 18.2
2. Does he/she think about what he wants to do with the rest of his
life? (R)
23.4 18.2
3. Does he/she plan things ahead? (R) 39.6 27.4
Boredom Susceptibility (α = .46)
1. Is he/she easily bored? 44.3 35.2
2. Does he/she need to have things be exciting? 43.9 33.4
3. Does he/she concentrate well on things (R)? 28.4 17.8
4. Does he/she do dangerous or risky things for the fun of it? 4.7 2.1
5. Does he/she stay away from scary things and places? (R) 24.7 18.8
Impulsivity (α = .52)
1. Does he/she think before doing or saying something? (R) 32.6 23.6
2. Does he/she think about his actions and behavior? (R) 20.1 14.0
3. Does he/she have a hard time waiting for things he wants? 48.1 34.9
Behavioral Dyscontrol (α = .75)
1. Does he/she get mad or irritated over little things? 45.1 42.3
2. Does he/she lose his temper easily? 33.8 26.5
3. Is he/she easily frustrated? 39.7 39.5
Callousness (α = .47)
1. Is he/she mean to other people? 7.7 5.5
2. Does he tease and pick on other people? 15.9 9.0
3. Is he kind and thoughtful of other people? (R) 3.2 1.6
4. Does he/she try not to hurt other people's feelings? (R) 14.2 9.8
5. Is he/she able to tell how other people feel? (R) 16.5 12.3
98
Table A2 (Table A2, continued)
Item endorsements from CPS-RE caregiver reports for each factor by sex
Item % endorsement
Factor 2-Charming/Manipulative
Males Females
Manipulative (α = .53)
1. Does he/she try to get others to do what he wants by trying to get
on their good side?
36.8 34.0
2. Does he/she try to act charming or likable in order to get his way? 40.6 35.2
3. Does he/she try to take advantage of other people? 6.8 5.9
Failure to Accept Responsibility (α = .33)
1. Does he/she try to blame other people for things he has done? 30.1 26.7
2. Does he/she think he gets blamed for things he did not do? 61.5 56.3
3. When he/she gets in trouble, can he talk his way out of it? 47.4 40.5
Glibness (α = .61)
1. Does he/she show off to get people to pay attention to him? 28.9 19.3
2. Does he/she try to be the center of attention? 34.8 29.2
3. Does he/she tell stories to make himself look good? 16.7 11.2
4. Is he/she talkative? 69.7 71.9
5. Is he/she shy? (R) 53.9 51.8
Parasitic Lifestyle (α = .37)
1. Does he take a lot and not give much in return? 11.7 10.1
2. Does he give or share things? (R) 3.0 4.6
3. Does he usually return what he borrows? (R) 7.5 6.6
4. Does he try to see how much he can get away with? 48.5 38.1
Untruthful (α = .56)
1. Is he a good liar? 12.1 10.9
2. Will he usually tell a lie if he thinks he can get away with it? 31.4 26.8
3. Is he open and honest? (R) 7.6 3.2
4. Do people usually believe him when he/she tells a lie? 38.5 38.2
5. Can he be trusted? (R) 3.2 1.9
Table A2 (Table A2, continued)
Scales dropped from analyses due to poor loadings
Item % endorsement
Males Females
Grandiosity
1. Is he very sure of himself? 75.5 76.6
2. Does he have a low opinion or think badly of himself? (R) 88.0 87.4
3. Does he need to have people tell you that he is doing well or
ok? (R)
47.5 45.0
Lack of Guilt
1. Does it bother him/her when he/she does something wrong?
(R)
10.2 9.0
2. Does he usually feel bad or guilty after doing something
wrong? (R)
7.8 8.5
3. Does he wish that he could take back many things he has done?
(R)
58.1 63.2
99
Table A3.
Correlations for Psychopathic Traits and Other Behavioral Problems by Sex
CD ADHD ODD Reactive
Proactive GA
Total Psychopathy .46** .58** .51** .55** .51** .18**
Callous/Disinhibited .37** .57** .44** .51** .41** .22**
Boys
Charming/Manipulative .44** .40** .48** .46** .46** .15**
Total Psychopathy .43** .57** .39** .55** .52** .25**
Callous/Disinhibited .32** .58** .33** .50** .44** .31**
Girls
Charming/Manipulative .46** .42** .43** .52** .48** .15**
**. p<.01. Note: Sample size ranges from N=589-592 for boys; and N=612 to 624
for girls.
100
Table A4. Multiple Regression analyses of Psychopathic Traits With Other
Behavioral Problems (males, females)
Males Females
Internalizing
R
2
Change
.03* .06*
β Anxiety
-.07* .39
Externalizing
R
2
Change
.52* .52*
β CD
.11* .18*
β ADHD
.40* .37*
β ODD
.09* -.03
β Reactive
.20* .24*
β Proactive
.22* .21*
*p<.05
101
Figure A1. Illustration of Genetic and Environmental Correlations
Correlated factor model: circles indicate variance components (a
2
, c
2
, e
2
) and
rectangles indicate observed phenotypes, in this case: Psychopathic Personality
Traits (Callous/Disinhibited or Charming/Manipulative) and CD: a
2
= additive
genetic component, c
2
= shared environmental component, e
2
= non-shared
environmental component. r
g
= genetic correlation, r
c
= shared environmental
correlation, r
e
= non-shared environmental correlation. Note: for simplicity,
only one twin in a pair is shown.
A E
Psychopathic Traits
r g
a
2
C A E
CD
C
r c
r e
c
2
e
2
a
2
e
2
c
2
102
Figure A2.
Illustration of alternative factor models of psychopathic traits (adapted to the
CPS-RE subscales)
Hare Two-Factor Model
Social Deviance
Impulsive/Antisocial
Interpersonal/Affective
Callousness
Poverty of Affect
Lack of guilt
Glib
Grandiose
Untruthful
Manipulative
Fail to Accept Responsibility
Parasitic Lifestyle
Behavioral Dyscontrol
Lack of Goals
Impulsive
Boredom Susceptibility
Unreliable
103
Figure A2 (Figure A2, continued from previous page)
Illustration of alternative factor models of psychopathic traits (adapted to the CPS-
RE subscales)
Three-Factor Model
Grandiose/Manipulative
(Interpersonal)
Callous/Unemotional
(Affective)
Impulsive/Irresponsible
(Behavioral)
Glib
Grandiose
Untruthful
Manipulative
Lack of Guilt
Poverty of Affect
Callousness
Fail to Accept Responsibility
Impulsive
Parasitic Lifestyle
Lack of Goals
Unreliable/Irresponsible
Boredom Susceptibility
104
Table A5.
Oppositional Defiant Disorder Symptoms (ODD Symptoms, DISC-IV)
Oppositional Defiant Disorder Symptoms
1 Lost temper in past year
2 Argued/talked back to caretakers in past year
3 Done things on purpose that caretakers said not to do in past year
4 Refused to do things that caretakers said to do in past year
5 Done things just to annoy people/make them mad in past year
6 Blamed others for their mistakes or bad behavior in past year
7 Grouchy/easily annoyed in past year
8 Seemed mad at people or things in past year
9 Felt things were unfair and got angry in past year
10 Done mean things to people on purpose in past year
11 Gotten even with people in past year by hurting them/messing up their
things/telling lies about them
12 Swore/used dirty language in past year
105
Table A6.
Conduct Disorder Symptoms (CD Symptoms, DISC-IV)
Conduct Disorder Symptoms
1 Ever stolen money or other things from people he lives with
2 Ever shoplifted
3 Ever stolen from someone when they weren't around or weren't looking
4 Faked a name on a check/used someone else's credit card without permission
5 Ever snatched purse or jewelry
6 Ever held someone up or attacked someone to steal from them
7 Ever threatened someone in order to steal from them
8 Ever gotten into trouble for staying out at night more than 2 hours past the
time supposed to return
9 Ever run away from home overnight
10 Ever lied to get money or something else
11 Ever lied to not pay back money or to get out of something important he was
supposed to do
12 Ever skipped school
13 Ever broken into a house/building/car
14 Broken/damaged someone else's things on purpose
15 Broken something/messed up a place on purpose
16 Ever started a fire without permission
17 Ever been physically cruel to an animal
18 Ever bullied someone smaller who wouldn't fight back"
19 Ever been in a physical fight in which someone was hurt or could have been
hurt
20 Ever tried to hurt someone badly/been physically cruel to someone"
21 Ever hurt someone with a weapon (like a bat, brick, broken bottle, knife, gun
22 Ever been expelled from school for misbehavior
23 Ever been suspended from school for misbehavior
24 Ever had an in-school suspension (went to school but not allowed to attend
usual classes
25 Ever been in trouble with the police
106
Table A7.
Attention Deficit Hyperactivity Disorder Symptoms (ADHD, DISC-IV)
Attention Deficit Hyperactivity Disorder Symptoms
1 Trouble keeping mind on task for more than a short time
2 Often tried to avoid doing things that required sustained attention
3 Often disliked doing things that required sustained attention
4 Couldn't keep mind on one thing when other things were going on
5 Disorganized
6 Trouble finishing homework
7 Often lost things
8 Forgot what they were supposed to do
9 Often made a lot of mistakes because it's hard for him to do things carefully
10 Often didn't listen when people were speaking to him
11 Started activities without finishing them
12 Been 'on the go' more active than usual as if 'driven by a motor'
13 Fidgety/ restless
14 Left seat (as in at school, movie, restaurant) when not supposed to
15 Often climbed on things/ran around when not supposed to
16 Often talked a lot more than other children
17 Often made much more noise while playing than other children
18 Often interrupted others while they were talking or busy
19 Often butted in on what others were doing
20 Blurted out answers before hearing whole question
21 Often had trouble waiting for turns (as in line or playing a game)
107
Table A8.
Generalized Anxiety Disorder Symptoms (GAD, DISC-IV)
Generalized Anxiety Disorder Symptoms
1 Often worried before taking test/handing in an important assignment
2 Often seem worried before playing a sport, game, or other activity
3 Worried a lot about small mistakes made on homework, projects, or activities
4 Often worried about being on time
5 Often worried about having an illness
6 Type of person who is tense and finds it very hard to relax
7 Often been worried that they made a mistake or did something the wrong way
8 Been worried that they made a fool of themselves in front of other people
9 Been very concerned with being liked by others
10 Often had headaches
11 Often had stomachaches
12 Other aches and pains
108
Table A9.
Best Fitting Models from Bivariate Cholesky Analyses of Covariation Among Psychopathic Traits and CD,
ADHD, ODD
Overall fit Model difference test
Callous/Disinhibited-CD -2LL AIC BIC df ∆ AIC ∆χ
2
∆ df P
1 Saturated covariance* 6247.80 1509.80 -4476.72 2363
2 Boys ≠ Girls (r
g
and r
c
fix)-Drop
Shared Environmental Covariation
6292.81 1482.81 -4569.71 2405 -26.99 45.01 36 0.14
Charming/Manipulative-CD
1 Saturated covariance* 6176.26 1438.26 -4512.49 2369
2 Boys ≠ Girls (r
g
and r
c
fix)-Drop
Shared Environmental Covariation
6219.38 1409.38 -4606.43 2405 -28.88 43.12 36 0.19
Callous/Disinhibited-ADHD
1 Saturated covariance* 5992.30 1262.30 -4591.64 2365
2 Boys ≠ Girls (r
g
and r
c
fix) 6030.18 1240.18 -4668.95 2395 -22.12 37.88 30 0.15
3 Boys = Girls
6041.04 1233.04 -4692.39 2404 -29.26 48.74 39 0.14
3a Boys = Girls compared to Model # 2 6041.04 1233.04 -4692.39 2404 -7.140 10.86 9 0.28
3b Boys=Girls Drop Shared
Environmental Covariation
Compared to model # 3
6041.39 1227.39 -4701.84 2407 -5.65 0.35 3 0.95
Charming/Manipulative-ADHD
1 Saturated covariance*
2 Boys ≠ Girls (r
g
and r
c
fix) 6309.90 1519.90 -4529.09 2395 -19.19 44.81 32 0.07
3 Boys = Girls
6325.92 1517.92 -4549.95 2404 -21.17 60.83 41 0.02
4 Boys = Girls compared to model # 2 6325.92 1517.92 -4549.95 2404 -1.98 16.02 9 0.07
5 Boys = Girls Drop Shared
Environmental Covariation
6326.52 1512.52 -4559.28 2407 -26.57 61.43 44 0.04
5a Boys = Girls Drop Shared
Environmental Covariation
compared to Model # 3
6326.52 1512.52 -4559.28 2407 -5.40 0.60 3 0.90
*Models comparisons made to saturated covariance model unless otherwise specified. Best-fitting models are in
bold.
109
Table A9 (Table A9, continued)
Model Overall fit Model difference test
-2LL AIC BIC df ∆ AIC ∆χ
2
∆ df P
Callous/Disinhibited-ODD
1 Saturated covariance* 6243.39 1515.39 -4462.88 2364
2 Boys ≠ Girls (r
g
and r
c
fix) 6280.16 1488.16 -4547.16 2396 -27.23 36.77 32 0.26
3 Boys = Girls Full
6290.15 1480.15 -4571.04 2405 -35.23 46.76 41 0.25
3a Boys = Girls Full compared to
Model #2
6290.15 1480.15 -4571.04 2405 -8.01 9.99 9 0.35
Charming/Manipulative-ODD
1 Saturated covariance* 6232.13 1500.13 -4474.93 2366
2 Boys ≠ Girls (r
g
and r
c
fix) 6274.67 1482.67 -4549.91 2396 -17.46 42.54 30 0.06
3 Boys = Girls Full
6286.37 1476.37 -4572.94 2405 -23.76 54.24 39 0.05
3a Boys = Girls Full Compared to
Model #2
6286.37 1476.37 -4572.94 2405 -6.30 11.70 9 0.23
*Models comparisons made to saturated covariance model unless otherwise specified.
Best-fitting models are in bold.
110
Table A9 (Table A9, continued)
Model Overall fit Model difference test
Callous/Disinhibited-
Reactive Aggression
-2LL AIC BIC df ∆ AIC χ
2
∆ df P
1 Saturated model 6174.21 1418.20 -4542.39 2378
2 Boys ≠ Girls (r
g
free)
6210.36 1394.36 -4620.57 2408 -23.84 36.16 30 0.20
3 Boys ≠ Girls (r
c
free)
6220.08 1404.08 -4615.71 2408 -14.13 45.87 30 0.03
4 Boys ≠ Girls (r
g
and r
c
fix) 6211.96 1391.96 -4626.18 2410 -26.24 37.76 32 0.22
5 Boys = Girls
6223.73 1385.73 -4649.17 2419 -32.48 49.52 41 0.17
5a Boys = Girls (compared to #4) 6223.73 1385.73 -4649.17 2419 -6.227 11.77 9 0.23
6 Drop genetic co-variation 6228.17 1388.17 -4650.16 2420 -30.04 53.96 42 0.10
7 Drop shared environmental
co-variation
6226.97 1386.97 -4650.76 2420 -31.24 52.76 42 0.12
Callous/Disinhibited
Proactive Aggression
1 Saturated model 6315.09 1559.09 -4471.95 2378
2 Boys ≠ Girls (r
g
free)
6345.34 1529.34 -4553.07 2408 -29.75 30.25 30 0.45
3 Boys ≠ Girls (r
c
free)
6342.98 1526.98 -4554.26 2408 -32.12 27.88 30 0.58
4 Boys ≠ Girls (r
g
and r
c
fix) 6350.39 1530.39 -4556.97 2406 -28.71 35.29 32 0.31
5 Boys = Girls
6365.28 1527.28 -4578.40 2419 -31.81 50.19 41 0.15
5a Boys=Girls (compared to #4) 6365.28 1527.28 -4578.40 2419 -11.11 14.89 13 0.31
6 Drop genetic co-variation 6369.07 1529.07 -4579.71 2420 -30.02 53.98 42 0.10
7 Drop shared environmental
co-variation
6367.57 1527.57 -4580.46 2420 -31.53 52.47 42 0.13
*Models comparisons made to saturated covariance model unless otherwise specified.
Best-fitting models are in bold.
111
Table A9 (Table A9, continued)
Overall fit Model difference test
Charming/Manipulative-
Reactive Aggression
-2LL AIC BIC df ∆ AIC χ
2
∆ df P
1 Saturated model 6250.45 1494.45 -4504.27 2378
2 Boys ≠ Girls (r
g
free)
6291.55 1475.55 -4579.97 2408 -18.90 41.10 30 0.08
3 Boys ≠ Girls (r
c
free)
6292.97 1476.97 -4579.26 2408 -17.48 42.52 30 0.06
4 Boys ≠ Girls (r
g
and r
c
fix)
6294.83 1474.83 -4584.75 2410 -19.62 44.38 32 0.07
5 Boys = Girls
6304.33 1466.33 -4608.87 2419 -28.11 53.885 41 0.09
5a Boys = Girls (compared to #4) 6304.33 1466.33 -4608.87 2419 -8.49 9.505 9 0.39
6 Drop genetic co-variation 6310.06 1470.06 -4609.21 2420 -24.39 59.61 42 0.04
7 Drop shared environmental co-
variation
6309.08 1469.08 -4609.70 2420 -25.37 58.63 42 0.05
Charming/Manipulative-
Proactive Aggression
1 Saturated model 6297.20 1541.20 -4480.89 2378
2 Boys ≠ Girls (r
g
free)
6332.79 1516.79 -4559.35 2408 -24.407 35.593 30 0.22
3 Boys ≠ Girls (r
c
free)
6330.77 1514.77 -4560.36 2408 -26.43 33.573 30 0.30
4 Boys ≠ Girls (r
g
and r
c
fix)
6333.93 1513.93 -4565.20 2410 -27.27 36.73 32 0.26
5 Boys = Girls
6347.94 1509.94 -4587.07 2419 -31.26 50.74 41 0.14
5a Boys=Girls (Compared to #4) 6347.94 1509.94 -4587.07 2419 -3.993 14.007 9 0.12
6 Drop genetic co-variation 6350.02 1510.02 -4589.23 2420 -31.18 52.82 42 0.12
7 Drop shared environmental co-
variation
6353.63 1513.63 -4587.43 2420 -27.57 56.43 42 0.07
*Models comparisons made to saturated covariance model unless otherwise specified.
Best-fitting models are in bold.
Abstract (if available)
Abstract
Psychopathic personality traits and their genetic and environmental etiology were investigated in 9-10 year old twins. Specifically, the psychometric properties of psychopathic personality traits were examined by exploring the factor structure of these traits, investigating their relationship to other problem behaviors that occur in childhood, and investigating the extent of the genetic and environmental influences on these traits. A community sample of 1219 children (born between 1990 and 1995) from the Southern California Twin Project was employed. Psychopathic personality traits were assessed through caregiver reports of the twins psychopathic behaviors using the Child Psychopathy Scale-Revised Extended (CPS-RE
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Construct validity of psychopathic personality traits in a cohort of young twins
PDF
Heritability of schizotypal traits during adolescence
PDF
Multivariate biometric modeling among multiple traits across different raters in child externalizing behavior studies
PDF
Early temperamental and psychophysiological precursors of adult psychopathic personality
PDF
Biometric models of psychopathic traits in adolescence: a comparison of item-level and sum-score approaches
PDF
Patterns of EEG spectral power in 9-10 year old twins and their relationships with aggressive and nonaggressive antisocial behavior in childhood and adolescence
PDF
Psychophysiological reactivity to stress, psychopathic traits, and developmental trajectories of antisocial behavior: a prospective, longitudinal study
PDF
The development of electrodermal fear conditioning from ages 3 to 8 years and relationships with antisocial behavior at age 8 years
PDF
Heritability and longitudinal stability of impulsivity in adolescence
PDF
Different genetic and environmental structures for the overlap of three antisocial behavior factors with alcohol initiation
PDF
Developmental etiology of reactive and proactive aggression, electrodermal reactivity and their relationships: a twin study
PDF
Biopsychosocial and forensic clinical correlates of schizophrenia and homicide
PDF
Dementia, co-morbid depressive symptoms, and mortality in older adults: a twin study
PDF
Detecting joint interactions between sets of variables in the context of studies with a dichotomous phenotype, with applications to asthma susceptibility involving epigenetics and epistasis
Asset Metadata
Creator
Bezdjian, Serena
(author)
Core Title
Psychometric and biometric evaluation of psychopathic personality traits in 9 and 10 year old twins
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Psychology
Publication Date
07/29/2008
Defense Date
04/29/2008
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
heritability,OAI-PMH Harvest,psychopathic personality traits,twin children
Language
English
Advisor
Baker, Laura A. (
committee chair
), Prescott, Carol A. (
committee member
), Raine, Adrian (
committee member
), Schwartz, David (
committee member
), Trickett, Penelope K. (
committee member
)
Creator Email
bezdjian@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m1442
Unique identifier
UC1117143
Identifier
etd-Bezdjian-20080729 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-92742 (legacy record id),usctheses-m1442 (legacy record id)
Legacy Identifier
etd-Bezdjian-20080729.pdf
Dmrecord
92742
Document Type
Dissertation
Rights
Bezdjian, Serena
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
heritability
psychopathic personality traits
twin children