Elsevier

Journal of Criminal Justice

Volume 43, Issue 4, July–August 2015, Pages 327-336
Journal of Criminal Justice

A multi-wave cross-lagged regression analysis of the Youth Psychopathic Traits Inventory and Self-Reported Offending

https://doi.org/10.1016/j.jcrimjus.2015.04.002Get rights and content

Highlights

  • sought to clarify the psychopathy-offending relationship

  • cross-lagged Youth Psychopathic Traits Inventory (YPI) with offending

  • controlled for demographic, criminological, and antecedent covariates

  • reciprocal relationship between YPI and offending, significant only in males

  • psychopathy predicts offending and offending predicts psychopathy

Abstract

Purpose

This study sought to clarify the relationship between psychopathy and self-reported offending using a measure of psychopathy that did not directly assess criminal behavior; i.e., the Youth Psychopathic Traits Inventory (YPI).

Methods

Cross-lagged correlations between five pairs of waves from the Pathways to Desistance study were used to assess the psychopathy-offending relationship in 1354 (1170 male, 184 female) previously adjudicated delinquents. Structural equation modeling was used to assess these five pairs of waves, controlling for important demographic and criminological covariates and the outcome’s antecedent.

Results

All five zero-order cross-lagged YPI  offending correlations and all five zero-order cross-lagged offending  YPI correlations were significant. All five YPI  offending regressions and all five offending  YPI regressions were significant, after controlling for the demographic/criminological/antecedent variables, although there were no significant differences between the two paths (YPI  offending vs. offending  YPI). All ten cross-lagged regressions were significant in male participants but only three of the regressions were significant in females.

Conclusions

Psychopathy, as measured by the YPI, and self-reported offending are reciprocally related, at least in boys and young men, such that offending is as likely to shape psychopathy as psychopathy is to shape offending.

Introduction

Early factor analyses of the Psychopathy Checklist and its derivatives—the Psychopathy Checklist-Revised (PCL-R: Hare, 2003), the Psychopathy Checklist: Screening Version (PCL: SV: Hart, Cox, & Hare, 1995), and the Psychopathy Checklist: Youth Version (PCL: YV: Forth, Kosson, & Hare, 2003)—identified two primary factors: a personality factor, characterized by the selfish, callous and remorseless use of others, and a behavioral deviance factor, characterized by impulsivity and antisocial behavior (Harpur et al., 1988, Harpur et al., 1989). This two-factor model was eventually replaced by a three-factor model that divided the personality factor into affective and interpersonal subcomponents and retained only the impulsive component of the behavioral deviance factor (Cooke & Michie, 2001). Hare and Neumann (2008) subsequently proposed a four-factor model that included all three components from the three-factor model but added the antisocial behavior component that was left out of the Cooke and Michie (2001) model. This theoretical/conceptual disagreement about whether or not antisocial behavior is a core feature of psychopathy has resulted in an intense and unabated controversy.

There is consensus that psychopathy and antisocial behavior are linked, but there is no consensus as to whether antisocial behavior is a defining characteristic of psychopathy or an entirely separate phenomenon predicted and explained, at least in part, by psychopathy. Over the last several decades psychologists have debated this issue. In their review of the research, including their own extensive work, Skeem and Cooke (2010) concluded that antisocial behavior was nothing more than a secondary manifestation or sequela of psychopathic personality disorder. In contrast, Hare and Neumann (2010) argued that their and others research support the perspective that antisocial behavior is a core feature of psychopathy. They concede, in part, that criminal behavior can be considered a dependent phenomenon. Theoretically, the debate over this issue has involved Cleckley’s (1941/1976) groundbreaking book, The Mask of Sanity, upon which modern-day conceptualizations of psychopathy are based. In his book, Cleckly classified inadequately motivated antisocial behavior as a core feature of psychopathy. Fraud, forgery, adultery, theft, and fighting, often in the absence of a discernable long-term goal, were all identified as symptoms of inadequately motivated antisocial acts. Subsequent research by Hare (1999), which resulted in the original 20-item PCL-R, was based on prison samples. This research identified goal-directed, cold-blooded, and proactive forms of antisocial conduct, including vicious and serial forms of extreme violence, as central to the measurement of psychopathy. More recent research comparing the three- and four-factor PCL-R models, using a variety of statistical techniques such as confirmatory factor analysis and item response theory (IRT), have supported one position or the other but not both (Cooke et al., 2007, Neumann et al., 2007, Vitacco et al., 2005, Walters, 2012). In this article, it will be argued that this controversy very likely will never be resolved empirically no matter how innovative the statistical methodology.

The PCL-R is regarded by some as a “gold standard” measure of psychopathy based largely on the perception that it predicts both general and violent recidivism (Hare, 1996, Hemphill et al., 1998). However, this key form of validity has typically been associated with the anti-social behavioral dimension or domain of the PCL- R (Walters, 2003). In essence, the key and now long standing validity concern of the PCL-R and PCL-YV has been that its key personality dimensions have not been predictive of central measures of criminality. More recently, in a series of studies encompassing ten different samples and nearly 2900 subjects, Walters and colleagues confirmed that the antisocial facet of the PCL consistently predicted general and violent recidivism above and beyond the contributions of the first three facets (interpersonal, affective, and lifestyle), but the first three facets consistently failed to display incremental validity relative to the antisocial facet in predicting these same two outcomes (Walters and Heilbrun, 2010, Walters et al., 2008, Walters et al., 2011). Walters (2012) subsequently reported that the first three facets of the PCL-R, which according to the results of factor and IRT analysis constituted the optimal model for PCL-R psychopathy in a group of Canadian offenders, failed to predict general or violent recidivism above and beyond the contributions of age and criminal history. These large-scale studies suggest that the PCL-R’s ability to predict recidivism may be the result of a predictor-outcome confound created when antisocial and criminal items are included on the PCL-R. In other words, a major validity question with regards to this popular measure of psychopathy is whether its relationship with serious criminality is largely tautological.

Hare and Neumann (2010) have consistently asserted that general antisocial behavior, rather than criminality per se, is the focus of the PCL-R and its derivatives. Yet, three of the five items on Facet 4 (antisocial) of the PCL-R directly reference criminal behavior (i.e., juvenile delinquency, revocation of conditional release, and criminal versatility) and a fourth antisocial item (poor behavioral controls) can potentially be scored from criminal outcomes. In order to address this issue, there have been several studies that have excluded the explicitly criminal indicators from the Psychopathy Checklist (e.g. Corrado et al., & Vincent et al. plus) and found that the remaining antisocial and behavioral indicators remained the strongest predictors of violent criminality and recidivism (Corrado et al., 2004, Marshall et al., 2006, Vincent et al., 2008). Another option would be to use an instrument such as the Youth Psychopathic Traits Inventory (YPI: Andershed, Kerr, Stattin, & Levander, 2002) that does not incorporate antisocial behavior, or at least criminality, as indicators in its definition of psychopathy.

There are several advantages to using the YPI to assess the relationship between psychopathy and crime. First, it utilizes Cooke and Michie’s (2001) three-factor model of psychopathy and does not consider antisocial behavior a core feature of psychopathy. Accordingly, it is less vulnerable to charges of criterion contamination than a measure like the PCL-R/YV when correlated with offending behavior. Second, the YPI is a self-report measure. While self-report measures have obvious and inherent internal validity limitations (e.g., response styles, reading difficulties), they may provide a more reliable estimate of the cognitive and emotional features of psychopathy than a rating procedure based on data from an interview and a review of available records (Lilienfeld & Fowler, 2006). More specifically, the YPI exhibits strong internal consistency and temporal stability (Andershed et al., 2002) and adequate to good convergent (Andershed et al., 2007, Seals et al., 2012), predictive (Chauhan et al., 2014, Skeem and Cauffman, 2003), and construct (Declercq, Markey, Vandist, & Verhaeghe, 2009; Poythress, Dembo, Wareham, & Greenbaum, 2006) validity.

Using data from a large Swedish twin cohort, Forsman, Lichtenstein, Andershed, and Larsson (2010) discerned that the total YPI score measured at age 16–17 predicted subsequent antisocial behavior at age 19–20, and that persistent antisocial behavior from age 8–9 to age 16–17 predicted the total YPI score measured at age 19–20. This is one of the first studies to report a potential reciprocal relationship between psychopathic personality traits and antisocial behavior. These results remain tentative, however, in light of several significant methodological limitations. First, the correlations used to identify the reciprocal relationship were not true cross-lag correlations. The T1 measure of persistent antisocial behavior ran from age 8–9 to age 16–17, whereas the T1 measure of psychopathy was taken at age 16–17. Second, there was a three-year lag between T1 and T2, a fairly long period of time that might allow additional variables to impact on the relationship. Third, other than sex, there were no control variables included in the analyses and there were no antecedent measures of the predictor or outcome that might help rule out the possibility that the reciprocal relationships observed in the Forsman et al. (2010) study were simply a function of variable relationships in place before T1.

The purpose of this study was to ascertain whether a measure of psychopathy that did not include items directly related to criminal behavior (i.e., the YPI) would form a reciprocal causal relationship with self-reported offending from adolescence through early adulthood. There were five pairs of cross-lagged correlations tested in this study. It was hypothesized that the majority of lags would be significantly different from zero but that none of the lag pairs (e.g., YPI-1  self reported offending (SRO)-2 vs. SRO-1  YPI-2) would be significantly different from each other. To test this hypothesis, three demographic control variables (age, race, sex) and seven control variables taken from currently popular theories of crime (differential association, biosocial theory, low self-control, family monitoring, social capital, and routine activities) and measured at baseline were included in the regressions. In addition, an antecedent measure of the outcome variable was included in each individual regression. For instance, SRO-1 was controlled in the YPI-1  SRO-2 regression and YPI-1 was controlled in the SRO-1  YPI-2 regression. Lastly, bias-corrected bootstrapped confidence intervals were used to evaluate the significance of specific lags and determine whether lag pairs (YPI  SRO vs. SRO  YPI) differed significantly from one another.

Section snippets

Participants

Adolescent male and female adjudicated delinquents from the Philadelphia, Pennsylvania and Phoenix, Arizona areas were enrolled in the Pathways to Desistance study (Mulvey, 2012) sometime between November 2000 and January 2003. All 1354 members of the Pathways study (1170 males, 184 females) served as participants in this study. The ethnic breakdown of the Pathways sample was 20.2% white, 41.4% black, 33.5% Hispanic, and 4.8% other, the average age of participants at the time of the baseline

Results

As was mentioned in the previous section, cross-lagged analysis is based on two assumptions: synchronicity and stationarity. The synchronicity assumption was only partially satisfied. It was satisfied to the extent that the YPI and SRO were administered at the same point in time during Waves 1–6 of the Pathways study but was not fully satisfied because, whereas the YPI evaluated the individual’s thoughts, actions, and behaviors at the time of assessment, the SRO considered behaviors over the

Discussion

The results of this study offer preliminary support for the hypothesis that psychopathy, as measured by the YPI, and offending, as measured by the SRO, are reciprocally and perhaps causally connected, at least in males. Previous research has consistently shown that psychopathy and offending are related (DeLisi et al., in press, Laurell et al., 2010, Walsh and Walsh, 2006, Walters, 2003) but this is one of the first studies to establish that the two could be causally linked using a measure of

Glenn D. Walters, Ph.D., is an Associate Professor of Criminal Justice at Kutztown University in Kutztown, Pennsylvania where he teaches courses in research methods, corrections, and substance abuse and crime. His research interests include the assessment of criminal thinking and behavior, classification and the taxometric method, mediation analysis, and creation of an overarching psychological theory of criminal behavior.

References (61)

  • G.M. Vincent et al.

    The PCL:YV and recidivism in male and female juveniles: A follow-up in young adulthood

    International Journal of Law and Psychiatry

    (2008)
  • H. Andershed et al.

    Convergent validity of the Youth Psychopathic Traits Inventory (YPI)

    Assessment

    (2007)
  • H. Andershed et al.

    Psychopathic traits in non-referred youths: A new assessment tool

  • M. Caldwell et al.

    Treatment response of adolescent offenders with psychopathy features: A 2-year follow-up

    Criminal Justice and Behavior

    (2006)
  • P. Chauhan et al.

    Comparing the Youth Psychopathic Traits Inventory (YPI) and the Psychopathy Checklist–Youth Version (PCL–YV) among offending girls

    Assessment

    (2014)
  • H. Cleckley

    The mask of sanity

    (1976)
  • L. Cohen et al.

    Social change and crime rate trends: A routine activity approach

    American Sociological Review

    (1979)
  • D.A. Cole et al.

    Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling

    Journal of Abnormal Psychology

    (2003)
  • D.J. Cooke et al.

    Refining the construct of psychopathy: Towards a hierarchical model

    Psychological Assessment

    (2001)
  • D.J. Cooke et al.

    Understanding the structure of the Psychopathy Checklist—Revised: An exploration of methodological confusion

    British Journal of Psychiatry

    (2007)
  • R.R. Corrado et al.

    Predictive validity of the Psychopathy Checklist: Youth Version for general and violent recidivism

    Behavioral Sciences and the Law

    (2004)
  • F. Declercq et al.

    The Youth Psychopathic Traits Inventory: Factor structure and antisocial behaviour in non-referred 12–19-year-olds

    Journal of Forensic Psychiatry and Psychology

    (2009)
  • M. DeLisi et al.

    Not my fault: Blame externalization is the psychopathic feature most associated with pathological delinquency among confined delinquents

    International Journal of Offender Therapy and Comparative Criminology

    (2014)
  • M. Forsman et al.

    A longitudinal twin study of the direction of effects between psychopathic personality and antisocial behavior

    Child Psychology and Psychiatry

    (2010)
  • A.E. Forth et al.

    Psychopathy Checklist: Youth Version technical manual

    (2003)
  • M. Gottfredson et al.

    A general theory of crime

    (1990)
  • R.D. Hare

    Psychopathy: A clinical construct whose time has come

    Criminal Justice and Behavior

    (1996)
  • R.D. Hare

    Without conscience: The disturbing world of the psychopaths among us

    (1999)
  • R.D. Hare

    The Hare Psychopathy Checklist–Revised manual

    (2003)
  • R.D. Hare et al.

    Psychopathy as a clinical and empirical construct

    Annual Review of Clinical Psychology

    (2008)
  • R.D. Hare et al.

    The role of antisociality in the psychopathy construct: Comment on Skeem and Cooke (2010)

    Psychological Assessment

    (2010)
  • T.J. Harpur et al.

    Factor structure of the Psychopathy Checklist

    Journal of Consulting and Clinical Psychology

    (1988)
  • T.J. Harpur et al.

    Two-factor conceptualization of psychopathy: Construct validity and assessment implications

    Psychological Assessment

    (1989)
  • S.D. Hart et al.

    Manual for the Psychopathy Checklist: Screening Version (PCL:SV)

    (1995)
  • A.F. Hayes

    Introduction to mediation, moderation, and conditional process analysis: A regression-based approach

    (2013)
  • J.F. Hemphill et al.

    Psychopathy and recidivism: A review

    Legal and Criminological Psychology

    (1998)
  • T. Hirschi

    Causes of delinquency

    (1969)
  • D. Huizinga et al.

    Are there multiple paths to delinquency?

    Journal of Criminal Law and Criminology

    (1991)
  • J. Joseph

    The equal environment assumption of the classical twin method: A critical analysis

    Journal of Mind and Behavior

    (1998)
  • Cited by (0)

    Glenn D. Walters, Ph.D., is an Associate Professor of Criminal Justice at Kutztown University in Kutztown, Pennsylvania where he teaches courses in research methods, corrections, and substance abuse and crime. His research interests include the assessment of criminal thinking and behavior, classification and the taxometric method, mediation analysis, and creation of an overarching psychological theory of criminal behavior.

    View full text