xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001498280
007 cr mnu|||uuuuu
008 041209s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000569
Falkenbach, Diana M.
The subtypes of psychopathy and their relationship to hostile and instrumental aggression
h [electronic resource] /
by Diana M. Falkenbach.
[Tampa, Fla.] :
University of South Florida,
Thesis (Ph.D.)--University of South Florida, 2004.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 161 pages.
ABSTRACT: Psychopathy is not a diagnostic category, however theories of psychopathy have been discussed throughout psychological history. While the construct of psychopathy is associated with important psychological outcomes, there are inconsistencies in the literature with regard to correlates, etiology and treatment. These inconsistencies suggest that there may be several subtypes of psychopathy. This paper discusses the heterogeneity of psychopathy and considers the existence of psychopathic traits in nonclinical populations. Measures of etiology (Behavioral Activation System and Behavioral Inhibition System; Gray, 1985), psychopathy (Levenson's psychopathy measure; Levenson, Keihl, & Fitzpatrick, 1995) and anxiety (State Trait Anxiety Inventory; Speilberger, Gorsuch & Lushene, 1970) were used in Model based cluster analysis to investigate the existence of subtypes analogous to primary and secondary psychopathy in college students.Four clusters emerged with cluster profiles differing in theoretically coherent ways. Two of the clusters were representative of subclinical primary and secondary psychopathy and the other two represented non-psychopathic groups. Research (Buss, 1961, Dodge, 1991) regarding aggression discriminates between two types of aggression: instrumental and hostile. The current study considered whether the subjects in clusters created by psychopathy data differ in terms of the types of aggression used. As expected, the Psychopathic Traits groups used more aggression than the Non-psychopathic Traits groups, and the Primary Psychopathic Traits group used more instrumental aggression than the Secondary Psychopathic Traits group. Overall, these results support the existence of subclinical subtypes of psychopathy that resemble, in meaningful ways, hypothetical clinical variants. The results also suggest that subtyping may have clinical and forensic utility in risk assessment.
Co-adviser: Poythress, Norman.
Co-adviser: Epps, James.
Antisocial Personality Disorder.
t USF Electronic Theses and Dissertations.
i The Subtypes of Psychopathy and Their Rela tionship to Hostile and Instrumental Aggression. by Diana M. Falkenbach A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Psychology College of Arts and Sciences University of South Florida Co-Major Professor: Norman Poythress, Ph.D. Co-Major Professor: James Epps, Ph.D. Cythia Cimino, Ph.D Kevin Thompson, Ph.D Paul Spector, Ph.D Paul Stiles, Ph.D Date of Approval: October 28, 2004 Keywords: Antisocial Personality Disorder violence, Sub-types, Cluster Anlysis Copyright 2004, Diana M. Falkenbach
ii DEDICATION Finishing this dissertation woul d not have been possible with out the loving support of so many people. Essentially, I dedicate this disse rtation to those who ha ve read it Â– all my editors: To my mom and dad who showed me how to love and li ve life to the full est, told me I could be anything and do anythi ng I wanted to, and have prov ided endless support in all areas of my life. To my Dad who, despite my constant argumenta tiveness as a kid, is my hero. He taught me resolve and determination, and most of all how to think for myself. Thanks, Daddy. You're the best! To my Mom my friend, therapist, assistant, and so much more who gave me the beautiful, loving, open and trusti ng model of life that I liv e to this day. She is the "wind beneath [all of our] wi ngs". Mom, you are there for everyone, everyday. I love you. To my sister Joannawho always makes me feel connected to family even when I am far away (physically and mentally). Jo, your constant smile and unconditional love always brings joy to my life. To my niece Delaney whose scribbles on my drafts ke pt me from taking everything too seriously, and who, as an infant, kept me company through long hours of writing. To Mama and Papa -who have always been my bigge st fans. Mama, though Papa has passed on, the footprint you both left on my life is forever enduring. To all my friends Â– who just never stopped giving of themselves in c ountless ways and who put up with all my craziness. Your fr iendships make me the luckiest person alive! Finally to Scott Â– Whose unfailing support and belief in me got me through the many stages of grad school and this disse rtation. Scott, you truly bri ng out the best in me in all I do. You have gone through a very tough time with me and carried most of the burden. Despite that, you remain my sun on a cloudy day. You are ther e for me all the times when I need you. You are my best friend, and so much more! Today I finally join real life agai n with all of you. THANK YOU SO MUCH!
i TABLE OF CONTENTS List of Tables iv List of Figures v Abstract vi Introduction 1 Review of the Literature 2 Psychopathy 2 Antisocial Personality Disorder and Psychopathy 2 Definition of Psychopathy and Correlates 5 PCL-R 8 Factors of Psychopathy 9 Subtypes 9 History of Subtypes 10 Karpman 11 Lykken 11 Research on Subtypes an d Cluster Analyses 15 Haapasalo & Pulkkinen. 15 Alterman, et al., 1998 16 Cluster Analysis Conclusions. 18 Subtypes-Factors Distinc tion and Correlates 18 Learning 20 Fear and Anxiety 21 BAS/BIS 24 Implications for Subtyping 27 Aggression 28 The Development of Aggressive Behavior 29 Frustration-Aggression Hypothesis 29
ii Social Learning Theory 31 Hostile (Reactive) and Instrumental Aggression 32 Psychopathy and Aggression Overlap 34 PCL-R Factor Scores and Violence and Aggression 36 Instrumental Versus Hostile Aggression Questions 37 Interaction Between Psychopathy Type and Aggression Type 40 The Assessment of Psychopathy in Non-Institutionalized Samples 42 Self Report Measures 42 Studies of Psychopathy Using Non-Institutionalized Samples 43 Current Study 45 Hypotheses 48 Cluster Analysis 48 Aggression Hypotheses 49 Methods 50 Participants 50 Measures 51 Predictor and Clustering Measures 51 LevensonÂ’s Psychopathy Scale 51 Behavioral Inhibition System/Behav ioral Activation System Scales 52 State-Trait Anxiety Inventory 53 Validation Measures 53 Aggression Questionnaire 53 Aggressive Incidents Coding Form 54 Psychopathic Personality Inventory 55 Procedures 56 Analysis 58 Cluster Analysis 58 Aggression Analyses 62 Results 64
iii Preliminary Analyses 65 Validity 65 Scale Descriptions 66 Correlation Analyses 67 Aggression 69 Aggression Coding 69 Primary Analyses 71 Cluster Analyses 71 Aggression Analyses 78 Discussion 81 Primary Psychopathic Traits Group 83 Secondary Psychopathic Traits Group 86 Clusters Conclusions 89 Aggression 90 Limitations and Future Research 93 Measurement Issues 93 Other Issues 95 References 97 Appendicies 116 Appendix A: Demographics 117 Appendix B: LPS 118 Appendix C: BIS/BAS Scales 121 Appendix D: STAI 124 Appendix E: Aggression Questionnaire 127 Appendix F: Aggressive Incident s Report form and Coding Form 129 Appendix G: PPI 139 About the Author 152
iv LIST OF TABLES Table 1 Comparisons of APD, CleckleyÂ’s Ps ychopathy Criteria and PCL-R Items 4 Table 2 Three Clusters Derived by Haap asalo & Pulkkinen, (1992) 16 Table 3 Clusters Revealed by Al terman et al. (1998) 18 Table 4 Subscales of the PPI 56 Table 5 Model-based Clustering Crite ria with Corresponding Assumptions 61 Table 6 Pearson Correlations Between all Measures 66 Table 7 Mean, Standard Deviatio ns, and Ranges on Measures 67 Table 8 Zero Order and Partial Correlati on, and Differences between Correlations 68 Table 9 AWE Values for Altern ative Cluster Models 72 Table 10 Difference Between th e Clusters on Clus tering Variables 74 Table 11 Demographic Char acteristics of Four Clusters 76 Table 12 Differences Between the Clusters on Aggression 78
v LIST OF FIGURES Figure 1 LykkenÂ’s Antisocial Personalities (1995) 13 Figure 2 Profiles of the three main proposed clusters of psychopathy 48 Figure 3 Mean Z-Scores for Four Clusters 75 Figure 4 Proportions of Instrumental Versus Reactive Aggression for Psychopathic Trait Groups 80
vi The Subtypes of Psychopathy and Their Relationship to Hostile and Instrumental Aggression Diana M. Falkenbach ABSTRACT Psychopathy is not a diagnostic category, however theories of psychopathy have been discussed throughout psychol ogical history. While th e construct of psychopathy is associated with important psychological ou tcomes, there are inconsistencies in the literature with regard to corre lates, etiology and treatment. These inconsistencies suggest that there may be several subtypes of psychopathy. This paper discusses the heterogeneity of psychopathy and considers the existence of psychopathic traits in nonclinical populations. Measures of etio logy (Behavioral Activation System and Behavioral Inhibition System; Gray, 1985), psychopathy (LevensonÂ’s psychopathy measure; Levenson, Keihl, & Fitzpatrick, 1995) and anxiety (State Trait Anxiety Inventory; Speilberger, Gorsuch & Lushene, 1970) were used in Model based cluster analysis to investigate the existence of subt ypes analogous to primar y and secondary psychopathy in college students. Four clusters emerged with cluster profiles differing in theoretically coherent ways. Two of the clusters were representative of subclinical primary and secondary psychopathy and the other two represented non-psychopathic groups. Research (Buss, 1961, Dodge, 1991) regarding aggression discriminates between tw o types of aggression: instrumental and
vii hostile. The current study considered whet her the subjects in clusters created by psychopathy data differ in terms of the types of aggression used. As expected, the Psychopathic Ttraits groups used more aggression than the Non-psychopathic Traits groups, and the Primary Psychopathic Traits group used more instrumental aggression than the Secondary Psychopathic Traits group. Overall, these results support the existence of subclinical subtypes of psychopathy that resemble, in meaningful ways, hy pothetical clinical variants. The results also suggest that subtyping may ha ve clinical and forensic utility in risk assessment.
1 INTRODUCTION Psychopathy is not a diagnostic category, however theories of psychopathy have been discussed throughout psychol ogical history. While th e construct of psychopathy is associated with important psychological out comes, there are inconsistencies in the literature with regard to corre lates, etiology and treatment. These inconsistencies suggest that there may be several subtypes of psychopathy. This paper discusses the heterogeneity of psychopathy and considers the existence of psyc hopathic traits in nonclinical populations. The current study mo re closely scrutinizes psychopathy in a college population by considering various subc linical subtypes of psychopathy that may be identified through cluster analysis. The study will use measures of etiology (Behavioral Activation System and Beha vioral Inhibition System; Gray, 1985), psychopathy (LevensonÂ’s psychopathy measur e; Levenson, Keihl, & Fitzpatrick, 1995) and anxiety (State Trait Anxi ety Inventory; Speilberger, Gorsuch & Lushene, 1970) to determine clusters or subtypes of psychopathy. Research (Buss, 1961, Dodge, 1991) regarding aggression discriminates between tw o types of aggression: instrumental and hostile. The current study will determine whether the subjects in clusters created by psychopathy data differ in terms of the types of aggression used in ways consistent with the theories governing such clusters.
2 REVIEW OF THE LITERATURE Psychopathy Antisocial Personality Disorder and Psychopathy The diagnostic labeling of antisocial be havior has evolved through the decades using terms such as psychopath, sociopath, di ssocial personality di sorder and finally, Antisocial Personality Disorder (APD; Lykken, 1995; Rogers and Dion, 1991). Until recently, the terms psychopathy and antisocial personality were often used interchangeably. However, while psychopathy can be defined in terms of personality traits and behavioral features (Cleckle y, 1976; Hare, 1980), as shown in Table 1, the current Diagnostic and Stat istical Manual of Mental Disorders (DSM; APA, 1994) criteria for APD include mainly the behavi oral features of psychopathy (Wilson, Frick, & Clements, 1999; Hare, 1996). By marginalizi ng personality features, the DSM Â“Â…fails to recognize that the same fundamental pers onality structure, w ith the characteristic pattern of ruthless and vi ndictive behavior, is often displaye d in ways that are not socially disreputable, irresponsible, or illegalÂ” (Millon and Davi s, 1996, p. 443). The limitations imposed by marginalizing psychopathic personal ity features in diagnostic criteria is further evidenced by a strata within the criminal population that display a higher than average risk for negative outcomes such as vi olence, recidivism, institutional infractions and misconduct, and poor treatment respons iveness. The current research on
3 psychopathy indicates that this c onstruct, with both personality and behavioral features, is a more reliable predictor of these nega tive outcomes than a diagnosis of APD (Cunningham & Reidy, 1998; Hart & Hare, 1998).
4 Table 1 Comparison of APD, CleckleyÂ’s Psychopathy Criteria and PCL-R Items 301.7 Antisocial Personality Disorder CleckleyÂ’s Psychopathy Criteria PCL-R items A. Pervasive pattern of disregard for and violation of the rights of others occurring since age 15, as indicated by three (or more) of the following: (1) Failure to conform to social norms with respect to lawful behaviors as indicated by repeatedly performing acts that are grounds for arrest. (2) Deceitfulness, as indicated by repeated lying, use of aliases, or conning others for personal profit or pleasure. (3) Impulsivity or failure to plan ahead. (4) Irritability and aggressiveness, as indicated by repeated physical fight or assaults. (5) Reckless disregard for the safety of self/others. (6) Consistent irresponsibility, as indicated by repeated failure to sustain consistent work behavior or honor financial obligations. (7) Lack of remorse, as indicated by being indifferent to or rationalizing having hurt, mistreated, or stolen from another B. Individual must be a least 18 years C. There is evidence of conduct disorder (CD) with onset before age 15 years. D. The occurrence of antisocial behavior is not exclusively during the course of Schizophrenia or Manic Episode 1. Superficial charm and good intelligence 2. Absence of delusions and other signs of irrational thinking 3. Absence of nervousness or psychoneurotic manifestations 4. Unreliability 5. Untruthfulness and insincerity 6. Lack of remorse or shame 7. Inadequately motivated antisocial behavior 8. Poor judgment and failure to learn by experience 9. Pathological egocentricity and incapacity for love 10. General poverty in affective reactions 11. Specific loss of insight 12. Unresponsiveness in general interpersonal relations 13. Fantastic and uninviting behavior with or without drink 14. Suicide rarely carried out 15. Sex life impersonal trivial and poorly integrated 16. Failure to follow any life plan Factor 1 (interpersonal/affective) Glibness/superficial charm Grandiose sense of self-worth Pathological lying Conning/manipulative Lack of remorse/guilt Shallow affect Callous/lack of empathy Failure to accept responsibility for own actions Factor 2 (Behavioral/Lifestyle) Need for stimulation Parasitic lifestyle Poor behavioral controls Early behavioral problems Lack of realistic, long-term plans Irresponsibility Impulsivity Juvenile delinquency Revocation of conditional release (APA, 1994, pp 649-650) Cleckley (1941, 1976) Hare (1990)
5 Definition of Psychopathy and Correlates Psychopaths are often described as selfis h and disrespectful of the rights and welfare of others, while simultaneously displaying a lack of guilt or concern for the consequences of their actions. They app ear carefree and may act on a whim to satisfy their personal desires. They can be well li ked, displaying superficial charm, but they are insincere and incapable of participating in su stained affective relationships with others. They may also be irresponsible and lack c oncern for their future. Consequently, they may not maintain consistent em ployment (Hare, et al, 1990). The construct of psychopathy, while not a DSM diagnostic category, has a longstanding history in psychological theory. L ykken (1995) states: Â“sin ce the beginnings of psychiatry in the early 19th century, it has been recognized that there are persons whose persistent antisocial behavior cannot be understood in term s of mental or emotional disorder, neurotic motivations, or incompet ent parentingÂ” (p. 113). Pinel (1806) first described psychopaths as people who display a deficit in affect as well as increased impulsivity (Bodholt, Richards, Gacano, 2000), without mental illne ss or intellectual deficits (Lykken, 1995). Kraepelin (1907) sim ilarly described a group of criminals that were particularly immoral. The current conceptualization of psychopathy is based on CleckleyÂ’s (1941, 1976) work. Cleckley used 16 (s ee Table 1) Â“interpersonal, affective, cognitive and behavioral [characteristics] asso ciated with an impulsi ve, irresponsible, and deceitful lifestyleÂ” to identify psychopat hy (Bodholt et al., 2000, p.56). He described people with deficits of conscience who acted in ways unacceptable to society and showed
6 no concern for the consequences of their behavior (Lykken, 1995). It is generally believed that: [the personality structure of the psychopa th is first evident from an early age (Frick, OÂ’Brian, Wooton & McBurnett, 1994), and is well defined by early adolescence (Forth, Hart & Ha re, 1990). It is stable across time (Harpur & Hare, 1994), manifested across a broad range of situations and environments (for reviews see Hare 1996; Hart & Hare, 1997), and likely contributes to the maintenance of antisocial behaviors throughout the individualÂ’s lifespanÂ” (Hemphill, Hare, & Wong, 1998, p. 1401). For over a decade, studies have demonstrated that the construct of psychopathy is associated with a variety of outcomes importa nt to psychology (Ha rris & Cormier, 1995; Hart & Hare, 1997; Serin 1996). Psychopathy is negatively related to startle potentiation (Patrick & Berthot, 1995), fear (Lykken, 1995), learning wi th negative consequences (Schmauk, 1970), inhibitions against aggression (Megargee 1982), time to re-offense (Serin, 1996, Serin & Amos 95), and the perpetration of murder (Williamson, Hare & Wong, 1987). There is empirical support for a positive relationship between psychopathy and violence convictions (H are, 1983; Forth et al., 1990) violence and misconduct in prison (Harris, Rice & Cormier, 1991; To ch, Adams & Grant, 1989), and violence recidivism (Serin, 1996; Serin & Amos 1995) While only a small group of inmates cause problems within the pr ison system, those few consume a large amount of time and 1 Some contemporary researchers do not believe that su fficient evidence exists to indicate the existence of psychopathic traits early in childhood or adolescence.
7 resources, and commit a disproportionate am ount of violent acts towards other inmates and staff (Blackburn & Coid, 1998). Psychopa ths also commit crim es of a greater number, variety and severity (Hare & Jutai, 1983; Hare & McPherson, 1984; Blackburn & Coid, 1998. Toch, et al. ( 1989) looked at a sample of in mates and found that those with multiple misconducts received higher psychopathy scores. These authors found that psychopaths averaged 15.8 misconducts a nd non-psychopaths averaged 5.3 throughout their sentence periods. Hart, Kropp, & Hare (1988) measured psychopathy in male parolees. Psychopathy scores predicted out come better than any other variables, including criminal history and demographi cs. Additionally, the higher the psychopathy score, the higher the percentage of criminals who violated the conditi ons of release; 23.5, 48.9, and 65.2, for low, medium, and high scorers, respectively. Those participants with high psychopathy scores caused more pr oblems for parole supervisors than nonpsychopathic offenders. While consistent relationships have been found between psychopathy and some outcome variables, researchers have b een less successful at understanding the relationship between psychopathy and correlates such as anxiety, fear and learning. For example, beginning with Cleckley (1976) psyc hopathy has been associated with low anxiety, however, evidence of a clear correla tion is lacking. In fact Lykken (1957) found that psychopaths and non-psychopaths have comparable levels of anxiety, but psychopaths demonstrated less fear. Add itionally, ambiguous findings have resulted from studies of physiological responses to anxiety provoking situati ons and neurological response systems (Fowles, 1980). Similarly, the ability of psychopa ths to learn from
8 typical reward-punishment paradigms has b een questioned (Wallace, Schmitt, Vitale & Newman, 2000). PCL-R The gold standard for the measur ement of psychopathy in adult male2 forensic and correctional populations is the Psychopa thy Checklist-Revised (PCL-R; Hare, 1991). The PCL-R assesses psychopathy in terms of seve ral of CleckleyÂ’s orig inal criteria. The measure has been shown to be reliable and va lid (Hare, Harpur, Ha kstain, Forth, Hart, & Newman, 1990; Hare 1991; 2003), and research indicates that it predicts recidivism, violence and criminal behavior better than APD, criminal history and personality variables (Harpur, et al., 1989; Hart & Hare 1989; Kosson et al., 1990). While recent research has suggested the possibility of a three-factor solution, the PCL-R has traditionally been conceptualized as possessi ng an internal structure comprised of two factors (Hare, 1991; Hare, et al., 1990) 3. The two factors are referred to herein as Factor 1 and Factor 2. The two-factor conceptualiz ation is emphasized in this paper as the subtyping theories were devel oped based on research discussi ng the correlates of the two2 Research is underway assessing th e psychometric characteristics of th e PCL-R in female populations, see Vitale, Stevens, Brinkley & Newman, 2002; Hare 2003). 3 More recent research indicates th at a three-factor model may better fit the data (Cooke & Michie, 2001; Skeem, Mulvey, & Grisso, 200 3, Hare, 2003). This three-factor co nceptualization still places emphasis on the affective, interpersonal, and be havioral aspects of the disorder. Research exploring that model is underway; however, Hare (2003) maintains that the th ree factors actually are subf actors of the two primary factors.
9 factors of the PCL-R. Diagnos tically, Â“psychopathyÂ” is traditionally defined as a PCL-R score greater than 29. In general, for someone to score high enough to be diagnosed with psychopathy they must demonstrate both the behavioral and personality features of psychopathy (Wilson, Frick, & Clements, 1999; Harpur, Hare, & Hakstain, 1989). Factors of Psychopathy Factor 1 and Factor 2 of the PCL-R corre spond to the personal ity and behavioral features of psychopathy, respec tively (see Table 1). Factor 1, or the personality-based items (Lilienfeld & Andrews, 1996), describes th e affective and interpersonal aspects. As Cleckley (1941, 1976) and others (Karpm an, 1948; McCord & McCord, 1964) noted, antisocial behaviors are not a necessary component of psychopathy (Wilson, et al., 1999), therefore, these personality f eatures are currently believed to be the core features of psychopathy. Factor 1 includes superficial charm, grandiosity, ma nipulation, callousness, lack of empathy and guilt, and lack of respect or care for others. Fact or 2 of the PCL-R is composed of behavior-based items and is similar to the criteria for APD (Lilienfeld & Andrews, 1996; Lilienfeld, 1994; Widiger & Corb itt, 1993). Factor 2 reflects chronically antisocial or socially deviant behavior, j uvenile delinquency, impulsivity, and criminal versatility. While the two f actors are highly correlated (.50; Harpur, Hakstain, Hare, 1988), they have different external correlates (Harpur et al., 1989). Subtypes There is some controversy in the psychopat hy literature with rega rd to correlates, etiology and treatment. These inconsistencies suggest that there may be several variants of psychopathy. For instance, while there are psychopaths in the prison system, many
10 psychopathic inmates do not reoffend after release (Hemphill, et al., 1998). Additionally, there are those psychopaths that never find th eir way into a cell, due to different life situations, luck or skill (Lykken, 1995). The mu ltiple factor structure of the PCL-R also indicates that there may be numerous groupi ngs within the constr uct of psychopathy; some people tend to exhibit more personali ty features, others demonstrate more behavioral features of psychopathy, while ye t others may score high on both factors. Qualitative differences exist among psychopa ths and this heterogeneity within psychopathy might be clarified by the exploration of subtypes. The clinical l iterature is rich with descriptions of different theoretica l "variants" of psychopath. These theoretical descriptions can guide empirical explor ation of subtypes and may clarify some inconsistent findings regardi ng correlates. A clearer under standing of the variants of psychopathy may also lead to better outco me and prediction, as well as improved modalities for the treatm ent of psychopathy. History of Subtypes Although several subtyping theo ries have been proposed (for a review see Skeem, Poythress, Edens, Lilienfeld, & Cale, 2003), th is paper focuses solely on theories based on modern conceptions and measures of psychopathy, those that conceptualize psychopathy in terms of personality and behavi oral features, and t hose that demonstrate some empirical support in the literature. While CleckleyÂ’s criteria initially defined primary psychopathy, (Karpman, 1948) others have proposed th e existence of another type of psychopath; the secondary psychopath (Hare, 1991, Harpur et al., 1988; Harpur et al., 1989, Karpman, 1948; Lykken, 1995). The theories of Karpman and Lykken are in
11 the forefront in terms con ceptualizing psychopathy as consisting of primary and secondary variants. Karpman One of the first subtyping theories and arguably the basis for current typology theories (Skeem et al., 2003) is th at of Karpman (1941; 1955). Karpman (1948) made a clear distinction betw een primary and secondary psychopathy. Behaviorally the two types of psychopaths are similarly anti social, aggressive and amoral, and both display a lack of regard for the feelings of others. Despite these similarities, Karpman (1948) indicated etiological and motivational differences between primary and secondary psychopaths. He described the primary ps ychopath as the Â“true psychopathÂ”, with a constitutional deficit of conscience that allows for a tendency towards callous, manipulative, glib, selfish and untruthful behavior. Secondary psychopaths display similar characteristics, however etiologically their behavior results from an underlying emotional problem that includes impulsivity a nd elevated anxiety re lated to neuroticism (Lynam, Whiteside, & Jones, 1999; Levens on, Kiehl, & Fitzpatrick., 1995; Karpman, 1955). He described primary psychopaths as planful, purposeful, and goal directed, as well as selfish and unconcerned about the feel ings or thoughts of others. Secondary psychopaths were described as hastier, emotionally reactive (Blackburn & Maybury, 1985), and more environmentally influenced (due to poor parenting). Secondary psychopaths experience some emotions such as guilt and love, and appear at times to seek the affections of others (K arpman, 1941; Skeem et al., 2003). Lykken As shown in Figure1, Lykken (1995) discussed several variants of antisocial personality including the sociopath and the psychopath. He described these
12 variants in terms of differences in temperamen t versus parenting resources. APD is more prevalent in inner cities where Lykken believ es greater exposure to bad parenting and poor socialization causes the number of crimin als within lower Socio Economic Status (SES) groups to increase; Â“unsocialized people tend to do a poor job of socializing their own childrenÂ” (Lykken, 1995, p.11). These an tisocial individuals with normal temperaments and unsociali zed parents are LykkenÂ’s sociopaths. There is a second set of amoral individuals; psychopaths, who come fr om any SES and do not necessarily have a history of Â“bad childhoodsÂ”. Their poor be havior and aggression is hypothesized to be related to difficult temperaments rather than poor parenting. In other words, a Â“psychopath [is] an individual in whom the nor mal processes of socialization have failed to produce the mechanisms of conscience a nd habits of law-abidingness that normally constrain antisocial impulsesÂ” (Lykken, 1995, p. 6) He proposes a constitutional deficit or biological basis for psychopa thic behavior. This theory is supported by LivesleyÂ’s (1995) research that found a higher concorda nce of self-reported psychopathic traits between monozygotic twins than di zygotic twins (Livesley, 1995). Lykken (1995) asserts that psychopaths e ngage in antisocial behavior because their lack of conscience, or indifference to wards punishment, allows them that liberty. Psychopaths seem to be difficult children to raise, and Â“lack conscience, moral values, and habits of good conduct because of some peculia rity in themselves rather than in their upbringingÂ” (p.134). Lykken (1995) believes that it is a deficit in fearfulness that puts people at risk for psychopathic behaviors. The idea that psychopath s are less fearful is supported by research indicating that people with high psychopathy scores display less
13 Figure 1. LykkenÂ’s Antisoci al Personalities (1995)4 autonomic responses (e.g., lower GSR in anticip ation of electric shoc k), and lower startle response to fear inducing imagery (Patrick, 1995; Patrick, Bradley, & Lang, 1993). Typically, children are punished for yieldi ng to Â“immoralÂ” impulses and acting in an antisocial manner, and therefore learn to act in appropriate ways. Those that are indifferent to punishment, Lykken (1995) argues, are Â“unlikely to develop an effective conscienceÂ” (p. 62). Children low in fearfulne ss are at higher risk for behavior problems because they are not as intimidated by punishments. Those less concerned with punishment and the opinions of others do not experience guilt, and social conformities are not internalized (Lykken, 1995). An interaction exists between environm ent and biology. Parenting plays a role but research also supports a constitutional component to antisocial behavior (Wilson & 4 Lykken does not limit his description of antisocial pers onality variants to what is noted in this table. See Lykken (1995) for a discussion of several other, more speculative, variants. Antisocial Personalities Sociopaths Psychopaths Primary Psychopath Secondary Psychopath
14 Herrnstein, 1985). Research has identified ge netic components for many characteristics related to psychopathy, such as aggression and fearlessness. These characteristics have also been identified in infa nts and are quite stable thr oughout life (Kagen, 1992; Eron & Huesman, 1990; Huesman, Eron, Lefkowitz & Wa lder, 1984). Glueck & Glueck (1950) conducted a study with 500 delinquent and 500 non-delinquent Boston boys matched on age, race, IQ and SES. Sampson & Laub ( 1993) utilized this data and found that delinquent youth were more difficult from infa ncy and engaged in more temper tantrums. Longitudinal research (Loeber, 1982) indica tes a connection between childhood behavior problems and psychopathy in adults. Hues man et al.Â’s, (1984) study indicated a relationship between aggression in parents at age 8 and that of their children at age 8 (r=.65). Also, individual differences in le vels of fear can be exaggerated through interaction with environment. Those yout hs who experience less fear may become involved in more daring and riskier activities, and success in those situations allows them to be even less concerned for consequences (Lykken 1995). Lykken (1995), like Karpman, distinguish es between primary and secondary psychopathy. Primary psychopaths are believed to have constitutional deficit in the inhibition of behavioral responses in certain contexts. They are less sensitive and responsive to, and therefore may not inhibit their behavior in response to, cues for punishment or non-reward. Consequently, they may not feel anxiety or fear as others would in similar situations. Secondary psyc hopaths are thought to engage in excessive behavioral response. Their focus is on the attainment of goals, and they may continue reward-seeking behaviors despite adequate awareness of the consequences. Lykken
15 (1995) uses Gray and FowlesÂ’ concepts of the Behavioral Inhibition System (BIS) and Behavioral Activation System (BAS) as a theoretical framework for distinguishing separate etiologies for primary and seconda ry psychopathy. The concepts of BIS and BAS will be discussed further later in this paper. Research on Subtypes and Cluster Analyses Few empirical studies have attempted to identify the subtypes of psychopathy. However, the current theories of psychopathy suggest a constellation of traits that make up at least two variants. Furt her research and empirical evidence is needed to determine whether these subtypes exist beyond theore tical speculation (Blackburn, 1975). The PCL-R factors are a starting poi nt for discussing the idea of variants within psychopathy. However, subtypes are not distinguished solely by considering the relationships between the factors and individual correlates; multiple traits and features need to be taken into account. Cluster analysis allo ws a population to be split into groups based on scores on multiple measures (Magnusson, 1988; Magnusson & Bergman, 1988; Haaspasalo & Pulkkinen, 1992). Thus far two important clus ter analysis studies have been conducted in adults looking at psychopathic traits5 6. Haapasalo & Pulkkinen. In a study by Haapasalo & Pulkkinen (1992) PCL-R items were cluster analyzed for 92 nonviolent Swedish, adult male offenders. Outcome 5 Blackburn (1975) cluster analyzed male offenders using personality scales, however the clustering variables used are not consistent with the Hare/Cleckley conceptualization. 6 Herv, Ling, and Hare (2000) also conducted a cluster analysis study, however, clusters were based solely on the PCL-R and not related to any external variables.
16 variables included criminal be havior and personality scal es such as the Minnesota Multiphasic Personality Indicator (MMPI) Scal e 4 (Pd), California Personality Inventory (CPI) socialization scale (So) and sensation seeking. As s hown in Table 2, analysis of the PCL items distinguished 3 groups: Cluste r 1 were those partic ipants with higher Factor 1 scores and this group may corres pond with primary psychopathy. Subjects in Cluster 2 had higher Factor 2 scores and this group l ooks more like secondary psychopaths. Cluster 3 consisted of low scorers on both factors. Most of the personality measures did not correlate with psychopathy scores, however, the MMPI scale 4 was positively related to Factor 2. Â“It might, not unreasonably, be assumed that the personality questionnaires have more to do with antisocial and criminal behavior starting in early life than with CleckleyÂ’s description of the psychopathÂ’s personalityÂ” (Haapasalo and Pulkkinen, 1992, p.326). While the clusters found by Haapasalo and Pulkkinen are in line with theories of the variants of psychopathy, these authors did not pose any a priori hypothese s about subgroup differences on the measures (Haapasalo & Pulkkinen, 1992). Alterman, McDermott, Cacciola, Ruth erford, Boardman, McKay & Cook, 1998, Alterman et al. (1998) used the PCL-R, m easures of Conduct Disorder (CD) and APD symptoms, and the So scale of the CPI to cl uster analyze 252 methadone patients. Due to difficulties distinguishing clusters based on the factor scores of the PCL-R, the second stage of cluster analysis used individual item s of the PCL-R. Six clusters were revealed representing variants of antis ocial behavior (see Table 3). External criterion validation was conducted using (among others) measures of anxiety, depression, hostility and guilt.
17 Table 2 Three Clusters Derived by H aapasalo & Pulkkinen, (1992). Cluster 1 (n=27) Higher on Factor 1 personality items, (glibness/charm, grandiosity, lying, conning and manipulative, lack of remo rse, shallow affect, callous, lack of realistic long term goals, failure to accept responsibility). Received longest sentences. Primary Psychopathy Cluster 2 (n=23) Higher on Factor 2 targeting behavioral aspects such as early antisocial behavior and poor impulse control. Higher MMPI scale 4. Received more frequent convictions. Secondary Psychopathy Cluster 2 (n=42) Lower scores on all items, more impu lsive than cluster 1 but not as criminally versatile or poorly cont rolled as Cluster 2. Fewer behavior problems than both of the other clusters. Not Psychopathy Mean PCL-R scores are reported for each cluste r as well as the percent of each cluster found to be psychopathic. It is important to not e that the authors used a very liberal PCLR cut-off score of greater or equal to 20 to determine psychopat hy. Three groups were identified as psychopathic. All 3 groups had some antisocial behavior, yet scored differently on external criteri ons such as anxiety. Severa l of the groups have parallel score patterns (but PCL-R scores below 30) to the theoretical descri ption of primary and secondary psychopathy. Group 5 is similar to primary psychopaths with antisocial behavior and no anxiety and Groups 1 and 2 are similar to secondary psychopaths with antisocial behavior, anxiety and depression. These findings support th e idea of variants
18 within psychopathy (Alterman, et al., 1998). Additionally, this study lends support for the existence of variants in a population with subclinical levels of psychopathy as many of the participants did not meet cr iteria for diagnosis of psychopathy. Cluster Analysis Conclusions. The two studies reviewed here use different samples and clustering variables. While this research is suggestiv e of the existence of variants, no consensus can be drawn as to which variables to include in the defi nition of subtypes. However, each study identified clusters of indi viduals that, descriptively, appear similar to the prototypes of primary and secondary psychopaths described in the literature. Haapasalo & PulkkinenÂ’s Cluster 1, is indi cative of primary ps ychopathy with higher scores on Factor 1 items and more severe sentences. Cluster 2 is representative of secondary psychopathy with higher scores on Fa ctor 2 items, more frequent convictions and more antisocial behavior. Alterman et al.Â’s Type 5, is reflective of primary psychopathy with psychopathic behavior a nd low anxiety. Secondary psychopathy is seen in Type 1 and Type 2 with high anxiety and antisocial behavior. Subtypes-Factors Distinction and Correlates Some authors (e.g. Levenson et al., 1995) have suggested that the PCL-R factors themselves distinguish the two subtypes of psychopathy. Factor 1 items have been thought to be indicative of primary psychopa thy and Factor 2 items are thought to be indicative of secondary psychopathy. Differences exist in the respec tive correlates of the two factors. For instance, some authors ha ve considered impulsivity to be a core characteristic of psychopathy, however, Factor 1 does not correlate w ith impulsivity. The relationship between impulsivity and total psychopathy score is attributed to a correlation
19 Table 3 Clusters Revealed by Alterman et al. (1998) Clusters PCL-R mean; % psychopathic Descriptors Type 1 (n =27) 21; 52% psychopathic Early onset of antisocial be havior (high CD); high on antisocial behavior, anxiety, depression and hostility (assaultiveness, indirect, verbal). Secondary psychopathy Type 2 (n = 30) 23; 70% psychopathic Later onset of antisocial be havior (moderate CD); high antisocial behavior, anxiety and depression. Secondary psychopathy Type 3 (n = 45) 17; 0% psychopathic Emotionally unstable moderate antisocial behavior. Type 4 (n = 42) 15; 0% psychopathic Not antisocial; drug related adult antisocial behavior. Type 5 (n = 38) 20; 47% psychopathic Psychopathic criminal, modera te antisocial behavior. Lowest anxiety, depression and guilt. Primary psychopathy Type 6 (n = 70) 9; 0% psychopathic Low antisocial behavior.
20 with Factor 2 only (Harpur et al., 1989). Low anxiety levels have al so been associated with psychopathy (Cleckley, 1976). Resear ch considering the factors individually indicates that anxiety is either unrelated to the factors or po sitively related to Factor 2 and negatively related to Factor 1. Specifica lly, Hare (1991) and Ha rpur et al. (1989) considered this relationship and found that Fact or 1 was negatively re lated to anxiety and neuroticism and Factor 2 was positively related. Patrick et al. (1993) reported that Factor 1 was negatively related to anxiety, fear, startle and distress responses. People with higher Factor 2 scores tended to experience more fear, distress, anger, anxiety, and negative affect. While differences in the PCL-R factor scores provide a starting point for distinguishing the subtypes of psychopathy, a clearer distinction between primary and secondary psychopathy requires that additiona l criteria be considered. Lykken (1995) suggests Â“one way to segregate the primary ps ychopaths more clearly would be to try to identify patterns of two or more indicators that are characteristic of this groupÂ” (p.127). To this end, constructs such as anxiety, and fear, and indicators of Behavioral Inhibition System (BIS) and Behavioral Activation Syst em (BAS) functioning have been identified as potential means for discriminating primar y psychopathy from secondary psychopathy. Learning Studies have found that psychopaths have impaired learning in response to punishment. Psychopaths are known for their quest for self-fulfillment ev en at the risk of punishment, and they appear to be insensitiv e to cues of impending adverse consequences (Patterson, & Newman, 1993; Newman, Wi dom, & Nathan, 1985). Lykken (1957)
21 showed that psychopaths did not inhibit beha viors for which they received shock as a punishment. In research that followed, Schmauk (1970) found that, while psychopaths did not learn as well from physical punishment they performed as well as controls when there was a loss of a reward ra ther then a punishment to avoi d. The authors indicated that psychopaths could use passive avoidance (the inhibition of behavior) when motivated.7 Newman & Kossen (1986) used a go/no-go task to show that psychopaths have difficulty with passive avoidance because of an inab ility to shift attent ion from a dominant response style. They found that psychopaths wo rked towards rewards even at the risk of losing the rewards. When there were co mpeting rewards and punishments psychopaths were deficient in avoiding punishments. However, psychopaths performed as well as non-psychopaths on punishment only tasks. So me researchers have interpreted these difficulties with traditional le arning paradigms to imply that psychopaths display deficits in response modulation, due to problems e ither anticipating adverse consequences, learning from past punishments, or shifting attention from go al-directed behavior, or due to lack of fear and anxiety8. Fear and Anxiety Cleckley believed that a lack of affective response was the essential deficit from which all psychopathic deficits follow, includi ng perhaps their inab ility to modify and change their behaviors in response to punishments (Steuerwald & Kossen, 2000). 7 It is important to note that several of these studies used outdated measures of psychopathy (Newman & Schmitt, 1998). 8 For more information on learning deficits in psychopathy see Newman (1998) or Wallace, et.al., (2000).
22 Cleckley (1976) and Karpman (194 8) believed psychopathic defi cits were associated with low anxiety, however, consistent empirical evidence of a negative correlation between psychopathy and anxiety is la cking. In fact Lykken ( 1957) found psychopaths and nonpsychopaths to have comparable levels of anxiety, but psychopaths demonstrated less fear. LykkenÂ’s theory is similar to Cleckle yÂ’s, however Lykken believed the basis for psychopathic deficits is a speci fic deficit in fearfulness. While the two terms are often used interchangeably, it is important to note that there are differences between fear and anxiety. Â“Fearfulness appears to be a sens itivity to cues of impending danger (Gray, 1982; Tellegen, 1978). In contrast, anxiet y appears to be distress produced by the perception that danger and related consequences are inevitable (T ellegen 1978)Â” (Lilienfeld, 1994, p. 31). CleckleyÂ’s, KarpmanÂ’s and LykkenÂ’s theori es appear to be supported by research examining biological responses to stressors. Fowles (1980) postulated that increased Heart Rate (HR) might result from an effort to cope with impending stress, while Hare (1978) suggested that increased Skin Conductance Levels (S CL) may be related to the subjective experience of anxiet y. Several studies indicate a decreased physiological response to aversive stimuli in psychopaths, relative to non-psychopaths. For example, psychopaths display less of an increase in SCL under stress than non-psychopaths (Hare, Frazelle, & Cox, 1978). Anticipatory HR fi ndings vary, with no difference between psychopaths and non-psychopaths in some studies (Hare & Quinn, 1971) while an increased HR was found for psychopaths in ot hers (Hare et al., 1978). Dengerink & Bertilson (1975) measured SCL and HR in psychopaths and non-psychopaths while using
23 Buss Aggression Machine (Buss, 1961). They found that as aggressi on levels increased, palmer SCL increased in non-psychopaths but not in psychopaths, while HR was higher in psychopaths than non-psychopaths. The authors proposed that the higher HR and lower SCL of psychopaths in anticipation of shock demonstrate an active versus passive coping stance. Olgloff and Wong (1990), found the expected lower SCL in psychopaths, indicating lower levels of anxi ety. They also found psychopaths to have an increase in HR in anticipation of an aversive stimulus but not when subjects were able to prevent the event from occurring. Taken together, psychopathsÂ’ physiological responses to impending aversive stimuli may indicate le ss subjective experiences of anxiety, or successful attempts to cope with aversive si tuations (less increase in SCL) while coping with and decreasing reactions to expected aver sive situations (increased HR; Hare, 1978). These studies are indicative of a rela tionship between psychopathy and anxiety and are consistent with the theories of Cl eckley and Lykken positing that decreased fear or anxiety is an integral pa rt of psychopathy. Interestingl y, the research also indicates some uncertainty about the relationship be tween anxiety and tota l psychopathy score (Kossen, Smith, & Newman., 1990; Schmitt & Ne wman, 1999; Hare, 1991; Lilienfeld, 1994; Harpur et al., 1989). Schmitt and Ne wman (1999) conducted a replication of LykkenÂ’s (1957) research and found that ps ychopathy scores did not correlate with measures of anxiety or fear. The inconsistent results of the various studies may be due to the heterogeneity of psychopathy; the s ubgroups may experience fear and anxiety differently. As such, it may be possible to us e fear and anxiety indicators to distinguish between primary and secondary psychopathy (Hicks, Markon, Patrick, Krueger, In
24 Press). Additionally, Lykken (1995) found: If the low-fear hypothesis is valid then th e Cleckley criteria or PCL-1-denominated psychopaths who are low in harm avoidan ce should constitute the purest group of primary psychopaths, whereas the prison inmates who are not psychopathic according to the PCL-1 and who are also high on harm avoidance should provide the greatest contrast. (p.127) Lykken (1957) used institution staff-rate d Cleckley psychopaths (n=19), neurotic sociopaths with traits sim ilar to secondary psychopaths (n = 20) and a group of nonpsychopaths (n=15). Using his Activity Pr eference Questionnaire (APQ), Lykken found that CleckleyÂ’s psychopaths scored lower than the other two groups on fearfulness. Primary psychopaths also showed less physio logical response (lower SCL) to fear eliciting stimuli (a loud buzzer). Finally, primary psychopaths had more difficulty with passive learning tasks in response to shock. In sum, LykkenÂ’s low fear hypothesis as well as KarpmanÂ’s theory that primary psychopaths may be low anxious while sec ondary psychopaths may experience anxiety helps to reconcile some of the inconsistenc ies in the literature. However, while the distinction between primary and secondary psychopathy ma y help better explain the relationship between psychopathy and anxiety, research has yet to empirically test these theories. BAS/BIS Despite studies indicating a reduced response to punishment and a reduced physiological response (e.g. SCL a nd startle responses) to fe ar and anxiety (Hare, 1972;
25 Hinton & OÂ’Neil, 1978) psychopaths may not be simply under responsive. Physiological research suggests that psychopathsÂ’ lack of response is a coping style that might help prevent the negative affect cau sed by aversive events and situations (Hare 1978; Ogloff & Wong, 1990). This coping style may be re lated to the BIS and the BAS (Gray, 1982; Fowles, 1980). The BIS is triggered by cues of punishment and aversive stimuli, and inhibits ongoing behaviors. It produces anxiety reactions (Pa tterson & Newman, 1993) and increased arousal following behavior interruption, and focuses attention on new environmental cues (Gray, 1982). In cont rast, the BAS is enacted when reward opportunities are presented and sh ifts attention to goal-direct ed behavior. The BAS is related to impulsivity and beha viors associated with attain ing a goal (Fowles, 1980). The BAS and BIS are each activated when a goal is presented. If systems are functioning properly, these systems conflict and provoke a nxiety (Lykken, 1995). For example, when a person sees a forbidden goal, like money le ft out on a desk, the BAS is activated and they feel the desire to take the money. Ho wever, to the degree that guilt and punishment is anticipated for taking the money, the BIS is also activated and anxiety occurs. In wellsocialized individual, the anxiet y due to the BIS overrides the desire thus they do not take the money inhibition wins out. The theory postu lates that antisocial behavior (taking the money) may be a result of either an underac tive BIS (in the case of primary psychopathy) or an overactive BAS (in the case of secondary psychopathy). Through measures of HR and SCL, Fowl es (1980) demonstrated that the physiological indicators of the BIS/BAS model explain psychopathic learning deficits. HR is believed to be reflective of the incenti ve effects of the BAS and SCL is related to
26 the BIS9. Gray (1982) suggested psychopathy was a result of a poorly developed BIS. This theory was supported by Fowles (1980) who demonstrated that the poor passive avoidance learning, lower anxiety levels, a nd low SCL in psychopaths were due to a weak BIS. Psychopaths have less of an incr ease in SCL in the pr esence of aversive stimuli, and Schmauk (1970) found psychopath s had no increase in SCL when awaiting shock punishments. Some psychopaths may also demonstrate an over active BAS as indicated by impulsivity and reward seeki ng behavior (Fowles, 1980). Once a response is established and a punishment contingency is added, psychopaths do not adjust their behavior as well as non-psychopaths. In othe r words, psychopaths are goal orientated, even at the risk of punishment (Patters on & Newman, 1993). Psychopaths also display an increase in HR with aversi ve stimuli that is indicative of more activity in the BAS (Fowles, 1980). As a group, psychopaths do not appear to alte r their behavior in new situations. They also demonstrate more of an increase in HR and less of an increase in SCL than non-psychopaths during aversive situations. Therefore, psychopathy may be related to either an overactive BAS which leads to extr a focus and pursuance of reward (increased HR), and/or an under active BIS, or a lack of behavioral inhibition in response to punishment (decreased SCL; Patterson & Newman, 1993). Patrick, Cuthbert & Lang, (1994) examined this theory more closely. They studied psychological responses to 9 Erhlich and Malmo (1967) found that the HR of psychopaths decreased during an extinction period even though their levels of responding did not change. Roberts (1974) demonstrated an electrodermal response to fear that was independent of HR and activity level.
27 imagined aversive stimuli in inmates. Those with high scores on both factors of psychopathy, as well as those with only high Factor 2 scores had lower SCL and HR than non-psychopaths. The lower HR associated with Factor 2 is consistent with a difference between primary and secondary psychopaths. Given that typology theorists have indicated differential anxiety and impulsivit y for primary and secondary psychopaths, it has been posited (Lykken, 1995; Fowles, 1980) th at a weak BIS may be indicative of primary psychopathy and the overactive BA S may be indicative of secondary psychopathy. A weak BIS relates to Cleck leyÂ’s primary psychopath and LykkenÂ’s low fear hypothesis. The antisocial behaviors of a person with a weak BIS will have little anxiety consequences or accompaniments. A secondary psychopath, or someone with high BAS will actively seek rewards and act impulsively to achieve goals, but may still experience anxiety (Lykken, 1995). Implications for Subtyping The theory and research on psychopathy provides some evidence of the existence of primary and secondary psychopathy. Existin g theories indicate specific etiological, personality, and behavioral differences betw een the variants of psychopathy. Future research must confirm these ideas and determin e their clinical relevance. Psychopathy is a predictor of treatment non-compliance a nd failure (Ogloff, Wong, & Greenwood, 1990; Seto & Lalumiere, 2000), and many believe that a diagnosis of psychopathy is synonymous with being untreatable. Ogloff et al. (1990) considered the treatment outcome of male criminals. Psychopathic in mates improved less, were less motivated, and left treatment earlier than non-psychopath ic inmates. Karpman (1946) specifically
28 referred to the primary psychopath as Â“incurableÂ” and recommended permanent incarceration. A better understand ing of the subtypes might help in the development of more specific and targeted treatment plans for psychopaths. Even if primary psychopathy is untreatable, the identification of subtype s may allow for better treatment of other variants. Secondary psychopathsÂ’ experi ence of anxiety and reactive aggression may indicate the need for better coping strategies. Studies with aggressive children find that they can learn to moderate aggressive reac tions. Additionally, perhaps fearless children can be taught more appropriate outlets fo r their adventurousne ss such as becoming firefighters or poli ce officers (Lykken, 1995). As mentioned earlier, in male offenders, psychopathic traits are associated with risks for negative outcomes such as violen ce, recidivism, diversity and severity of criminal behavior, and institution infractions (Hart & Hare, 1997; Hare & Jutai, 1983; Harris & Cormier, 1995; Forth et al., 1990; Hare & McPherson, 1984; Blackburn & Coid, 1998; Toch et al., 1989). In a meta-analyti c study by Salekin, Rogers, & Sewell (1996) psychopathy (high PCL-R score) was strongly re lated to criminal behavior and violence, both previous and future. The ability to pr edict specific outcomes such as violent and aggressive behaviors may improve with th e identification of subtypes. Accurate prediction may lead to res earch on the development of preventative techniques. Aggression Violence is prevalent in psychopaths and aggression is a key element in understanding the relationship between psychop athy and violence. Â“Human aggression is any behavior directed toward another individual that is ca rried out with the proximate
29 (immediate) intent to cause harm. In addi tion, the perpetrator must believe that the behavior will harm that target and that th e target is motivated to avoid the behaviorÂ” (Bushman &Anderson, 2001, p. 274) The etiology of aggression can not be explained by a single factor. A combination of envir onmental, psychological and biological factors probably best explains aggressi ve behavior. While aggression is important to study in adults, most of the research explaining the development of aggressive behavior has been conducted with youthful samples. The Development of Aggressive Behavior Frustration-Aggression Hypothesis Dollard, Doob, Miller, Mower, Sears, Ford, Hovland, & Sollenberger (1939) proposed that a reciprocal relationship exis ts between frustration and aggression where aggression can not occur without a preceding frus tration and vice versa. This theory is known as the frustration-aggression hypothesis. Dollard and colleagues suggested that frustration results when goal-di rected behavior is blocked. Aggression is an innate drive to injure another person or object in response to frustrat ion. Although aggression is an automatic and necessary response it can be di splaced. As such, aggression is not always directed at the person or object re sponsible for the frustration. There are several criticisms of the frustration-aggressi on hypothesis. First, not every frustration is followed by aggression. Children whose level of aggression has been reduced through prior punishment do not resort to aggressive behavior in the face of frustration. If the frustra tion-aggression hypothesis held tr ue then aggression in these children would not be diminished by punishme nt (Bandura, 1973). Miller (1941) later
30 modified the frustration-aggr ession hypothesis to state th at all frustrations do not necessarily lead to aggressive behavior. Another criticism of the fr ustration-aggression hypothesi s is that the theory does not consider the cognitive and affective aspects of aggressi on. People have individual interpretations regarding frustrating events a nd these processes are not considered in the frustration-aggression hypothesis. Additionall y, frustration is typically followed by an emotional reaction that can lead to ag gression (Brown & Farber, 1951), and that emotional state is typically anger (Berkowitz, 1962). Ange r is a fairly common emotion (Averill, 1982) that often occurs in rela tionships with others (Steuerwald, & Kosson, 2000). When anger is combined with appr opriate stimulation, a ggression can occur; anger can instigate aggressi on (Brown & Farber, 1951). Finally, Berkowitz (1989) criticizes the frustration-aggression hypothesis because it focuses solely on aggression with the goal of injuring another pe rson or object. The frustration-aggression hypothesi s fails to consider aggressi on for alternative purposes. BerkowitzÂ’s (1989) modified frustration-a ggression hypothesis posited that anger is included in the model as a mediator, that fr ustration produces anger which then leads to aggression. Berkowitz also notes that ange r can occur without aggression. Additional modifications have broadened the definition of the mediator to include any negative affect (i.e. irritation, annoyance, sadness, anger), not just anger. Â“Certain kinds of appraisal or attributional beliefs can intensify or w eaken the anger experienceÂ” (Berkowitz, 1990, p. 496). So affective respons e is a mediator in the relationship between frustration and aggressi on; the strength of the ange r or affect produced by the
31 frustration is directly related to th e motivation to act aggressively. Social Learning Theory While the frustration-aggression hypothesis posits that aggression is an innate reaction, social learning theory proposes that aggression is a learned behavior. Bandura (1973) agreed with Berkowitz that frustration elicits an emotional response, but believed the arousal to be more generali zed. Social learning theory in dicates that arousal can lead to many responses, not just aggression. A personÂ’s previous learning and environment determine reactions to frustration. A person may have a biological predisposition to aggression, but the specific reac tion, and the type, frequency a nd victim of aggression are learned through social experien ces. Social experiences de termine how the arousal is labeled (e.g., anger, sadness, etc.) and which responses are considered aggressive. Bandura (1973) indicated that learning pr inciples apply to the acquisition of aggressive behavior. For example, when a rewa rd is presented for an aggressive act, the aggression will likely be repeated in a simila r situation. Inversely, if one is punished for aggression, aggression is less likely to be used on another occasion. The use of aggressive behavior therefore, is typically shaped during upbr inging. In most cases there is a need to be accepted by others. Reward is a childÂ’s indication that they are acting appropriately and they are accepted by societ y, while a punishment indicates they are not acting in accordance with societal rules. For appropriate socializ ation and behavioral shaping to occur, a child must possess a need for approval from others. Parenting plays a large role in this process. For example, a child who is inconsistent ly disciplined may not learn what behaviors are accepta ble and aggressive behaviors may continue. If a child is
32 not cared for appropriately or rejected by a parent, they may never learn to desire acceptance and love of others. The feelings of rejection may also cause hostility and consequential aggressive behaviors (Bandura, 1973). Lykken (1995) describes a similar process for the development of sociopathy. Additionally, LykkenÂ’s primary psychopaths have less need for acceptance from others and are consequently difficult to socialize. Similarly, the children described by Bandura (1 973) lack a need for approval and do not experience guilt or remorse for aggressive behaviors. Like Lykken, Bandura believes that poor parenting or disregard for the feelings of others, leads to de ficiencies in learning to act appropriately and the development of an aggressive and antisocial repertoire. Hostile (Reactive) and Instrumental Aggression Social learning theorists address the overgeneralization of the frustrationaggression hypothesis by separating out hostile/reactive (the terms are used interchangeably) aggression from instru mental aggression. Many researchers discriminate between hostile aggression that is a reaction to provocation or threat, and instrumental or goal-directed aggres sion (Buss, 1961, Dodge, 1991, Feshback, 1964, Parke & Slaby, 1972). Reactive aggression is a r eaction to a frustrati on or threat and is what is described in the frustr ation-aggression hypothesis. It is an angry reaction that is intended to cause harm (Bushman, & Anders on, 2001), typically to someone with whom a prior relationship exists (Cornell, Wa rren, Hawk, Stafford, Oram, & Pine, 1996). Instrumental aggression is aimed at attain ing a particular goal and is not generally directed at harming another person for the sa ke of doing harm. Â“Instrumental aggression is premeditated, calculated behavior that is motivated by some other goal (e.g., obtain
33 money, restore oneÂ’s image, restore justice) Â” (Cornell et al., 1996, p. 274). For instance, when a person attacks someone to obtain his wa tch, the purpose of the attack is not to injure the other person but to obtain the watch. Several characteristics of an aggressive act seem to distinguish instrume ntal from reactive aggression. Instrumental acts typically are planned (or at least require some t hought prior to action) and motivated by an incentive or goal, and are not in response to provocation (Cornell et al., 1996). While reactive aggression is usually aimed at a pe rson with whom a prio r relationship exists, and involves some level of arousal and provocation. The hostile and instrumental aggression di stinction is theoretical. People commit both hostile and instrument aggression, and Be rkowitz (1989) cautione d that individuals can exhibit both types of aggression th roughout their life. Some acts can be conceptualized as combining instrumental and hostile features as most aggressive behaviors can be viewed as achieving some re sult for the aggressor as well as inflicting some harm on the victim. Bushman & Ande rson (2001) suggests that the dichotomy of instrumental and hostile aggression may no long er be a useful distin ction, stating that Â“specifically, the same motives can drive eith er type of aggression, different motives can drive the same aggressive behavior, and ma ny aggressive behaviors are mixtures of hostile and instrumental aggres sionÂ” (p. 276). Because seve ral contributing factors and motivations for aggressive beha vior exist, dichotomizing type s of aggression is somewhat artificial. However, while the distincti on between hostile and instrumental aggression can be problematic, these problems are inherent in research in general. It is these inherent problems that lend importance to this study. The problem is not as simple as
34 determining whether a distinction between in strumental and hostile aggression exists or not. To study constructs that exist in nature, creating clea r distinctions and examining specific variables is often he lpful and necessary. Such simplification is a common flaw in analog research because di stinctions are created that are not obviously reflected in nature. The artificial distin ction not withstanding, such mo dels aid our understanding of naturally occurring constructs. Additionally, as will be demonstrated below, empirical research does support the distinction betw een hostile and instrumental aggression. Psychopathy and Aggression Overlap Psychopaths are at a higher risk for violent behaviors. The research on psychopathy and aggression will assist in the development of methods to determine which individuals are at higher risk for vi olent criminal behavior (Monahan, Steadman, Silver, Appelbaum, Clark Robbins, Mulve y, Roth, ,Grisso, & Banks, 2001; Tengstrom, Grann. Langsrom & Kullgren, 2000). As previously discussed, an overlap exists between psychopathy and aggression (Blackburn & Maybury, 1985; Leary, 1957; Forth et al., 1990; Hare & McPherson, 1984; Kossen et al, 1990; Rice, Harris, & Quincy 1990; Serin, 1991; Harris, Rice, & Cormier 1991; Salekin et al., 1996). Psychopathic aggression is related to so mething other than the experience of anger. The ability of psychopa ths to experience anger is ev en up for debate. Cleckley (1976) believed that primary psychopaths do not experience anger, while McCord and McCord (1964) believed that in response to a deficit in ability to cope with frustrations of everyday life, psychopaths often become angry. More recent research is still unclear. Serin (1991) used the PCL to identify ps ychopathy in inmates. Using hypothetical
35 scenarios, those scoring high on psychopathy self-reported more anger in response to the vignettes of provocation. When film clips we re used, the results we re less clear (Forth, 1992). Forth (1992) used the PCL and measured self-reported anger, HR, SCL and facial expressions in response to anger evoki ng films, and found no differences between psychopaths and non-psychopaths. However, film clips and vignettes may be problematic and weak indicators of anger (Averill, 1982). Steuer wald, & Kosson (2000) indicate that: the exclusive reliance on se lf reports and focus on hypothe tical situations temper generalizations to emotional reactions in real life situations . psychopaths in these studies, could have reported what they thought they should feel instead of what they would actually feel had the scenarios been real. (p. 119-120) Psychopaths may experience anger. However, th e data is unclear, and more research is needed to gain a clearer understanding of anger and aggression in psychopaths. Blackburn and Lee-Evans (1985) used th e Special Hospitals Assessment of Personality and Socialization (SHAPS; Bl ackburn, 1979) to measure psychopathy. The SHAPS has four subscales; P) high psychopa thy and low social withdrawal, S) high psychopathy and high social withdrawal, C) low psychopathy and low social withdrawal, and I) low psychopathy and high social withdraw al. The authors considered differences in aggressive reactions to va rious situations. Participants self-reported their level of reaction to situations presen ted to them. The psychopathic group reported stronger anger reactions than the non-psychopathic gr oup (means = .96, and .67, respectively). Additionally, those with higher levels of withdrawal had st ronger reactions than the low
36 social withdrawal group (means = 0.97 and 0.66) Those subjects with the combination of psychopathy and high social withdrawal c onsistently reported st rong reactions. Those psychopaths with low social withdrawal appeared to differ in their types of reactions, and they Â“respond with less arous al, but with similar levels of aggression and angerÂ” (Blackburn & Lee-Evans, 1985, p.14). The meas ures used to determine psychopathy in this study are not consistent with the theories of Cleckley or Hare therefore these results should be considered with caution. However, it is possible to thi nk of the high social withdrawal group as being similar to s econdary psychopathy and the low social withdrawal group as being similar to primar y psychopathy. If the aggression results are considered with this conceptu alization, there is reason to th ink that different psychopathy profiles might differentia lly display aggression. PCL-R Factor Scores and Violence and Aggression The lack of clarity in the psychopathy and aggression literature may result from a failure to distinguish among va riants (sub-types) of psychopa thy. While research has not considered variants of psychopathy and aggr ession, the two factor s of the PCL and PCLR have been studied separately in respect to violence. Fact or 1 and Factor 2 have each been found to predict violence (Harpur et al. 1989) independently as well as conjointly (Harpur & Hare, 1991). Serin (1996) found that Factor 1 was uniquely related to violent recidivism in 93 prison inmates. So while the behavioral aspects of psychopathy are related to general recidivism, the personality features are uniquely predictive of violent criminal behavior. Kosson, Kelly and White (1997) found that psychopathic traits, specifically Factor 1 of the PCL: Screening Version (SV), predicte d sexual aggression in
37 college men. Blackburn & Coid, (1998) indicate that there is a slightly higher correlation between Factor 1 than Factor 2 with crimes of violence such as robbery and firearms charges. They found that Factor 2 was more related to general criminal behavior and non-violent property crimes. On the othe r hand, Harpur and Hare, (1991) found PCL-R Factors 1 and 2 to be more equally correlate d with violent recidivism. Cornell et al. (1996) noted that those with vi olent histories would be more likely to score high on items assessing previous antisocial behavior (Factor 2). Hare and McPherson, (1984) used the PCL to discriminate between violent a nd nonviolent offenders and the items that discriminated the best were from Factor 2. Ko sson et al., (1997) also found that Factor 2 of the PCL-R predicted violent offenses. Instrumental Versus Hostile Aggression Questions The various findings with regards to a ggression and psychopathy may be a result of aggression being considered as a homogeneous construct wh en, in fact, different types of aggression may exist. As discussed ear lier, aggression can be instrumental or hostile/reactive. Even research that does not directly m easure the relationship between psychopathy and types of aggression, provide s some evidence of a connection between the two constructs. There is some evidence that Â“because of their insensitivity to social, moral, or emotional prohibitions against violence, psychopaths may be more willing to engage in violent behavior for instrumental purpos esÂ” (Cornell et al., 1996, p. 784). A study by Williamson et al. (1987) evaluated the type s of crimes and victims of PCL rated psychopaths. Psychopaths tended to victimi ze people unknown to them, and their crimes
38 involved material gain. N on-psychopaths were more likely to know their victims and their crimes were indicat ive of a strong emotional reaction. Psychopaths were instrumental and non-psychopaths were more reactive. Other studi es indicate a higher degree of weapons use (Hare & McPherson, 1984; Serin, 1991) among psychopaths than among non-psychopaths. Weapon use is a way to control others wit hout physical attack, and can be threatening if used to obt ain goals (Patrick & Zempolich, 1998). Two studies by Cornell et al. (1996) looked at the relationship between psychopathy and types of aggression more di rectly by considering whether psychopathy could distinguish between instrumental and reactive offenders. The PCL-R and the PCLSV were used but no cutoff scores were assi gned; psychopathy scores were considered on a continuum. Three groups were formed based on types of offenses in the records: instrumental, reactive, and non-violent offenders. A coding system developed by the authors was used to rate the type of o ffense and determined instrumentality and reactivity, where one instrumental offense put the offender in the instrumental category. As would be expected, instrumental offenders used more plans a nd had specific goals. Reactive offenders more often knew their vic tim and acted in anger as a reaction to a perceived provocation. Even when lengt h of time served was considered, the instrumental group had significantly higher psyc hopathy scores, than the other groups. One problem with research on instrume ntal and hostile aggression is that offenders are not exclusively instrumental or reactive; no absolute distinction between the types of aggression exists. The authors note that some reactive offenders planed their aggressive acts and reported no anger or specific provocation. Some instrumental
39 offenders reported acting in a state of anger and in a more impulsive manner. Fewer people commit instrumental aggression than reactive aggression and it is possible that instrumental aggression is indicative of more serious pathology. Cornell and colleagues (1996) note Â“capacity to inflict serious inju ry on a person for goal-directed criminal purposes is made possible by the relative lack of wellÂ–internalized social standards and associated feelings of concern and respect for others that otherwise would inhibit the offenderÂ” (p. 788). While these studies indicate a relations hip between psychopathy and instrumental aggression, Serin (1991) found th at PCL psychopaths were similar to non-psychopaths on hostility (Overcontrolled Hostility Scale; Megargee et al, 1967) and reactive anger (NovacoÂ’s Anger Inventory; 1975). Psychopa ths reported more a nger in response to frustrating vignettes, but had the same attrib ution of hostile intent as non-psychopaths. Psychopathy also predicted the use of more in strumental aggression and threats, but less serious harm to victims and fatalities. Oddly, psychopaths appeared angrier than nonpsychopaths but their aggression seemed to be more instrumental. Broad cut-off scores were used to determine psychopathy, but the hi gher anger reaction is inconsistent with previous studies. Toch et al. (1989) f ound a similar relationship between psychopathy and angry or reactive types of aggression. The authors coded Â“antiauthoritarianÂ” (aggression when given an order, thwarted or chastised) as one of the most frequently committed types of misconduct of psychopaths. Most research indicates that psychopathy is related to the us e of instrumental aggression. Psychopaths use more threats a nd weapons than non-psychopaths and have
40 distinct motivations such as sadism, or money. Non-psychopathic offenders are more likely to be involved in sexual assault and to know their victims, and they are more commonly motivated by an angry reaction to perceived provocation or jealousy (Cornell et al. 1996; Meloy, 1988, 1992; Serin, 1991; W illiamson et al. 1987). However, some inconsistency exists in that some studies ha ve found increased reactive anger and reactive aggression in psychopaths. Interaction Between Psychopathy Type and Aggression Type Patrick & Zempolich (1998) stated that Â“[ c]learer interpretation would be possible if studiesÂ…explicitly assessed both components of psychopathyÂ” (p. 331). Research has not considered the relationship between the two types of psychopathy and the two types of aggression. The inconsistency in the literature suggests th e possibility of an interaction effect. Primary psychopaths, posit ed to experience less emotionality, anxiety and fear, and to be more calculating, may utili ze instrumental violence to a greater extent than secondary psychopaths. Secondary ps ychopaths, who experi ence more anxiety, impulsivity, and attachments to others, may be more inclined towards reactive violence. No studies have specifically considered th is interaction, but the little research that has considered aggression in relation to the factors of the PCL-R offers some support for this hypothesis. Factor 1 is related to c onning and emotional detachment, so individuals with high Factor 1 scores should be less prone to angry and violent reactions and more likely to aggress for instrumental purposes. Fact or 2 is related to a nxiety, poor behavioral controls and impulsivity, suggesting a tende ncy to use reactive or angry aggression (Patrick & Zempolich, 1998). Hart and Dempster (1997) used hierarchical regression to
41 consider the unique relationship between the factors of the PCL-R and types of violent offenses. Factor 1 was related to instrume ntal offenses that were planned and goaldirected, while people high on Factor 2 co mmitted more spontaneous offences against victims known to them. In the second study by Cornell et al. (1996) instrumental offenders had higher scores on both factors of the PCL-SV (the authors indicated conservative rating on the personality-based items such as Â“superficial charmÂ” and Â“shallow-affectÂ” so there may be less people w ith high factor 1 scores). In PulkinnenÂ’s (1987) study of juvenile offenders, offensiv e aggression (instrumental aggression) was more stable across time and more predictive of later criminal behavior than defensive (reactive) aggression. This study indicates that instrumental aggression may be related to personality traits and more in dicative of serious pathology. The studies that have considered th e relationship between the factors of psychopathy and the types of aggression give rise to hypotheses about how primary and secondary psychopaths might differ in their use of aggression. It appears that instrumental aggression may be more related to the personality features of psychopathy as measured by Factor 1. It is reasonable to predict then that primary psychopaths, who are higher on Factor 1 than non-psychopaths and, demonstrat e low anxiety and guilt, and a deficiency in the BIS, would be more likely to engage in aggressive behavior in the pursuit of personal goals. It is also reasonable to propose that secondary psychopaths, those with high Factor 2 scores, more anxi ety and impulsivity, and an overactive BAS, might be more reactive to provocation and act aggressively out of anger.
42 The Assessment of Psychopathy in Non-Institutionalized Samples Multiple measures of psychopathy have been compared (Hare, 1985) and correlations between various indices of psychopathy tend to be low (Hare, 1985; Hundleby & Ross, 1977; Newman et al., 1985). The PCL-R is the gold standard for the measurement of psychopathy in correctional and forensic populations, but it is not without its shortcomings. The PCL-R is res ource intensive, requires several hours for completion, requires specialized training and is limited to use with institutionalized populations for which there is access to files of past behaviors. The PCL-R was developed using incarcerated populations (Har e, 1991) and previous research has been limited to institutionalized samples and drug a nd alcohol abusers. There are questions about the applicability of PCL to non-institutionalized populations (Widom, 1977). The antisocial behavior features of Factor 2 were not part of Cl eckleyÂ’s original conceptualization of psychopathy and make th e PCL-R ill-equipped to identify successful psychopaths lacking a criminal history (L ilienfeld & Andrews, 1996). The PCL-RÂ’s emphasis on items assessing criminality also makes criterion contamination an issue when scores on this measure are used to consider the relationship between psychopathy and violence (Patrick and Zempolich, 1998). Self Report Measures Self-report measures alleviate the time and training issues of the PCL-R, but conventional measures are also problematic. Self-report indices from general personality inventories, such as the scale 4 (psychopath ic deviate) and 9 (mania) of the MMPI and the [So] scale of the CPI, have been shown to be related primarily to Factor 2 of the PCL-
43 R (Harpur et al, 1989). However, these scales typically corrolate poo rly, or not at all, with Factor 1. In other words, they are in adequate measures of the core personality features of psychopathy id entified by Cleckley (1976). Newer self-report measures have been developed that are based on CleckleyÂ’s conceptualization of psychopathy and do not si mply measure past antisocial behavior or violence. Several scales have been de veloped, including the LevensonÂ’s Psychopathy Scale (LPS; Levenson, Kiehl, & Fitzpatr ick, 1995), the Psychopathic Personality Inventory (PPI; Lilienfeld and Andrews, 1996), the Psychopathy Checklist: Short Version (PCL:SV; Forth, Brown, Hart, & Hare, 1996), and HareÂ’s (1985) Self Report Psychopathy Scale (SRP-II). Self-report m easures, with the exception of the PCL:SV10 conceptualize psychopathy as dimensional, wh ere higher scores are indicative of more psychopathic traits. Using this methodology psychopathy can be investigated in noninstitutionalized samples (Levenson et al., 1995; Zagon & Jackson, 1994). Studies of Psychopathy Using Non-Institutionalized Samples Several self-report measures of psyc hopathy have been developed using noninstitutionalized samples. In developing the c onsidered psychopathic traits in male and female college students and found a two-factor solution similar to the PCL-R. However, both scales have been found to correlate w ith antisocial behavior. Lilienfeld and Andrews (1996), developed a self-report measure of psychopathy and tested it on university samples as well. They used tw o male samples (n=241 and 253 respectively), 10 The PCL:SV requires a file review, similar to the PC L-R and therefore is not easily used in community samples.
44 and a combined gender sample (n=610). A f actor analysis of the PPI indicated eight discrete dimensions of psychopathy. In a st udy by Forth, et al. (1996), using the PCL-SV with 150 college students, the two-factor solu tion was not identified. Less research is available with the SRP-II, however, Zagon and Jackson (1994) conducted a study to consider the construct validity of HareÂ’s (1985) SRP-II. No factor analysis was conducted (Wilson et al., 1999), therefore this measure does not fit the two factor model proposed by theory. Levenson, (1992) commented that one of the difficulties in developing a selfreport measure of psychopathy is targeting Factor 1 traits in a sample that is unlikely to have a clinically significant level of these tr aits. The difficulties are consistent with the problems inherent in earlier measures of psychopathy by self-report. However, although the base-rate in college samples for some ps ychopathy items is low (e specially Factor 1), studies have found enough responses and variabi lity to allow for interpretation (Levenson et al 1995). With a low base -rate, these scales consider psychopathy on a continuum. In this manner, the relative strength of psyc hopathic traits is cons idered rather than assigning a cut-off score that deems someone psychopathic. So while the newer selfreport measures of psychopathy have not alle viated the difficulties involved in the measurement of psychopathy in non-institutionalized samples, some of them appear to be adequate research measures of psychopathic features.
45 CURRENT STUDY This study used cluster analysis to id entify groups of individuals who differ on psychopathic features and to inve stigate whether those groups differ in the use of instrumental versus hostile aggression. As this study util ized a non-institutionalized sample of college students, cluster analysis was us ed to identify sub-clinical variants of psychopathy within the sample. Within the psychopathy literature, theories ha ve been posited regarding the existence of subtypes with differing etiologies and correlates. Cleckley described primary psychopaths as lacking nervousness and Karpma n (1941) described secondary psychopaths as having an underlying neurotic urge and a tendency to act impulsively. Primary psychopaths plan their behaviors whereas secondary psychopaths are mo re impulsive. Lykken (1995) postulated that a constitutional deficit exists for psychopaths, an d Fowles (1980) related th at deficit to GrayÂ’s (1985) BIS/BAS model. These authors believ e that primary psychopaths may have an under active BIS and that secondary psychopaths may have an overactive BAS. The current study combined these theories into a model to use for cluster analysis. As such, the choice of clustering variables was di ctated by the etiological theories proposed. The LPS (Levenson et al, 1995), the BIS/BA S scales (Carver & White, 1994) and the State Trait Anxiety Inventory (S TAI; Spielberger et al., 1970) trait scale were included as clustering variables to explore for subtypes analogous to primar y and secondary psychopathy. It was not anticipated that there would be a la rge number of true clinical psychopaths within the sample, however, since psychopathy is measured dimensionally on the LPS, participants in
46 the sample could be identified th at display sub-clinical levels of psychopathic traits. Existing data indicates the existence of psychopathy in correctional samples and most of the previous research has focused on examining psychopathy in correctional or forensic samples. However, while the literature discusses the existence psychopathy in non-institutionaliz ed populations, there is little empirical eviden ce to support these theories. Ther e are clinical anecdotes and a few studies that indicate the existence of a parallel construct in non-forensic groups. Additionally, many theories of ps ychopathy indicate that ther e are Â“successfulÂ” psychopaths outside of forensic settings. For example, L ykken (1995) states that while psychopathic traits may incline someone towards criminal behavior many factors may keep that person out of prison. The existence of psychopathy in non-in stitutionalized populations and the concept of the Â“successfulÂ” psychopath needed further ex ploration. The current study used college students in an effort to provide evidence of psychopathy in non-inst itutionalized populations and to identify subclinical or Â“successfulÂ” psychopaths. Although several other self-repo rt measures are available, the LPS was chosen for several reasons. Thus far, studies have not concentrated on the relationship between instrumental and reactive violence and the two aspects of psychopathy in student or community samples. Given the different correlates of the f actors, more research is needed in this area (Harpur et al., 1989, Patrick, 1994, 1995, Kossen et al., 1997). The LPS is the only self-report measure that has reliably demonstrated a two-fa ctor solution consistent with the theories for primary and secondary psychopathy11. Additionally, several studies have used this measure in 11 Since the proposal of this paper, Benning, Patrick, Hicks, Blonigen, & Krueger have found a two-factor
47 college samples and found it was similarly related to variables that have been correlated with the PCL-R in forensic samples (Lynam et al ., 1999; McHoskey, Worzel, & Szyarto, 1998). The PPI was administered as a construct valida tion measure for the LPS, with the hope that there would be a signif icant positive correlation between the two measures to confirm that the LPS is measuring a construc t analogous to psychopathy. Most of the research on psychopathy has been conducted solely with male samples. Women were not used in this st udy because there is concern that the current conceptualization of psychopathy may not capture psychopathic fe atures as they are manifest in women (Hamburger, Lilienfled & Hogben, 1996). In prev ious studies of non-institutionalized samples (Lilienfeld & Andrews, 1996), that utilized men and women, men were found to have considerably higher levels of psychopathy (Lev enson et al., 1995). A dditionally, at the time that the earlier research was ta king place, there was no research from which to base predictions or hypotheses for outcome with women. Due to these concerns and the possibility that additional variability due to ge nder may cause problems in the cl uster analysis results, women were excluded from the study. There is limited data regarding psychopa thy and racially or ethnically diverse populations. For the most part, previous resear ch has used racially homogeneous samples of Caucasian males. Therefore, it was difficult to predict how racially or ethnically diverse subjects would perform. However the LPS wa s developed using a racially heterogeneous group of student subjects (Levenson, et al., 1995). Since the university population is solution for the PPI.
48 heterogeneous, this study utilized whatever composition of subjects was available. Hypotheses Cluster Analysis A person centered cluster an alysis was conducted and wa s expected to yield 4-6 clusters. Upon examination, two clusters were e xpected to be representative of or similar to primary psychopathy and secondary ps ychopathy. Figure 2 depicts the expected variable profiles of these clusters. Factor 1Factor 2BISBASTrait AnxietyHigh Medium Low Primary Psychopathy Secondary Psychoapthy NonPsychopathy Figure 2. Profiles of the three main proposed clusters of psychopathy.
49 a. LPS results Those participants in both the psychopathy groups were expected to have higher scores on both Factors of the LPS than non-psychopathy groups. Those in the primary psychopathy group we re expected to have high scores on Factor 1 and moderate to high scores on Factor 2 of the LPS. Those in the secondary psychopathy cluster were expect ed to have higher scores on Factor 2 than Factor 1 of the LPS. b. BIS/BAS results The primary psychopat hy group was expected to have low scores on the BIS scale and normal scores on the BAS measure. Those in the secondary psychopathy group were expect ed to have high scores on the BAS measure and normal scores on the BIS scales12. c. STAI trait scale results Those partic ipants in the primary psychopathy group were expected to have lower scores on th e STAI trait scale, while those in the secondary group were expected to scor e higher on the STAI trait scale. Aggression Hypotheses These empirically derived clusters were to be used to predict aggression as measured by the Aggression Questionnaire (AQ; Buss & Perry, 1992), and use of hostile and instrumental aggression us ing a modified version of the coding guide for violent incidents (Cornell et al., 1996). This coding guide is used to rate specific aggressive 12 McHoskey et al. (1998) found that college students w ith high scores on both the BIS and BAS scales had the highest secondary psychopathy scores on the LPS, so a secondary hypothesis was that the secondary psychopathy cluster may have high scores on both the BAS and the BIS scales.
50 events as primarily reactive or instrumental in nature. The particular hypotheses were as follows: a. Higher aggression scores on the AQ for the psychopathy clusters than the non-psychopathy clusters. b. The primary psychopathy group, because of their cunning and manipulative means of atta ining their goals, were expected to be rated as using more instrumental aggressi on. The secondary group, because of their greater susceptibility to stress and higher degree of impulsivity were expected to use more hostile or r eactive aggression. No research is available on type of aggression e xpected by the non-psychopathy groups, however given the hypothesis that instrumental aggr ession is indicative of more severe pathology, it is expect ed that the non-psychopathic groups will use more hostile aggression. Methods Participants The participants were 96 male college students in psychology classes from the University of South Florida who participated in the study for extra credit points. The ages of the participants ranged from 18 to 46 years old (M: 21.46, SD: 4.56). The sample was racially diverse with 59.4 % (n=57) Cau casians, 20.8 % (n=20) African Americans, 19.8 % (n=19) from other racial backgrounds, and 1% who did not indicate their racial group. As for ethnicity, 12.4% (n=12) of the sample was Hispan ic, 54.6% (n=53) categorized
51 themselves as other, and 33% (n=32) of th e sample did not specify their ethnic group. Most of the participants we re single, (93.8%, n=90), 4.1% (n= 4) were married, and 2.1% (n=2) were divorced. The sampleÂ’s socio-econo mic status was skewed toward the upper range of the spectrum with 48.5% (n=47) be longing to families with incomes of $60,000 or more, and 59.8% (n=58) of whose fa mily income was over $40,000. On the other extreme, a total of 19.6% (n=19) of partic ipants came from families with incomes under $19,000, including 11.3% (n=11) of the sample who reported incomes under $9,000, which put them under the poverty level. Measures Basic demographic information on each pa rticipant was obtained from a selfreport demographics form. This form include d psychosocial history data (see Appendix A). Predictor and Clustering Measures LevensonÂ’s Psychopathy Scale (LPS; Appendix B ). Levenson et al. (1995) created the 26-item Primary (16 items) and Secondary (10 items) psychopathy scales in order to assess Factor 1 and Factor 2, respectively, of the PCL-R in non-inst itutionalized groups. The Primary scale measures callous, selfish and manipulative attit udes and the Secondary scale targets more behavioral features such as impulsivity, lack of long term goals, low frustration tolerance and self-d efeating tendencies. The factor analytically-derived scales consider psychopathic traits and behaviors that non-forens ic samples might endorse. Levenson and colleagues (1995) de veloped their measure using 487 undergraduate students. The two scales were correlated at r = .40 (cons istent with HareÂ’s
52 findings with the PCL-R; r =.50) and the co efficient alphas for the scales were .82 (Primary) and .63 (Secondary). Both scales we re significantly correla ted with antisocial behavior at .44 (Primary) and .29 (Seconda ry). As hypothesized, primary psychopathy was weakly related to trait anxiety an d Secondary Psychopathy was significantly correlated with trait anxiety. Both scales were significantly relate d to disinhibition and proneness to boredom, with no co rrelations with sensation se eking. These findings have been replicated by McHoskey et al. (1998) and Lynam et al. (1999). Lynam and collegues (1999) found that high scorers on the LPS reported more serious, and a greater variety of, antisocial behaviors over their lifetime as well as over the previous year. Testretest reliability over an average inte rval of eight weeks was .83. With minor modifications, the two-factor solution was re plicated in college students (Lynam et al., 1999). Behavioral Inhibition System/Behavioral Activation System Scales (BIS/BAS; Appendix C ). Carver and White (1994) began th e development of the BIS/BAS scale by generating a pool of items reflecting BIS or BAS sensitivity and the role of theses systems in generating emotional responses. BIS items consisted of asking people about responses to potentially punishing events. BAS scale items ask a bout Â“strong pursuit of appetitive goalsÂ…responsiveness to rewardÂ…a tendency to seek out new and potentially rewarding experiencesÂ” (p.322). Factor analysis of respons es by 732 university students revealed four scales: the BIS scale, consisti ng of 7 items and three-BAS related scales Â— Drive (4 items); Reward Responsiveness (5 items); and Fun Seeking (4 items). Correlations between the BIS and the three BA S scales were low, providing support for
53 the independence of the systems (-.12, -.08, .28). At the same time, the three BAS scales were positively correlated with each other, ranging from .34 to .41. Internal consistency as measured by coefficient alpha ranged from .66-.76, and test-retest reliability (8 week interval) ranged from .59 (Reward Responsiven ess) to .69 (Fun Seeking). In the current study, as in McHoskey et al. (1998) the BAS scale total was calculated by summing scores for the three BAS scales. State-Trait Anxiety Inventory (STAI; Speilberger et al., 1970; Appendix D ). The STAI consists of two 20-item scales. The state scale examines the current anxiety functioning and asks people about how they feel at the time they are completing the questionnaire. The trait scal e considers long-term manifest ations of anxiety and asks people about how they typically feel. A study by Gaudry, Vagg, & Spielberger (1975) indicated excellent internal c onsistency, high retest reliabil ity, and construct validation. Concurrent validity for the trait anxiety scal e is demonstrated by the correlations from .70-.85 (college students and neur opsychiatric patients) with th e Taylor Manifest Anxiety Scale (Taylor, 1953) and the Anxiety Scale Questionnaire (C attell & Scheier, 1963). Patients with anxiety neurosis also had higher scores on the trait scal e than other types of patients (Speilberger, 1985). Similarly, those in stressful situati ons score higher than when they are relaxed (Speilberger et al., 1970). Validation Measures Aggression Questionnaire (AQ; Appendix E). Buss and Perry (1992) developed the AQ in an effort to address the weaknesses of the older Buss Durkee Hostility Inventory (BDHI; Buss & Durkee, 1957). Most importantly, Buss and Perry noted that
54 the standards for constructing questionnaires at the time the BDHI was developed were less stringent than todayÂ’s sta ndards. The AQ item set was ad ministered to three groups of college students (612 men and 641 women) who rated the items on a scale from 1 (extremely uncharacteristic of me) to 5 (extreme ly characteristic of me). Factor analyses of these initial samples resulted in a four factor solution for the AQ: physical aggression, verbal aggression, anger, and hostility. Twenty-nine items were retained which met the criteria of loading above .35 on one factor and less than .35 on the other factors. The internal consistencies of the total score and the individual factors were evaluated using coefficient alpha and found to be satisfactory (Total score, .89; Physical Aggression, .85; Verbal Aggression. .72; Ange r, .83; and Hostility, .77). Satisfactory test-retest correlations (.72-.80) for 372 of the original subjects were obtained over a 9-week interval. Aggressive Incidents Coding Form (Adapted from Cornell et al., 1996; Appendix F ). Previous research on the relationship between psychopathy and hostile versus instrumental aggression has been conducte d using only forensic populations. The outcome measure in those studies was the ty pe of crime committed, and the rating scale developed by Cornell et al. (1996) was used to determine whether an aggressive act was instrumental or hostile/reactiv e. In the current study, pa rticipants reported acts of aggression and trained raters assessed these incidents on five different aspects of aggression that are believed to be relate d to instrumental and hostile aggression: Planning, Goal Directedness, Provocation, Arous al, and Relationship to Victim. Raters
55 used these scales to classify each past a ggressive incident as either reactive or instrumental. Psychopathic Personality Inventory (Lilie nfeld & Andrews, 1996; Appendix G). The PPI was developed to measure the core personality features of psychopathy in noninstitutional populations. A literature review revealed 24 focal constructs relevant to psychopathy, and all were included in the origin al items of the scale. A series of studies with undergraduates yielded the final 187-item measure. Items are rated on a four point Likert-type scale with 1 being false, 2 being mostly false, 3 being mostly true, and 4 being true. Factor analysis revealed the ei ght subscales shown in Table 4, plus three validity scales designed to a ssess response styles. The Va riable Response Inconsistency scale (VRIN) assesses how consistently re spondents answer items that have similar content. The Deviant Responding (DR) scale is designed to consider random or careless responding, which is thought to be indicative of malingering. The Unlikely Virtues scale is from TellegenÂ’s Multidimensional Pers onality Questionnaire (MPQ, 1982), which assesses impression management and social desirability in responding. In multiple samples the PPI has demonstrated excellent psychometric properties. Internal consistency was excellent for both th e total (from .90-.93) and the subscales (.70.90). Test-retest reliability was also very good. With an average of 26 days between testing, the correlations for the subscales ranged from .82-.94 and for the test-retest reliability for the Total scale was .95. The construct validity has been demonstrated through positive correlations ( .37-.91) with other self-report psychopathy measures, and
56 interview-based as well as interview rated measures of antisocial personality features (.59, .60, respectively). Table 4 Subscales of the PPI Machiavellian Egocentricity An orie ntation of Â“ruthless practicalityÂ” and manipulativeness in relationships with others Social Potency A tendency to be charming and adeptness at influencing others Cold-heartedness A propensity toward callousness, guiltlessness, and unsentimentality Carefree Nonplanfulness The nonplanning component of impulsivity Fearlessness Absence of an ticipatory anxiety concerning harm, and eagerness to take risks Blame Externalization Tendency to blame others for oneÂ’s difficulties and to rationalize oneÂ’s misbehavior Impulsive Nonconformity Reckless la ck of concern regarding social mores Stress Immunity Absence of ma rked reactions to anxiety provoking events Procedures Students in classes offering extra credit poi nts for research study participation were invited to volunteer for the current study. C onsent of participants was obtained using procedures approved by the University Instit utional Review Board. The consent form explained the procedures of the study and pr ovided information about confidentiality and the voluntary nature of the study. Upon recruitment into the study each student completed the demographics form first. Th e personality and psychopathy measures were given next (LPS, PPI, BIS/BAS scale, STA I). The aggression questionnaires were
57 administered last so that the aggression que stions did not influence the answers on the psychopathy questionnaires. The final questi onnaires were the AQ, and the aggression incidents reporting form, given in that orde r. The AQ asks specific questions about aggression and it was anticipate d that those questions would prime the participants to think of aggressive incidents from their past and therefore help them in completing the aggression reporting form. Administ ration time took approximately 1 hour. Participants were next asked to report on previous acts of aggression. After reporting past acts of general aggression, participants were specifically asked to report any acts of instrumental a ggression. For the purpose of this study instrumental aggression was defined as an act committed Â“for a clearly identifiable purpose other than responding to provocation or frustration.Â” (Cornell, et. al, 1996, p. 785), while reactive/angry aggression was defined as an act that was considered a Â“reaction to a dispute or interpersonal c onflict with the victimÂ” (Cor nell, et. al, 1996, p. 785). Cornell et al, (1996) distinguished betw een various aggression characteristics relevant to the distinction between in strumental and reactive aggression. The characteristics were, (a) pla nning, (b) goal-directedness, (c ) provocation, (d) anger, and (e) victim relationship. Three raters (undergraduate and graduate research assistants in a research lab focusing on aggression) were trained on the ratin g scale designed by Cornell et al (1996). Three pilot cases were chosen for training the raters and all raters agreed on the ratings for the test cases. Participants were classi fied as instrumental or reactive aggressors based on their report of past aggressive in cidents. The coding scheme that was
58 established for this study (bas ed on Cornell et al, 1996) had raters code each act on five characteristics, which were used to help inform the decision about reactivity versus instrumentality. Based on previous res earch (Cornell et al., 1996), instrumental aggression is presumed to be a rare occurr ence, and therefore a ny subject who reported instrumental aggression was placed in the Inst rumental Group. If all reported acts were rated as reactive then the participant was placed in the Reactive group. Two raters independently rated all participants as Instru mental aggressors or Reactive aggressors, and a third rater was employed as a tie breake r for the cases where the primary raters did not agree. Analysis Cluster Analysis The software package mclust, which is integrated into the computer program SPlus, was utilized for cluster analyses (Fra ley, 1998). The Statistical Package for the Social Sciences (SPSS/Graduate Pack 11.0 fo r Windows) was used for all preliminary and follow-up analyses. Cluster analysis identifies patterns of association among variables and creates subgroups in samples. Cluster analysis is indicated when corre lations alone may not capture the Â“true underlying structure of the dataÂ” (Rapkin & Luke, 1993, p. 252). Cluster analysis identifies cases in a sample w ith similar scores on all variables of interest and puts them together to form clusters or subgroups of cases (Rapkin & Luke, 1993). The clusters that emerge can be characteri zed by their profile of mean scores on each variable (Aldenderfer & Blashfield, 1984). Cluster analysis places no restrictions on the
59 number of variables used to create the prof iles, suggesting that the selection of five variables for the present study was acceptable. The cluster analysis was conducted followi ng a series of steps suggested in the literature (see Rapkin & Luke, 1993). First, the scores on th e predictor variables were standardized. An agglomerative hierarchical method was used with this data. This method was selected as it does not require pr e-specification of the number of clusters to be considered for the analysis. The agglomer ative hierarchical appr oach initially treats all cases as separate. Cases are combined (o r agglomerated) into clusters based on their similarity of ratings on all five of the pr edictor variables. This method produces nonoverlapping clusters (Blashfield & Alde nderfer, 1988; Rapkin & Luke, 1993). The hierarchical agglomerative method can use di fferent methods to determine the structure of the data. Model-based cluster analysis is a type of hierarchical cl ustering that is based on the Â“assumption that the data are gene rated by a mixture of underlying probability distributions;Â” (Insi ghtful Corporation, 1988-2001, p. 141) where each person is assumed to belong to a number of subpopulations. A f it criterion is used to estimate how many subpopulations exist within the sample popula tion, which participants belong to which clusters, and the mean vector and covariance matrix for each subpopulation. Multiple models are generated, each with different as sumptions about the size, shape, di stribution and orientation of clusters. In order to de termine the best model, an index of fit is calculated and goodness of f it is determined (Average Weight of Evidence; AWE statistic). Six different models are tested in mo del based clustering, each with different
60 assumptions about the covariance matri ces, and the assumptions determine the characteristics that determine the shape and structure of the data (see Fraley, 1998 for a more detailed discussion of the a ssumptions). One characteristic, the size of the distribution, which is indicated by the larges t eigenvalue, specifies the variances and covariances of the covariance matrix. Th e second characteristic specifies the shape of the distribution, and is ratio of th e eigenvalues to the highest ei genvalue. The eigenvectors of the covariance matrix, specify the third characteristic; orientation (Insightful Corporation, 1988-2001, p. 142). The six alternative models are determ ined by variably constraining these characteristics. As shown in Table 5, the Sum of Squares model assumes the distributions of the clusters are all the same size and are spherical in shape. The Spherical model allows the size to vary across clusters, but ag ain holds the shape constant as spherical. The Determinant model constrai ns the size and orientat ion of the clusters, and restricts the shape to be ellipsoidal. The S model holds the size constant and constrains the shape to ellipsoid al, but allows the clusters to differ in orientation. The S* model assumes ellipsoidal shape but allows the orientation and size to vary across clusters. Finally, the Unconstrained model allo ws the size, shape, and orientation of the clusters to vary. In order to determine the best number of clusters for a sample, the fit of each model is evaluated by using the Bayesian Info rmation Criterion (BIC) which specifies the odds that one model is the best fit compared to the other models. There are several benefits of using a goodness-of-f it index such as BIC, including the ability to select the
61 Table 5 Model-based clustering criteria with corresponding assumptions Criterion Reference Shape Orientation Size Sum of Squares Ward (1963) Spherical None Same Spherical Banfield and Raferty (1992) Spherical None Different Determinant Friedman and rubin (1967) Ellipsoidal Same Same S Murtagh and Raferty (1984) Ellipsoidal Different Same S* Banfield and Raferty (1992) Ellipsoidal Different Different Unconstrained Scott and Symons (1971) Ellipsoidal (can vary) Different Different Note. Table adopted from S-Plus 6 fo r Windows Guide to St atistics (Insightful Corporation, 1988-2001) optimal solution, removing much of the subj ectivity otherwise invol ved in selecting a cluster solution, and allowing for testing multiple models without the models having to be nested. The AWE statistic is an approximati on of the BIC factor, and larger AWE values indicate more evidence for that particular number of clusters. Negative values indicate
62 no evidence for that particular number of cl usters. The difference between AWE values relates to the posterior odds, or the ratio of the probability that one model is correct over the probability that another model is correct. A difference of 10 indi cates that the odds are 150:1 that the model with the higher value is the better f itting model. Raftery (1995) suggests that a difference of 10 is a very str ong indicator that the model with the higher values is a better fit. Whereas some clustering programs attempt to force all subjects into clusters, mclust includes a Â“robust clusteringÂ” option (Noise T) that pe rmits individual Â“outliersÂ” to be identified and to remain apart (i.e., as 1-member clusters) from primary clusters in seeking the optimal solution. In the presen t study, an initial an alysis using robust clustering revealed that no i ndividuals were considered to be outliers. Therefore, the standard mclust method (Noise F) for clustering partic ipants was used to derive clusters. The five clustering variables (LPS Primary, LPS Secondary, BIS, BAS, and STAI) were used for the cluster analysis and all six models were fit. Participants with similar scores on all of the variables we re grouped together in the same cluster. A one-way MANOVA was conducted as a fo llow-up to the cluster analysis to consider the differences between the cluste rs on the clustering variables. ANOVA were used as follow-up tests to cons ider the between-groups effects on the individual variables. Post hoc analyses of the univariate ANOV A were conducted using the Bonferoni method to compare the clusters on each of the clustering variables. Aggression Analyses Once the sample was combined into the appr opriate clusters, a se ries of analyses
63 were conducted comparing the profiles/clusters on the dependent/outcome variables: selfreported aggression (AQ scores ), and type of aggression perpetrated (hostile or instrumental aggression). A one-way ANOVA was used to consider the between-cluster differences on aggression. Post hoc analyses of the univariate ANOVA were conducted using the Bonferoni method. A 2 by 2 contingenc y table analysis was used to evaluate if a statistical relationship exists between type of aggression and subtype of psychopathy. The two types of aggression, reactive and in strumental, were represented in the two columns of the table. The number of rows wa s determined by the number of clusters that indicate psychopathic traits. A Chi-square analysis evaluated the heterogeneity of the clusters, by considering the propor tion of participants in the di fferent clusters that have committed instrumental versus reactive aggressive acts. Follow-up tests, or pair-wise comparisons, that are conceptually similar to post hoc tests for Analysis of Variance, were conducted for closer analyses on the differences detected between the clusters (Green, Salkind, & Akey, 2000).
64 RESULTS The Statistical Package for the Social Sciences (SPSS/Graduate Pack 11.0 for Windows) was used for all preliminary and follow-up analyses. The software package mclust which is integrated into the S-Plus stat istical package, was utilized for cluster analyses (Fraley, 1998). For clarity, the results section is divided into three main sections -preliminary analyses, aggression coding and primary anal yses. Preliminary analyses included the assessment of protocol validity, a descripti on of the scales and the interrelationships between the variables used in the study. The next secti on of the results includes the coding of aggressive acts and comparisons between the aggression groups on LevensonÂ’s Psychopathy Scale (LPS) and subscales (Primary psychopathy and Secondary psychopathy). The primary analyses are divided into three parts: a) The cluster analysis section presents results from the first hypot hesis regarding the nu mber of clusters resulting from the model-based cluster anal yses; b) The cluster comparisons section presents results from the sub-hypotheses regarding how the clusters compare on the clustering variables of the LPS Primary sc ale, LPS Secondary Scale, Behavioral Inhibition System Scale (BIS), Behavioral Activation System Scale (BAS) and Trait Anxiety as measured by the State Trait Anxi ety Inventory (STAI); and c) The aggression analyses section addresses hypotheses re garding the relationship between the psychopathy clusters and aggres sion. First, scores on the Aggression Questionnaire (AQ)
65 Are compared across clusters. Then the pr evalence of instrument al (versus reactive) aggression across clusters with different patt erns of psychopathic features is examined. Preliminary Analyses Validity Protocol validity was determined by using the validity scales of the PPI. Lilienfeld (personal communication, 2002) suggest ed a cut off point of three standard deviations above the sample mean, where any score above this cut off would be considered invalid. All particip ants in this sample met this standard and were included in the analyses. The PPI was developed to measure the core personality traits of psychopathy and was used in this study as a construct validati on measure for the LPS. As shown in Table 6, as predicted, the pattern of correlations between these PPI Total and the LPS Total, Primary and Secondary scales respectively, were in the positive direction, ( r= .49, .49, and .23).
66 Table 6. Pearson Correlations Between all Study Measures Note: *Correlation is significan t at the 0.05 level (1-tailed). **Correlation is significant at the 0.01 level (1-tailed). Scale Descriptions Table 7 presents the means, standard deviations, and ranges of scores on all measures. The scores and distributions obt ained in the current study were comparable with those reported in other st udies with college student samples for the PPI (Lilienfeld, S.O., personal communication, September 30, 2004), LPS (Levenson, et al., 1995), BIS/BAS (Carver & White, 1994), STAI (Ga udry et al., 1975), and AQ (Buss and Perry, 1992). LPS Total LPS Primary LPS Secondary PPI Total BIS Total BAS Total STAI Trait AQ Total LPS Total 1 LPS Primary .88** 1 LPS Secondary .66* .23* 1 PPI Total .49** .49** .23* 1 BIS Total .01 -.13 .24* -.28** 1 BAS Total -.02 -.07 .07 .16 .15 1 STAI Total .49** .29** .55** .18 .42** -.03 1 AQ Total .58** .41** .56** .38** .28** .17 .54** 1
67 Table 7 Means, Standard Deviations, and Ranges on Measures SCALE Mean Std. Deviation Range PPI 449.84 35.23 373Â–533 (160) LPS 52.70 9.45 32Â–72 (40) Primary 31.54 7.31 18Â–47 (29) Secondary 21.16 4.54 10Â–33 (23) BIS 19.29 3.21 12Â–28 (16) BAS 41.53 4.58 31Â–52 (21) STAI 40.03 9.93 22Â–66 (44) AQ 64.65 13.32 37Â–94 (57) Correlation Analyses Correlations were calculated among all the measures, and are presented in Table 6. One-tailed tests of significance were us ed as there were directional hypotheses for specific correlations. Both zero-order and partial correlations were calculated. The Primary and Secondary psychopathy scales of the LPS were significan tly correlated with each other (r = .23, p.<.05), but they were expect ed to have different associations with some criterion measures (e.g., BIS/BAS, STAI). Statistical suppression obscures the differences in the relationships between tw o related scales and other variables. Therefore, as shown in table 8, to consider the unique sources of variance in the relationships between the psychopathy scales and other variables, partial correlations were conducted controlling for the relati onship between the LPS Primary and LPS Secondary scales (McHoskey et al., 1998). Next, T-tests were used to compare
68 correlations, and differences between relations of the LPS subscales to clustering and aggression variables were calculated (see table 8). Table 8 Zero Order and Partial Correlations, and Differences between Correlations Measure LPS Primary Zero Order (Partial) LPS Secondary Zero Order (Partial) Test For Differences Between Zero Order Correlations BIS Total -.13 ( -.20*) .24* (.28**) t (93) = 3.02** BAS Total -.07 (-.09) .07 (.09) t (93) = -1.09 STAI Trait .29** ( .20) .55** ( .52**) t (93) = 2.40* AQ Total .40** ( .33**) .56** ( .53**) t (93) = 1.53 Note: *Correlation is signifi cant at the 0.05 level (1-tailed). **Correlation is significant at the 0.01 level (1-tailed). Following the research considering the th eories discussed ea rlier regarding the factors of psychopathy, it was expected that the LPS Primary scale would be negatively correlated with the BIS and unr elated to the BAS. This pattern of correlations was obtained, although the zero-order correlation wi th the BIS was not significant. However, when the effect of the LPS Secondary s cale was removed, the LPS PrimaryÂ’s negative correlation with the BIS became significant. Counter to Hare (1990) and Harpur et al. (1989) findings that Factor 1 was negatively related to anxiety, th ere was a significant positive correlation between the LPS Primary and the STAI. Removing the effects of LPS Secondary reduced this as sociation from .29 to .20, alth ough this partial correlation
69 approached significance (p = .053). The pos itive correlation with the AQ scale was consistent with expectations that Factor 1 traits are related to aggressive behavior (Cornell, 1996). The correlations for the LPS Secondary with the BIS and the BAS did not conform to theoretical expecta tions. A significant positive association with BAS was not obtained even when the effect of the LPS Pr imary scale was contro lled. The significant positive correlations for LPS Secondary scal e with measures of trait anxiety and aggression were positive, as expected, alt hough the positive associa tion with the BIS was not consistent with theory. Finally, as shown in the right hand colu mn of Table 8, tests for differences between correlations with LPS Primary and LPS Secondary were as expected except for BAS, where the anticipated higher correlati on for LPS Secondary was not obtained. Aggression Aggression Coding When participants were asked to repo rt acts of aggre ssion, all ninety-six participants reported at least one aggressive act. About 98% (N = 94) reported 2 acts, 91% (N = 87) reported 3 acts, 70% (N = 67) re ported 4 acts, and 33% (N = 32) reported 5 acts. After reporting past acts of general aggression, participan ts were specifically asked to report any acts of instrumental aggressi on, and 51% (N = 50) of the participants
70 reported an additional act of aggression13. All reported aggressive acts were determined to be Instrumental or Reactive based on the ra tings of two independent judges. When the aggressive acts were coded a nd rated as either instrument al or reactive aggression, one subject did not provide enough information to be rated and was dropped from the aggression analyses. Of the remaining 95 pa rticipants, 38.5% (n=37) of subjects were determined to have used instrumental a ggression (Instrumental Aggressors) and 60.4% (n=58) were classified as Reactive Aggresso rs. Instrumental acts typically took place when Â“trying to earn the respectÂ” of peer gr oups, but also were reported for more severe incidents as well. Typically, reactive incidents were argumen ts with significant others, family members or disputes with strangers or acquaintances. The percent agreement for the two independent raters was 85%, with a Kappa coefficient of .70, indicating a substantial agreement (Landis & Koch, 1977). A third rater was employed as a tie breaker for the cases where the primary raters did not agree. The two aggression groups were compar ed on demographic characteristics in order to determine the existence of any poten tially confounding variab les. Descriptively, the instrumental group was composed of 70.3% (n=26) Caucasians, 18.9% (n=7) African Americans, and 10.8% (n=4) other. The m ean age of the instrumental group was 22 years (SD =5.82, range 18-46). One person in the Reactive group did not report his racial group. Of the 57 participants that reporte d race, the Reactive group consisted of 50% (n=29) Caucasians, 23% (n=13) African Amer icans, and 26% (n=15) other. Only 57 13Not all of these acts were actually coded as instru mental. Only those that fit the criteria for an instrumental act were coded as instrumental.
71 participants (98%) in this gr oup reported their age, and of that group the mean age was 21 years (SD = 3.5, range 18-34). Analyses indi cated no group differences on age, F (1,93) =1.2, p. < .28, or on racial background, 2 (1, N = 94) = 4.27, p. = .12, CramerÂ’s V = .213. Primary Analyses Cluster Analyses The LPS subscales, together with STAI tr ait anxiety scale and BAS/BIS variables, were subjected to model-based cluster anal yses. Whereas some clustering programs attempt to force all subjects into clusters, mclust includes a Â“robust clusteringÂ” option (Noise T) that permits individua l Â“outliersÂ” to be identified and to remain apart (i.e., as 1member clusters) from primary clusters in seeking the optimal solution. An initial analysis using robust clusteri ng revealed that no individu als were considered to be outliers. Therefore, the standard mclust method (Noise F) for cl ustering all participants was used in calculating the cluster analysis. Table 9 presents the AWE statistics from the model-based cluster analysis. The best solution was obtained using model S, which assumes the clusters to be elliptical, and lets the orientation vary while holding the si ze and shape constant. Th e best fitting model was the four cluster solution (AWE = 602.184). The difference between the AWE values communicates the posterior odds, that is, give n the data, the probability that the second model is correct, over the proba bility that the first model is correct. A difference of at least 10 is very strong evidence of the best fi tting model Raferty (1995 ). This difference reflects an odds ration of 150:1 where the mo re positive AWE value represents the better
72 Table 9 AWE Values for Alternative Cluster Models Cluster Characteristics Number of Clusters Model Shape Volume Orientation12 3 4 5 6 7 8 9 S Ellipsoidal Same Different 0353.93 585.22 602.18 525.72 579.41 428.92 337.42 293.13 Sum of Squares Spherical Same None 048.15 43.44 38.94 26.05 13.62 1.02 -18.63 -45.66 Determinant Ellipsoidal Same Same 0-12.74 3.71 -9.57 -11.23 -32.93 -48.18 -77.12 -106.07 S* Ellipsoidal DifferentDifferent 0-52.41 -100.92-151.09 -204.06-262.52-328.21 -399.6 -465.02 Spherical Spherical DifferentNone 0-70.94 -144.25-212. 66 -277.16-342.60-410.18 -475.42 -536.63 Unconstrained Different DifferentDifferent 0-188.75-378.78 557.65 -723.33-890.00-1058.52 -1221.41-1373.41
73 fitting model. In the current study, all other solutions were represented by significantly lower AWE scores (AWE < 585.218; -16.966). Table 10 presents standardized Z-scores (m=0. SD = 1) and raw score means for each of the clustering variables, as well as the results of a MANOVA comparing the clusters on each of the clustering variables. Figure 3 illustrates the relative level and shape of the mean z-score profile s for each of the four clusters. Relative to other participants those in Cluster 1 had re latively higher scores on the LPS Primary scale, moderate scores on the LPS Secondary scale, and about average scores on the BIS, BAS, and STAI. Although the BIS score is somewhat higher than would be required for a clear match with th e theoretical Primary psychopathy profile pattern shown in Figure 1, this group still ha s the second lowest me an BIS score and the profile is otherwise suggestive of that s ubtype. Accordingly, those participants were referred to as the Primary Psychopathic Traits group. Relative to the other groups, those in Cluster 2 had average scor es on the LPS Primary scale and higher scores on the LPS Secondary scale. They also had relatively higher scores on the BIS, BAS, and STAI. The elevated BIS score is inconsistent with th e theoretical Secondar y Psychopathy profile shown in Figure 214. Accordingly, this group was referred to as the Secondary Psychopathic Traits group. The participants in Cluste r 3 had relatively lower scores on both psychopathy scales, BIS, BAS, and the STAI, and this group was referred to as the No Psychopathic TraitsÂ–Low Anxiety group. Finally, those in Cl uster 4 had relatively low 14 Although, consistent with these data, in a previous study with college students, McHoskey et al. (1998) reported that the highest LPS Secondary scor es were associated with elevations on both BIS and BAS.
74 scores on both psychopathy meas ures, moderate to high scores on the BIS and BAS, but lower scores on the STAI, and were called the No Psychopathic TraitsÂ–Normal Temperament group. Table 10 Differences between the Clusters on Clustering Variables Psychopathic Traits No Psychopathic Traits Type Primary (N=41) Z raw Secondary (N=16) Z raw Low Anxiety (N=13) Z raw Normal Temperament (N=26) Z raw F values .69 36.56 -.01 31.44-.32 29.23-.92 24.85 LPS Primary M SD .74 5.4 1.047.60 .81 5.9 .51 3.72 F (3,92) = 24.77**, eta square = .45 .32 22.61 .68 24.25-.90 17.08-.48 19.00 LPS Secondary M SD .87 3.9 .79 3.6 .91 4.1 .84 3.8 F (3,92) = 12.84**, eta square = .30 .17 18.73 1.06 22.69 1.17 15.54 .21 19.96 BIS Total M SD .77 2.5 .71 2.3 .84 2.7 .84 2.7 F (3,93) = 20.39**, eta square = .40 .09 41.10 .53 43.94 -.45 39.46 .05 41.77 BAS Total M SD .98 4.5 1.0 4.7 .76 3.5 1.03 4.7 F (3,93) = 2.63 p. =.06, eta square = .08 .13 41.32 1.1351.25-.92 30.92-.44 35.65 STAI Trait Anxiety M SD .69 6.9 1.04 10.3 .65 6.4 .84 8.3 F (3,93) = 19.64**, eta square = .39 Note: *Significant at the 0.05 level. **Significant at the 0.01 level
75 Figure 3. Mean Z-Scores for Four Clusters Table 11 reflects the demographic charac teristics of the four clusters. The No Psychopathic TraitsÂ–Low Anxiety (n = 13) had one participant that did not report his racial group and in the No Psychopathic TraitsÂ–Normal Temperament (n=26), one participant did not report his age. Also, as shown in Table 11, the 4 clusters were compared on demographic characteristi cs. ANOVA results indicated no group differences on age, F(3,94) = .336, p = .80 Howe ver, there were differences in racial diversity, p = .042, CramerÂ’s V = .26, with different racial -2.00 -1.00 0.00 1.00 2.00LPS Prim a ry LPS Secondary BIS BAS Tra i t A n xi e tyz-scores Cluster 1 Cluster 2 Cluster 3 Cluster 4
76 compositions between Clusters one and four p = .03, CramerÂ’s V = .33. Table 11 Demographic Characteristics of Four Clusters Cluster N Age Race Primary Psychopathic Traits (cluster 1) 41 Mean = 21.3 SD = 3.72 Range 18-34 Caucasian = 63.4% (n=26) African American = 26.8% (n=11) Other = 9.8% (n=4) Secondary Psychopathic Traits (cluster 2) 16 Mean = 20.6 SD = 1.86 Range 18-25 Caucasian = 62.5% (n=10) African Americans = 6.3% (n= 1) Other = 31.3% (n=5) No Psychopathic TraitsÂ–Low Anxiety (Cluster 3) 13 Mean = 21.6 SD = 6.27 Range 18-42 Caucasians = 46.2% (n=6) African American = 38.5% (n=5) Other = 7.7% (n=1) No Psychopathic TraitsÂ–Normal Temperament (Cluster 4) 26 Mean = 22 SD = 5.94 Range 18-46 Caucasian = 53.8% (n=14) African American = 11.5% (n= 3) Other = 34.6% (n=9) F(3,94) = .336, p = .80 p = .042, CramerÂ’s V = .26, A one-way MANOVA was conducted to cons ider the differences between the clusters on the dependant vari ables, LPS Primary, LPS Seconda ry, BIS, BAS, and STAI. The omnibus test showed signi ficant differences between th e clusters on the dependant variables, WilksÂ’ = .214, F (15,243) = 12.14, p < .001, eta squared = .40. ANOVA were used as follow-up tests to cons ider the between-groups effects on the individual variables. Post hoc analyses of the univariate ANOV A were conducted using the Bonferoni method
77 to compare the clusters on each of the clus tering variables. As shown in Table 10, all main effects were significant except for the BAS. The main effect for LPS Primary was significant, F (3,92) = 24.77, p. < .001, eta squared = .45, where the Primary Psychopathic Traits group had a higher mean than all other clusters. The Secondary Psychopathic Traits group was statistically similar to the No Psychopathic TraitsÂ–Low Anxiety group The No Psychopathic Traits-Normal Temperment group was significantly lower than the other clusters on the LPS Primary scale. The main effect for the LPS Secondary was significant, F (3,92) = 12.84, p. < .001, eta squared .30. However, unlike the LPS Primary, there was no significant difference between the Primary Psychopathic Traits group and the Secondary Psychopathic Traits group. The No Psychopathic Traits groups were also statistically similar to each other, but they were significantly lower than the Psychopathic Traits groups. The ANOVA results for the BIS were significant F (3,92) = 20.39, p. < .001, eta squared .40. Post hoc analyses revealed that the Secondary Psychopathic Traits group had the highest BIS scores. The No Psychopathic TraitsÂ– Normal Temperment and the Primary Psychopathic Traits group were similar to each other, and significantly lower than the Secondary Psychopathic Traits group. The No Psychopathic Trait Â– Low Anxiety group was significantly lower than all the ot her groups. No significant main effects emerged on the BAS scale, F (3,92) = 2.6, p. =.06, eta squared .40. There were significant difference be tween the clusters on the STAI, F (3,92) = 19.64, p. < .001, eta squared .39. The Secondary Psychopathic Traits group had the
78 highest STAI scores. The Primary Psychopathic Traits groupÂ’s STAI scores were significantly lower. While the two No Psychopathic Traits groups were lower than the Psychopathic Traits groups, they were not statistically different from each other. Aggression Analyses Next cluster differences on the AQ were examined. As shown in Table 12, a oneway ANOVA indicated there we re significant differences, F (3,92)=9.72, p. < .001, eta squared .24. For the AQ Total there was no significant difference between the Primary Psychopathic Traits group and the Secondary Psychopathic Traits group. There was a Table 12 Differences between the Clusters on Aggression Group AQ Total M SD Number of Aggressive Acts M SD Primary Psychopathic Traits 68.27a 12.18 4.10 1.02 Secondary Psychopathic Traits 73.25a 14.03 3.81 .98 Nonpsychopathic Traits Â– Low Anxiety 53.85b 9.57 3.92 1.12 Nonpsychopathic Traits Â– Normal Temperment 59.04b 10.61 3.69 .97 F Values F (3,92)=9.72, p. < .001, eta squared .24 F (3,92)=.942, p. =.436 eta squared .03 Note. Means that share superscripts are not si gnificantly different and those that have differing superscripts differ at p < .05 using Bonferoni as a post hoc test.
79 significant difference between the Psychopathic Traits groups and the No-Psychopathic Traits groups. While lower than the Psychopathic groups, the No Psychopathic Traits groups were also statistically similar to each other. All participants were asked to describe at least three aggressive acts. They were then specifically asked about Instrumental aggression. When the reported number of acts were compared across clusters, there were no differences found (see Table 12). Figure 4 displays the proportion of indi viduals in the two psychopathic traits groups who were classified as instrumental versus reactive aggressors. A Chi-square analysis revealed that the groups differed significantly, Pearson p = .026, CramerÂ’s V = .30. Consistent with the final hypothesis, instrumental aggression was preferentially associated with a cl uster of individuals identified as the Primary Psychopathic Traits group, whereas proportionately fe wer individuals (18.8% versus 51.2%) from the Secondary Psychopathic Traits group self-reported involvement in instrumental aggression. Similar to the Secondary Psychopathic Traits group, instrumental aggression was rare among members of the No-Psychopathic-Normal Temperament cluster (32%) and the Non-psychopathic-Low Anxiety cluster (38.5%).
80 48.8 51.2 81.3 18.8 0 10 20 30 40 50 60 70 80 90 Primary Secondary N=20 N = 21 N=13 N = 3 Psychopathic Traits GroupsPercent of Instrumental and Reactive Aggressors in Each Cluster Reactive Aggressors Instrumental Aggressors Figure 4. Proportions of Instrument al Versus Reactive Aggression for Psychopathic Traits Groups
81 DISCUSSION The construct of psychopathy has a long-st anding history in the psychological literature. Psychopathy has been found to be related to many importa nt social outcomes, such as poor treatment amenability, viol ence and recidivism (Cunningham & Reidy, 1998; Hart & Hare, 1998). However, inconsis tencies have been found in the research regarding correlates, etiology, and treatment of psychopathy. These inconsistencies, along with theoretical writings and the multifactor structure of instruments designed to assess the construct, indicate th at psychopathy is essentially a heterogeneous construct. Although theoretical writings ha ve indicated the existence of subtypes of psychopathy, to date, little empirical resear ch has been conducted to conf irm the existence of these subtypes. The current study represents the first known attempt to empirically test the validity of the subtypes of psychopathy in an adult population. More specifically, the aim of the study was to determine if the constr uct of psychopathy is better understood as a unitary construct, or as two distinct s ubtypes referred to as Primary and Secondary Psychopathy (Cleckley, 1941, Karpman, 1948, L ykken, 1995), and if those subtypes use different types of aggression. In order to test these hypotheses, th eoretical writings on subtypes were used as a basis for determin ing the specific variables that would best discriminate between variants: scores that assess different facets of phenotypic psychopathy (i.e., LPS Primary and Secondary s cales), indicators of temperament (i.e., the BIS and BAS scales), and a nxiety (i.e., the STAI). Mode l-based cluster analysis was
82 then used to investigate the existence of subtypes analogous to Primary and Secondary psychopathy in college students. In order to evaluate the construct va lidity of the emerging subtypes, several results were expected; a) Higher scor es on both scales of the LPS for the Psychopathy groups than the Non-psychopathy groups. Those in the Primary Psychopathy group would have high scores on the LPS Primary s cale and moderate to high scores on the LPS Secondary scale. Those in the Secondary Psychopathy cluster would have higher scores on the LPS Secondary scale than the LPS Primar y scale; b) Low scores on the BIS scale and normal scores on the BAS measure for the Primary Psychopathy group. High scores on the BAS measure and normal to high scores on the BIS scales for the Secondary Psychopathy group were; c) Low scores on the STAI trait scale for the Primary Psychopathy group, and high for the Secondary Psychopathy group; d) Higher aggression scores on the AQ for the psychopathy clusters than the non-psychopa thy clusters; e) A higher proportion of instrumental aggre ssion to hostile aggression used by the Primary Psychopathy group, then the Secondary Psychopathy group and the N on-Psychopathic groups Four clusters emerged with cluster prof iles differing in theoretically coherent ways. Two of the clusters were representa tive of sub-clinical Primary and Secondary psychopathy and the other two represented non-psychopathic groups. As expected, the Psychopathic Traits groups used more aggression than the Non-psychopathic traits groups, and the Primary Psychopathic Traits group used more instrumental aggression than the Secondary Psychopathic Traits group. Overall, thes e results support the
83 existence of sub-clinical subtypes of psyc hopathy that resemble, in meaningful ways, hypothetical clinical variants. The results al so suggest that subtyping may have clinical and forensic utility in risk assessment. Primary Psychopathic Traits Group Primary psychopaths have been theorized to have more personality traits of psychopathy (i.e., grandiosity, lack of emotions, lack of guilt, manipulativeness, deficit in conscience, lack of empat hy), have some behavioral traits of psychopathy (i.e. irresponsibility, failure to follo w a life plan), have a deficit in the ability to inhibit behaviors, and not experience anxiety (C leckley, 1941; Karpman, 1945; Lykken, 1995). The results of the cluster an alysis found a statistically significant difference in the reported psychopathic traits in the different cl usters. For example, participants in the Primary Psychopathic Traits group were more likely to re port egocentric and entitled attitudes, and were more callous, selfis h, and manipulative than the other groups. Additionally, members of this cl uster reported moderate to higher levels of impulsivity and irresponsibility. The theories of Lykken (1995), Karpma n (1941), and Cleckley (1941) all posit that primary psychopaths do not manifest trai t anxiety. In contrast, anxiety may be common in secondary psychopathy. Congruent with this idea, pa rticipants in the Primary Psychopathic Traits group had less trait anxi ety than those in the Secondary Psychopathic Traits group. However, somewhat unexpectedly, those in the Primary
84 Psychopathic Traits group did experience some level of anxiety compared to the Nonpsychopathic Traits groups. Research has not been consistent in indicating a negative correlation between psychopathy and anxiety. Theory suggests that these inconsistent re sults are due to the heterogeneity of psychopathic populations, a nd it is hypothesized that subgroups may experience fear and anxiety differently. R ecent research by Hicks et al. (In Print) supports this distinction. Hi cks and colleagues (In Print) conducted a cluster analysis study and found subtypes that resembled Primar y and Secondary psychopathic groups. Their Stable psychopath group (more like Prim ary psychopathy) showed little anxiety or otherwise negative emotionality, while thei r Aggressive group (more like Secondary psychopathy) was very high on negative emotiona lity. Other research has indicated that anxiety is negatively related to Factor 1 a nd positively related to Factor 2 (Hare, 1991; Harpur et al., 1989; Patrick et al., 1993; and Levenson et al ., 1995). In the current study, consistent with Hicks et al (In press) no relationship was found with anxiety (partial correlations). One possible reason that the co rrelation was not negati ve could be related to the measure of psychopathy used in this st udy. In studies with o ffenders (Brinkley et al., 2001) and college students (Levenson, et al., 1995) both LPS scales correlated positively with PCLR Factor 2 or the behavior al traits of psychopathy, which more often correlate positively with anxiety or other negative emotions. Thus the LPS Primary scale may not be an optimal measure of the persona lity traits of primary psychopathy. Finally, it is possible that primary psychopaths expe rience some level of anxiety, and what actually distinguishes them fr om the other groups is a low level of fear (Lykken, 1995).
85 The BIS is activated when cues for punish ment are detected, eliciting an anxiety reaction and consequential inhibition of ongoing behaviors (Patterson & Newman, 1993; Gray, 1982). Theory predicts that, while Primary psychopaths may experience a normal level of behavioral response, and therefore a normal BAS (as was shown in this sample); they have a constitutional deficit in the BI S and are less sensitive to cues for punishments or nonreward (Lykken, 1995; Fowles, 1980). Th e findings in this study lend empirical validity to this theory; the Primary Psychopathic Traits group experienced less behavioral inhibition than the Secondary Psychopathic Traits group. Given a reduced sensitivity to punishment, one would expect the Primary Psychopathic Traits group to have the least ability to inhibit behavior th an all other groups, however, when compared to the Non-psychopathic Traits groups, the Primary Psychopathic Traits group had more behavioral inhibition. Previous research (Fowles, 1980; Gray, 1982) suggests that a relationship exists between Primary psychopathy and the BIS. The LPS scale was designed with that relationship in mind, therefore a negative corr elation between the L PS Primary scale and the BIS scale would be expected. In the current study, partial correlations revealed a negative relationship between BIS and LPS prim ary. Previous research indicates that both the LPS subscales are related to antis ocial behavior (Leven son, et al., 1995), and thus there may not be a precise enough distin ction between the subscales. Partialling out the interrelationship between the scales clarifie d the correlates of the separate constructs. Overall the Primary Psychopathic Traits group displayed higher psychopathic personality traits, a normal level of behavi oral activation, and some deficit in the
86 inhibition of behavior, lendi ng empirical support to the theories of Lykken (1995), CLecklye (1941) and Karpman (1941). Counter to theories of Primary psychopathy, this sample seems to experience some degree of an xiety and behavioral inhibition. While the levels of these traits are still lower than in the Secondary Psychopathic Traits groups, they are slightly higher than the levels for groups who do not have any psychopathic traits. One reason for these findings could be related to the sample used. This study measured sub-clinical levels of psychopathic tr aits and therefore the sample is considered a Â“normalÂ” population that was not actually psyc hopathic. In forensic settings those with psychopathic traits have actually engaged in antisocial behaviors dem onstrating a lack of inhibition. In this Â“normalÂ” population ther e is no antisocial be havior and therefore automatically higher levels of inhibition then found in forensic samples. Perhaps the distinction between sub-clinical and clinical may help explain why these subjects showed some level of anxiety and consequently mo re behavioral inhibition than would be expected. Secondary Psychopathic Traits Group Secondary psychopaths are hypothesized to display more of the behavioral aspects (e.g., impulsivity, so cial deviance, emotional reactivity) of psychopathy than Primary psychopaths without necessarily possess ing the same personality traits such as glibness and charm (Lynam, Whiteside, & Jone s, 1999; Levenson, et al., 1995). They are also believed to experience anxiety and an overactive behavioral activation system (Karpman, 1941; Lykken, 1995). The current resu lts validate theory in that those in the
87 Secondary Psychopathic Traits group did demonstrate more behavior features of psychopathy, including impulsivity, lack of long-term goals, and low frustration tolerance, while exhibiting moderate levels of personality features, such as a tendency to lie, callousness and lack of remorse. Consistent with the resear ch indicating that anxiety is more strongly associated with PCL-R Factor 2 than Factor 1 (Hare, 1991; Harpur et al., 1989; Patrick et al., 1993; Levenson et al., 1995), the curr ent study found a stronger re lationship between anxiety and LPS Secondary psychopathy, a scale designe d to capture the behavioral features associated with Factor 2. Etiologicall y, researchers have posited that Secondary psychopathy stems from an underlying emotiona l problem that includes elevated anxiety (Lykken, 1995; Karpman, 1941; Lynam, et al., 1999; Levenson, et al., 1995). In the current study, those participants in the Secondary Psychopathic Traits group indeed had higher levels of anxiety than the other groups. Secondary psychopaths, despite their increa sed level of anxiety, are thought to be extremely goal-directed and to engage in excessi ve approach behavior even in the face of punishment (Lykken, 1995; Fowles, 1980). This overactive BAS was not found in the current study. Those participants in the Secondary Psychopathic Traits group had relatively higher levels of beha vioral activation, but statistically there were no differences among the groups. While Carver and White ( 1994) indicated that the BAS can be considered as a unified construct, the BAS Total scale was not satisfactory in discriminating between the factors of psychopath y or the subtypes. Additionally, in post hoc analyses considering the subscales of the BAS, no relationship was found between
88 the LPS Secondary scale and the BAS subscales of Drive (pursuit of desired goals) or Reward (the positive response to the anticipation and occurrence of rewards). However, those who had higher scores on the LPS Secondary scale tended to ha ve more desire for novel rewards and a willingness to spontane ously pursue those rewards (Fun-seeking scale). Carver and White (1994) indicate it is Â“unclear at present whether any specific manifestation of BAS functioni ng is more important than othersÂ” (p. 324), and Lykken (1995) believes that a Secondary psychopath wi ll act impulsively to achieve goals. While in this study, the Fun-seeking scale is the only discriminating subscal e, the designers of the LPS scale (Levenson et al, 1995) used a m easure of sensation seeking as an indicator of discriminant validity for the scale, as they believed there should be no relationship between psychopathy and sensation seeking. To the extent that fun seeking and sensation seeking are similar constructs, there is so me evidence that this scale should not be considered to be the critical manifestation of BAS Lykken (1995) stated that Secondary psyc hopaths will actively seek rewards, but still experience anxiety and normal levels of behavioral inhibition. In the current study, those in the Secondary Psychopathic Traits group not only experienced behavioral inhibition, but also had highe r levels than all other groups. Blackburn and Lee-Evens (1985) proposed that a strong BAS (impulsi vity) and strong BIS (a nxiety) were each necessary features of Secondary psychopat hy (McHoskey et al., 1998). The current findings are consistent with the McHoskey et al. (1998) findings that college students with high scores on both the BIS and BAS s cales had the highest scores on the LPS Secondary scale.
89 While traditional models of Secondary psychopathy do not indicate elevated levels of behavioral inhib ition, the elevated levels s een in the current study are understandable given that the BIS is associated with a nxiety reactions. Significant positive associations were found between the L PS Secondary scale, anxiety, and the BIS. Secondary psychopathy, the LPS Secondary scal e and Factor 2 of the PCL-R have all been associated with anxiety, so it follow s that behavioral inhibition is positively correlated with anxiety. Add itionally, McHoskey et al. ( 1998) provided a rationale for Secondary features being associated with both high BAS and high BIS that is consistent with the findings in the cu rrent study. They theorized that Secondary psychopaths are conflicted by strong drive (BAS) that pushes th em to test the limits of their internal constraints (high BIS), which may be associat ed with expressions of frustration. This theory might be particularly the case in st udent or community samples, whose members are generally better socialized and to have more developed internal constraints (i.e., conscience) than would be found in offender samples. The Secondary Psychopathic Traits groupÂ’s behavioral activ ation is not as uniquely high as predicted. However, this study measured psychopathic traits with a Â“normalÂ” population. Therefore, this group may not show either a uniquely high level of behavioral activation or reduc tion in BIS that would be found in a more heterogeneous population. Clusters Conclusions Overall this study provided empirical suppor t for the previously only theoretical distinction between Primary and Secondary ps ychopathy that has been postulated in the
90 literature. This study used etiological variab les and showed that psychopathy indeed is a heterogeneous construct. Additionally, the clusters generated by model-based cluster analysis in this study indicat e that even in Â“normalÂ” popul ations there are viable subclinical variants of psychopathy whose patt erns on clustering va riables substantially parallel those hypothesized in th e literature. The results f ound in this research are an important first step in deve loping a clearer understanding of psychopathy and its causes, which is a necessary step in developing more specified and appropriate treatment modalities. Aggression Comparisons of the clusters on measures of aggression validated the sub-clinical variants of psychopathy by pr edicting the types of aggr ession different types of psychopaths may use. The results of this study lend further support to the existing literature that has consistently found a pos itive association between psychopathy in general and aggression (Blackburn & Maybury, 1985; Leary, 1957; Forth et al., 1990; Hare & McPherson, 1984; Kosson et al, 1990; Rice, Harris, & Quincy 1990; Serin, 1991; Harris, Rice, & Cormier, 1991; Salekin et al., 1996), by finding, as predicted, that participants in the Psychopathic clusters were more aggressive than those in the N onpsychopathic clusters. Primary psychopaths are described as cunning, manipulative, nonanxious, guilt free, and they are believed to have a deficit in behavior inhibition. They often display a narcissistic sense of entitlement and a tendency to use others to satisfy their needs. Given these traits, Primary psychopaths are theorized to employ more instrumental aggression
91 than the other groups. Secondary psychopaths, who have more anxiety, impulsivity and an overactive BAS, are belie ved to be more reactive to provocation and to act aggressively out of anger. Therefore, because of their greater sus ceptibility to stress and higher degree of impulsivity, Secondary Psyc hopaths are believed to use more hostile aggression than instrumental aggression. Addi tionally, given that in strumental aggression is indicative of more severe pathology, non-psychopaths are t hought to use more reactive rather than instrumental a ggression. The results of th e current study support these theories. There was a difference in the pr evalence of instrumental versus reactive aggression used between the Primary Psychopathic Traits group and the Secondary Psychopathic Traits group, with the Primary Psychopathic Traits group using a greater proportion of instrumental aggression than the Secondary Psychopathic Traits group. Comparisons of the clusters on t ypes of aggression is an impor tant step in the study of the construct of aggression, but also in terms of validating the subtypes of psychopathy. The current study empirically validated the id ea that subjects in clusters created by psychopathy data do indeed differ in terms of the types of aggression used, and such differences were consistent with th e theories governing such clusters. First, the results demonstrated that even in a Â“normalÂ” sample there are viable sub-clinical variants of ps ychopathy, therefore providing so me of the first empirical support for the understanding of psychopathy as a heterogeneous construct. Second, and more importantly, this was the first attemp t to use etiological indicators along with Primary and Secondary psychopathy facet scor es to empirically derive subtypes of psychopathy (previous research focused solely on the exploration of the correlates of the
92 psychopathy factors). This method, cluster an alysis, permitted for actual derivation of multidimensional subtypes rather than limiting the research to the relationships to the factors alone. Considering a combination of variables also allowed for the ability to obtain potentially great er differentiation among the subtype s than would be possible with clustering on psychopathy facets alone. Furthermore, the current study was the firs t to consider the specific types of aggression used by specific types of psychopa ths, expanding the research by Cornell et al., (1996), who separated offenders into inst rumental and hostile groups and considered the level of psychopathy used by each group. Th is study elaborated on CornellÂ’s research by creating psychopathy clusters and using those clusters to consider differences in motivations for aggression in adults (see Christian, Frick & Hill, 1997 and Vincent Vitacco, & Grisso, 2003; for examples of cluster analytic studies on juvenile psychopathy). Additionally, Wilson et al. ( 1999) criticized previous rese arch that examined the two types of psychopathy for being limited to institutionalized samples. This was the first study of psychopathic features to use cl uster analysis with a university population, which allowed for the opportunity to demons trate the viability of subgroups that have mainly sub-clinical levels of psychopathic traits. Student populations are more comparable to the general population than are offender populations, which make the present findings potentially more generalizab le to community samples. The use of the LPS self-report measure of psychopathic tr aits, which does not include items that explicitly assess tendencies toward violence or aggression, also reduced the problem of
93 criterion contamination that exists with so me other measures. And unlike incarcerated samples, these participants were not subjected to the negative effects of institutionalized life (Widom, 1977). In short, the current research is a pion eering effort to look at psychopathy in a non-institutionalized sample, to use variables that theoreti cally define psychopathy to define the clusters, and to consider the di fferent types of aggr essions used by those variants. Limitations and Future Research While this study was able to utilize a non-clinical sample to consider psychopathic subtypes and types of aggres sion, there were certain limitations. Measurement Issues Originally several self-report measures of psychopathy were considered for inclusion in this research. The LPS was c hosen because it had been used in several studies with student and community samples, and it offered the most promise given the objectives of the research. At the time of study design, it was the only self-report measure that reliably demonstr ated a two-factor solution cons istent with the theories of primary and secondary psychopathy (Lynam et al., 1999; McHoskey, et al., 1998). Although the results of the current research generally supported the hypotheses, the LPS scales did not always correlate with the other variables in the ways that would necessarily be expected and some of the re lationships of the variables to the clusters did not clearly fit theory.
94 Previous research has indicated the there are some limitations to the discriminant validity of the LPS scales (Lynam, et al., 1999; Levenson et al., 1995) ,where the Primary psychopathy scale seems to be a better measur e of the behavioral and social deviance traits of psychopathy than of primary psychopa thy. An alternative measure that future investigators might employ is the PPI (Lilienfeld & Andrews, 1996). The PPI was designed to target the co re personality tra its of psychopathy (Lilienfeld, et al., 1996). In the current study, the PPI wa s administered as a validation measure, which allowed for some post hoc anal yses to be performed. First, the PPI and LPS totals scores were correla ted at r= .49, which is lower th an what would be expected between two self-report measures of the same construct. A recent study by Lilienfeld and Skeem (2004) reported a correlation of r = .69 between the LPS and PPI in a large offender sample, and similarly strong relati onships have been f ound between self-report measures of psychopathic tra its in juveniles (r = .80, p<.01; Falkenbach, Poythress, & Heide, 2003). When the relationship between the cluster variables and the PPI was considered, the results were consistent with theory. The PPI and the STAI were not correlated (r = .18, ns) and the expected ne gative relationship betw een the PPI and the BIS scale was found (r= -.21, p < .05). Perhap s the PPI captures better the features of primary psychopathy, as it appears to be a mo re effective means of discriminating the inhibitory and anxiety defi cits associated with primary psychopathy than the LPS Primary psychopathy scale. Additionall y, since this study was completed, Benning, Patrick, Hicks, Blonigen, and Krueger (2003) have structurally analyzed the PPI and found it to contain two factor s (PPI-I and PPI-II). Lilienfel d and Skeem (2004) compared
95 the two factors of the PPI with the scales of the PCL-R and LPS. These authors found that the PPI-I displayed better discriminant validity for antiso cial behavior than did the LPS Primary scale. They also found that the PPI-I, but not the LPS Primary scale, demonstrated the theoretical negative relati onship with anxiety. Confirmatory factor analysis is needed for the PPI and a comparison of its ut ility for identifying psychopathy types to that the ability of the LPS. Howe ver, based on the results of the current study, future researchers should consider that the PPI factors may be a better means of distinguishing Primary and Secondary psychopathy. Other Issues Previous research found adequate base ra tes of psychopathic traits in college students (Lynam, et al. 1999), however, the curr ent study had several findings that were not as clear as anticipated. Specifically some of the relationships between the variables were not as expected. Research findings are often easier to interpret when the base rates for the variables of interest are highe r. While the evaluation of ps ychopathic traits and subtypes is important in non-forensic samples, the res earch on subtypes is relatively new. More research is needed in offender and forensic samples where there are higher base-rates and there is more available research on the relatio nships of the variables used. Additionally, perhaps some of the associations among variab les differ across more socialized, Â“normalÂ” samples and forensic groups. It will be in teresting to compare correlates of clusters derived from non-clinical sample with those of clusters derived from offender samples in which clinically meaningful levels of psychopa thic traits are more prevalent. Previous
96 research indicates that psychopathic tra its are more common among men then women (Levenson, et al., 1995). The current study utilized a male only samp le in order to in crease the likelihood of finding aggressive and psychopathic traits ; however, there is a need to evaluate psychopathy and aggression in female populatio ns. The constructs and relationships between the variables may be different acros s genders and future research should focus on both men and women.
97 REFERENCES Aldenderfer, M.S. & Blashfield, R.K. ( 1984). Cluster Analysis. Beverly Hills: Sage Publications. Alterman, A. I., McDermott, P.A., Cacciola, J.S., Rutherford, M.J., Boardman, C.R., McKay, J.R., & Cook, T.G. (1998). A typology of antisociality in methadone patients. Journal of Abnormal Psychology, 107 (3), 412-422. American Psychiatric Associ ation (1994). Diagnostic a nd Statistical Manual of Mental Disorders (Fourth ed.). Washington: Amer ican Psychiatric Association. Averill, L. A. (1982). Recollections of Clarc's G, Stanley Hall. Journal of the History of Behavioral Sciences, 18 (4), 341-346. Bandura (1973). Aggression: A social l earning analysis. Englewood Cliffs, NJ: Pretice Hall. Benning, S. D., Patrick, C. J., Hicks, B. M., Blonigen, D. M., & Krueger, R. F. (2003). Factor structure of the psychopath ic personality inve ntory: Validity and implications for clinical asse ssment. Psychological Aassessment 15(3), 340-350. Berkowitz, L. (1962). Aggression: A soci al psychological analysis. New York: McGraw-Hill. Berkowitz, L. (1989). Frustration-aggr ession hypothesis: Examination and reformulation. Psychological Bulletin 106, 59-73. Berkowitz, L. (1990). On the formation a nd regulation of anger and aggression: A cognitive-neoassociationistic an alysis. American Psychologist, 45 (4), 494-503.
98 Blackburn, R. (1975). An empi rical classification of psychopathic personality. British Journal of Psychiatry, 127 456-460. Blackburn. R. (1979). Psychopathy and pers onality: The dimens ionality of selfreport and behavior ratings da ta in abnormal offenders. Br itish Journal of Social and Clinical psychology 18, 111-119. Blackburn, R. & Coid, J.W. (1998). Psychopathy and the dimensions of personality disorder in violent offenders Personality and Individual Differences, 25, 129-145. Blackburn, R., Lee-Evans, J. Michael. (1985). Reactions of primary and secondary psychopaths to anger-evoking s ituations. British J ournal of Clinical Psychology, 246(2), 93-100. Blackburn, R., & Maybury, C. (1985). Identify ing the psychopath: The relation of Cleckley's criteria to the interpersonal dom ain. Personality & Indi vidual Differences, 6(3), 375-386. Blashfield, R. K., Aldenderfer, Mark S. (1988). The methods and problems of cluster analysis. In J. R. Nesselroade Raymond B. (Ed.), Handbook of multivariate experimental psychology (2nd ed.). Pers pectives on individual differences. New York, NY: Plenum Press. Bodholt, R.H., Richards, H.R., & Gacano, ,C.B. (2000). Assessing psychopathy in adults: The psychopathy checklist-revised and screening version. In C. B. Gacano (Ed.), The Clinical and Forensic Assessmen t of Psychopathy: A Practitioners Guide (pp. 111-136). Mahwah, NJ: Lawrence Erlbaum Associates.
99 Brinkley, C. A., Schmitt, W. A., Smith, S. S. ((2001). Construct validation of a self-report psychopathy scale: Does LevensonÂ’s self-report psychopathy scale measure the same constructs as HareÂ’s PsychopathyChe cklist-Revised? Persosnality & Indiidual Differences 31(7), 1021-1038. Brown, J. S., Farber, I. E. (1951). Emo tions conceptualized as intervening variables--with suggestions toward a theory of frustration. Psychological Bulletin 48, 465-495. Bushman, B. J., Anderson, C.A. (2001). Is it time to pull the plug on hostile versus instrumental aggression dichotomy? Psychological Review 108(1), 273-279. Buss, A. H. (1961). The psychology of aggression New York: John Wiley & Sons, Inc. Buss, A. H., & Durkee, A. (1957). An inve ntory for assessing different kinds of hostility. Journal of Consulting Psychology, 21 343-349. Buss, A. H., Perry, M. (1992). The A ggression Questionnaire. Journal of Personality & Social Psychology 63(3), 452-459. Carver, C. S., White, T.L. (1994). Behavi oral inhibition, behavioral activation, and affective responses to impending rewa rd and punishment: The BIS/BAS Scales. Journal of Personality & Social Psychology 67(2), 319-333. Cattell, R.B., & Scheier, I.H. (1963). Handbook for the IPAT Anxiety Scale (2nd edition). Champaign, IL: Institute for Personality and Ability Testing.
100 Christian, R.E., Frick, P. J., Hill, N. L. (1997). Psychopathy and conduct problems in children: II. Implications for subtyping children with conduct problems. Journal of the Â“American Academy of Child & Adolescent Psychiatry 36(2), 233-241. Cleckley. (1941). The mask of sanity; an attempt to reinterp ret the so-called psychopathic personality St. Louis: C.V. Mosby. Cleckley. (1976). The mask of sanity; an attempt to reinterpret the so-called psychopathic personality (5th edition) St. Louis: C.V. Mosby. C ooke, D. J., & Michie, C. (1997). An item response theory analysis of the Hare psychopathy checklist. Psychological Assessment, 9 3-13. Cooke, D. J., Kosson, D. S., & Michie, C. (2001). Psychopathy and ethnicity: Structural, item, and test generalizability of the Psychopathy Checklist-Revised (PCL-R) in Caucasian and African American pa rticipants. Psychological assessment 13(4), 531542. Cornell, D. G., Warren, J., Hawk, G. Stafford, E., Oram, & Pine. (1996). Psychopathy in instrumental and reactive vi olent offenders. Journal of Consulting & Clinical Psychology, 64(4), 783-790. Cunningham, M. D., & Reidy, T.J. (1998). Antisocial personality disorder and psychopathy: Diagnostic dilemmas in classi fying patterns of antisocial behavior in sentencing evaluations. Behavior al Sciences and the Law, 16 333-351. Dengerink, H. A., Bertilson, ,H.S. (1975). Psychopathy and physiological arousal in an aggressive task. Psychophysiology, 12 (6), 682-684.
101 Dodge, K. A. (1991). The structure a nd function of reac tive and proactive aggression. In D. J. Pepler, Ed; Rubin, Kenneth H., Ed (Ed.), The development and treatment of childhood aggression (pp. 201-218). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Dollard, J., Doob, L. W. Miller, N. E. Mo wrer, O. H. Sears, R. R. Ford, C. S., Hovland, C. I. & Sollenberger, R. T. (1939). Frustra tion and aggression New Haven: Yale University Press. Erhlich, D.J. & Malmao, R.B. (1967) Electrophysiological concomitants of simple operant conditioning in the rat. Neurophychologia, 5, 219-235. Eron, L.D., & Huesman, L.R. (1990). The st ability of aggressive behavior--even unto the third generation. In Lewis, M., Mi ller, S.M. (Eds). Handbook of developmental psychopathology. Perspectives in developmenta l psychology. New York: Plenum Press. Falkenbach, D. M.,Poythress, N. G., Heid e, K.M. (2003). Psychopathic features in a juvenile diversion populat ion: Reliability and predictiv e validity of tw o self-report measures. Behavioral Sciences & the Law 21(6), 787-805. Feshback. (1964). The function of aggressi on and the regulation of the aggressive drive. Psychological Review, 71, 257-272. Forth (1992). Emotion and Psychopathy: A Three Component Analysis. Unpublished Doctoral Dissertation, University of British Columbia, Vancouver, Canada. Forth, A. E., Brown, S.L., Hart, S.D., & Hare, R.D. (1996). The assessment of psychopathy in male and female noncriminals : Reliability and va lidity. Personality & Individual Differences, 20 (5), 531-543.
102 Forth, A. F., Hart, S.D., & Hare, R.D. (1990). Assessment of psychopathy in male young offenders. Psychological Assessment, 2 342-344. Fowles, D. C. (1980). The three arousal m odel: Implications of GrayÂ’s two-factor learning theory for heart rate, elec trodermal activity and psychopathy. Psychophysiology, 17 87-104. Fraley, C. (1998). Algorithms for mode l-based Gaussian hierarchial Clustering. SIAM Journal of Scientific Computing 20(1), 270-281. Frick, P.J., OÂ’Brian, B.S., Wooton, J.M., Mc Burnett, K. (1994). Psychopathy and conduct problems in children. Journal o f Abnormal Psychology 103, 700-707. Gaudry, E., Vagg, P.R., & Spielberger,C.D. (1975). Validation of the state-trait distinction in anxiety research. Multivariate Behavior Research, 10, 331-341. Glueck, S., & Glueck, E. (1950). Unraveling Juvenile Delinquency. Oxdford, England: Commonwealth Fund. Gray, J. A. (1982). The Neuropsychol ogy of Anxiety: an Enquiry into the Functions of the Septo-Hippocampal System New York, NY: Clarendon Press/Oxford University Press, pp. 548. Gray, J. A. (1985). The neuropsychology of anxiety. Issues in Mental Health Nursing, 7 (1-4), 201-228.
103 Green, S.B., Salkind, N.J., & Akey, T.M. (2000). Using SPSS for Windows: Analysing and Understanding Data. New Jersey: Prentice Hall. Haapasalo, J., & Pulkkinen, L. (1992). The psychopathy checklis t and non-violent offender groups. Criminal Beha viour & Mental Health, 2 (4), 315-328. Hamburger, M.E., Lilienfeld, S.O., & Hogben, M. (1996). Psychopathy, gender, and gender roles: Implications for antisocial a nd histrionic personality disorders. Journal of Personality Disorders 10, 41-55. Hare, R. D. (1972). Psychopathy and phys iological responses to adrenalin. Journal of Abnormal Psychology, 79 (2), 138-147. Hare, R. D. (1978). Electrodermal and car diovascular correlat es of psychopathy. In R. D. Hare, and Schalling, D. (Ed.), Ps ychopathic disorder: approaches to research (pp. 107-144). Hare, R. D. (1980). A research scale fo r assessment of psychopathy in criminal population. Personality and I ndividual Differences, 1 111-119. Hare, R. D. (1983). Diagnosis of antisoci al personality diso rder in two prison samples. American Journal of Psychiatry (149), 887-890. Hare, R.D. (1985). Comparison of proce dures for the assessment of psychopathy. Journal of Consulting and Clinical Psychology 53(1), 7-16. Hare, R.D. (1991). The Hare Psycho pathy Checklist Revised. Toronto: Multi-Health Systems.
104 Hare, R. D. (1996). Psychopathy: A clin ical construct whose time has come. Criminal Justice and Behavior, 23 (1), 25-54. Hare, R.D. (2003). The Hare Psychopathy Checklist Â– Revised; Second Edition. Toronto: Multi-Health Systems. Hare, R. D., Frazelle, J., & Cox, D.N. (1978). Psychopathology and physiological responses to threats of an aver sive stimulus. Psychopsysiology, 15 165-172. Hare, R. D., Harpur, T.J., Hakstain, A. R., Forth, A.E., & Hart, S.D., & Newman, J.P. (1990). The revised psychopathy checkli st: Reliability and factor structure. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 2 (3), 338341. Hare, R. D., and Jutai, J.W. (1983). Cr iminal history of the male psychopath: Some preliminary data. In K. Van Dusen & S. Mednick (Eds.), Prospective Studies of Crime and Delinquency. (pp. 23-32) Boston: Klener Mijhoff. Hare, R. D., & McPherson, L.M. (1984). Violent and aggressive behavior by criminal psychopaths. International J ournal of Law & Psychiatry, 7(1), 35-50. Hare, R. D., & Quinn, M.J. (1971). Psychopathy and autonomic conditioning. Journal of Psychology, 77, 223-235. Harpur, T. J. Hakstain, A.R., Hare, R.D. (1988). Factor structure of the Psychopathy Checklist. Journal of Consu lting & Clinical Psychology, 56(5), 741-747. Harpur, H., & Hare, R.D. (1991). Psychopa thy and violent behavior: Two factors are better than one. August, Paper presented at the meeting of APA, San Francisco.
105 Harpur, T. J., & Hare, R.D. (1994). Asse ssment of psychopathy as a function of age. Journal of Abnorma l Psychology, 103(4), 604-609. Harpur, T. J., Hare, R.D., & Hakstain, A. R. (1989). Two factor conceptualization of psychopathy construct validity and a ssessment implications. Psychological Assessment. A journal of Consulting and Clinical Psychology, 1(1), 6-17. Harris, G. T., Rice, M.E., and Cormier, C.A. (1991). Psychopathy and violent recidivism. Law and Huma n Behavior, 15(6), 625-637. Harris, G. T., & Comier, M.E. (1995) The effectiveness of therapeutic communities for psychopath":Reply. Therapeuti c Communities: International Journal for Therapeutic & Supportive Or ganizations, 16(2), 147-149. Hart, S. D., & Dempster, R.J. (1997). Im pulsivity and psychopathy. In C. D. Webster, Ed; Jackson, Margaret A., Ed (Ed.), Impulsivity: Theory, assessment, and treatment. (pp. 212-232). New Yo rk, NY: The Guilford Press. Hart, S. D., & Hare, R.D. (1997). Psychopa thy: Assessment and association with criminal conduct. In D. M. Stoff, Ed; Breiling, James, Ed; et al (Ed.), Handbook of antisocial behavior. (pp. 22-35). New Yor k, NY, US: John Wiley & Sons, Inc,. Hart, S. D., & Hare, R.D. (1998). Disc rimination validity of the psychopathy checklist in a forensic psychiatric popul ation. Psychological Assessment, 1(3), 211-218. Hart, S. D., Kropp, P.R., & Hare, R.D. (1988). Performance of male psychopaths following conditional release from prison. Journal of Consulting and Clinical Psychology, 56(2), 227-232.
106 Hemphill, J. F., Hare, R.D., & Wong, S. (1998). Psychopathy and recidivism: A review. Legal & Criminol ogical Psychology, 3(1), 139-170. Herve, H. F., Yong hui Ling, J, Hare, R. D. (2000). Criminal Psychopathy and its subtypes. Paper presented at the American Psychology-Law Society, New Orleans, ,LA. Hicks, B. M., Markon, K. E., Patrick, C. J ., Krueger, R. F. (In Press). Identifying psychopathy subtypes based on personality st ructure. Psychological Assessment. Hinton, J.W., O'Neill, M.T. (1978). Pilot research on psychophysiological response profiles of maximum security hospita l patients. British J ournal of Social & Clinical Psychology, 17(1), p 103. Huesman, L.R., Eron, L.D., Lefkowitz,,M.M ., & Walder, L.O. (1984). Stability of aggression over time and generations. Developmental Psychology, 20(6), 1120-1134 Hundleby, J. D., & Ross, B.E. (1977). Co mparison of measures of psychopathy. Journal of Consulting & Clin ical Psychology, 45(4), 702-703. Kagan, J. (1992). YesterdayÂ’s premises tomorrowÂ’s promises. Developmental Psychology, 28(6), 990-997 Karpman, B. (1941). On the need for se parating psychopathy into two distinct clinical types: symptomatic and idio pathic. Journal of Criminology and Psychopathology, 3, 112-137. Karpman, B. (1946). Psychopathy and the scheme of human typology. The Journal of Nervous and Me ntal Disease, 103, 276-288.. Karpman, B. (1948) The myth of psychopath ic personality. American Journal of Psychiatry, 104, 523-535
107 Karpman, B. (1955). Criminal psychodynamics : a platform. Archives of Criminal Psychodynamics, 1, 3-100. Kosson, D. S., Kelly, D.S., White, J.W. ( 1997). Psychopathy related traits predict self reported sexual aggression among college men. Journal of Interpersonal Violence, 12(2), 241-254. Kosson, D. S., Smith, S.S., & Newman, J. P. (1990). Evaluating the construct validity of psychopathy in Black and White male inmates: Three preliminary studies. Journal of Abnormal Psychology, 99(3), 250-259. Kraepelin. (1907). Clinical Psychiatrtry. (A.R. Diefendorf, Trans.) New York: Macmillan. Landis, J.R. & Koch, G.G. (1977). The measurement of observer agreement for categorical data. Bi ometrics, 33, 159-174. Leary, T. (1957). Interpersona l diagnosis of per sonality ; a functional theory and methodology for personality evaluation. Levenson, M. R. (1992). Rethinking ps ychopathy. Theory & Psychology, 2(1), 51-71. Levenson, M. R., Kiehl, K.A., & Fitzpatrick, C.M. (1995). Assessing psychopathic attributes in a noninstitutiona lized population. Journal of Personality and Social psychology, 68(1), 151-158. Lilienfeld, S. O. (1994). Conceptual pr oblems in the assessment of psychopathy. Clinical Psychology Review, 14(1), 17-38.
108 Lilienfeld, S. O., & Andrews, B.P. (1996). Development and preliminary validation of a self-report measure of psyc hopathic personality traits in noncriminal populations. Journal of Personality Assessment, 66(3), 488-524. Lilienfeld, S.O.,& Skeem, J.L. (2004, Marc h). Psychometric pr operties of selfreport measures of psychopathy. In N.G. P oythress (Chair), Contemporary Issues in Psychopathy Research. Sym posium conducted at the meeting of the American Psychology-Law Society, Scottsdale, AZ. Livesley, W.J. (1995). Phenotypic and genot ypic structure of psychopathic traits. Paper presented at the NATO Ad vances Study Institute of Psyc hopathy, Alvor, Portugal. In Millon, T et al (eds). Psychopathy Antisocial, Criminal and Violent Behavior, p. 260. Loeber, R. (1982). The stability of antis ocial and delinquent child behavior: A review. Child Development, 53(6), 1431-1446. Lykken, D. T. (1957). A study of anxiety in the sociopathic personality. Journal of Abnormal & Social Psychology. 55, 6-10. Lykken, D. T. (1995). The Antisocial Pe rsonalities. Hillsdale, NJ: Lawrence Erlbaum. Lynam, D.R., Whiteside, S. & Jones, S. (1999). Self-reported psychopathy: A validation study. Journal of Pers onality Assessment, 73(1), 110-132. Magnusson, B. (1988). Individual developmen t from an interactional perspective: A longitudinal Study. Hillsdale, NJ: Erlbaum.
109 Magnusson, B., & Bergman, L.R. (1988). Individual and variable based approaches to longitudinal research on early risk factors. In Rutter, Ed. Studies of Psychosocial Risk. The Power of Longitudi nal Data, pp. 45-61. Cambridge: Cambridge University Press. McCord, W., & McCord, J. (1964). The Psychopath: an Esay on the Ciminal Mind. Oxford, England: D Van Nostrand, xi, pp. 223. McHoskey, J. W., Worzel, W., & Szyart o, C. (1998). Machiavellianism and psychopathy. Journal of Personality & Social Psychology, 74(1), 192-210. Megargee, E.I. (1967). Development a nd validation of an MMPI scale of assaultiveness in overcontrolled individua ls. Journal of Abnormal Psychology, 72(6), 519-528. Megargee, E. I. (1982). Psychological determinants and correlates of criminal violence. in M. E. W. N. A,. Weiner (E d.), Criminal Violence (pp. 81-170). London: Cassel. Meloy, J. R. (1988). The Psychopathic Mind : Origins, Dynamics, and Treatment. Northvale, NJ: Jason Aronson, Inc., xix, pp. 474. Meloy, J.R. (1992). On the predictability of violent behavior: Considerations and guidelines: Discussion. Journal of Forensic Sciences, 37(4), 949-950. Miller, N.E. (1941). The frustrationaggression hypothesis. Psychological Review, 48, 337-342. Millon, T., & Davis, R.D. (1996). Disord ers of Personality: DSM-IV and Beyond (2nd Edition). New York: Wiley.
110 Monahan, J., Steadman, H.J., Silver, E., Appelbaum, P.S., Clark Robbins, P., Mulvey, E.P., Roth, L.H., Grisso, T., & Ba nks, S. (2001). Rethinking Risk Assessment: The MacArthur Study of Mental Disorder and Violence. Oxford: Oxford University Press. Newman, J. P., & Kosson, D.S. ( 1986). Passive avoidance learning in psychopathic and nonpsychopathic offenders Journal of Abnormal Psychology, 95(3), 252-256. Newman, J. P., & Schmitt, W.A. (19 98). Passive avoidance in psychoopathic offenders: A replication and extension. Journal of Abnormal Behavior, 107(3), 527-532. Newman, J. P., Widom, C.S., & Nath an, S. (1985). Passive avoidance in syndromes of disinhibition: Psychopathy a nd extraversion. Journa l of Personality & Social Psychology, 48(5), 1316-1327. Novoco 1975 (hostility scale) Ogloff, J. R., & Wong, S. (1990). Electrode rmal and cardiovascu lar evidence of a coping response in psychopaths. Criminal Justice and Behavior, 17, 231-245. Ogloff, J. R., Wong, S., Greenwood, A. ( 1990). Treating criminal psychopaths in a therapeutic community program. Beha vior Sciences and the Law, 8, 181-190. Overall, J.E., & Magee, K.N. (1992). Replication as a rule for determining the number of clusters in hierar chical cluster analysis. App lied Psychological Measurement, 16, 119-128. Parke, R. D., Ewall, W., Slaby, R. G., ( 1972). Hostile and helpful verbalizations as regulators of nonverbal aggression. J ournal of Personality & Social Psychology
111 23(2), 243-248. Parke, S. & Slaby (1983). Hostile and helpful verbalizations as regulators of nonverbal aggression. Journal of Personal ity & Social Psychology, 23(2), p 243-248. Patrick, C.J. (1994). Emotion and Ps ychopathy: Startling new insights. Psychophysiology, 31, 415-428. Patrick, C.J. (1995). Emotion and temperme nt in Psychopathy. Clinical scientce, fall 5-Psychophysiology, 31, 415-428. Patrick, C. J. Berthot, B.D. (1995). Star tle potentiation duri ng anticipation of a noxious stimulus. Active vs. passive response sets. Psychophysiology, 32, 72-80. Patrick, C. J., Bradley, M.M., & Lang, P.J. (1993). Emotion in the criminal psychopath: Startle reflex modulation. Journal of Abnormal Psychology, 102, 82-92. Patrick, C.J., Cuthbert, B.N., & Lang, P.J. (1994). Emotion in the criminal psychopath: Fear image processing. Journa l of Abnormal Psychology, 103(3), 523-534. Patrick, C. J., & Zempolich, K.A. ( 1998). Emotion and aggression in the psychopathic personality. Aggression & Violent Behavior, 3(4), 030-338. Patterson, M. C., Newman, J.P. (1993). Reflexivity and learning form aversive events: toward a paychological mechan ism for the syndromes of disinhibition. Psychological Review, 100, 716-736. Pinel. (1806). A Treatise on Insanity. (D. Davis translator) New York: Hafner. Pulkkinen, L. (1987). Offensive and defensive aggression in humans: A longitudinal perspective. Aggre ssive Behavior, 13(4), 197-212
112 Raftery, A. (1995). Bayesian model selec tion in social resear ch. Socialogical Methodology, 25, 111-163. Rapkin, B. D., & Luke, D.A. (1993). Cluster analysis in community research: Epistemology and practice. American J ournal of Community Psychology, 21(2), 247277. Rice, M. E., Harris,G.T., & Quincy, V. L. (1990). A follow-up of rapists assessed in a maximum-security psychiatric facility. Journal of Interpersonal Violence Dec 1990, 5(4), 435-448. Roberts, L. (1974). Comparative ps ychophysiology of the electrodermal and cardiac control systems. In P.A. Obrist, A. H. Black, J. Brener, & L.V. DiCara (Eds.). Cardiovascular Psychophysiology-Current Issu es in Response Mechanisms, Biofeedback and Methodology. Chicago: Aldine. Rogers, R., & Dion, K.L. (1991). Reth inking the DSM III--R diagnosis of antisocial personality disorder. Bulletin of the American Academy of Psychiatry & the Law, 19(1), 21-31. Salekin, R. T., Rogers, R., & Sewell, K.W. (1996). A review and meta-analysis of the Psychopathy Checklist and Psychopathy Ch ecklist--Revised: Predictive validity of dangerousness. Clinical Psychology: Science & Practice, 3(3), 203-215. Sampson, R. J., & Laub, J.H. (1993). Crime in the making: Pathways and turning points through life. Schmauk, F. J. (1970). Punishment, arousal and avoidance learning in sociopaths. Journal of Abnormal Psychology, 76, 325-335.
113 Schmitt, W.A., & Newman, J.P. (1999). Are all psychopathic individuals low anxious? Journal of Abnormal Psychology, 108(2), 353-358. Serin, R. C. (1991). Psychopathy and vi olence in criminals. Journal of Interpersonal Violence, 6(4), 423-431. Serin, R. C. (1996). Violent recidivism in criminal psychopaths. Law and Human Behavior(20), 207-217. Serin, R. & Amos, N.L. (1995). The role of Psychopathy in the assessment of dangerousness. International Journa l of Law and Behavior, 18, 231-238. Seto, M.C. & Lalumiere (2000). Psychopathy and sexual Guide (pp. 87-109). Mahwah, NJ: Lawrence Erlbaum Associates. ag gression. In C. B. Gacono (Ed.), The Clinical and Forensic Assessment of Psychopathy: A Practitioners Skeem, J. L., Mulvey, E.P, Grisso, T. (2003). Applicability of traditional and revised models of psychopathy to the Psychopathy Checklist: Screening Version. Psychological Assement 15(1),41-55. Skeem, J. L., Poythress, N., Edens, J. F., Lilienfeld, S.O., & Cale, E.M. (2003). Psychopathic personality or personalities? Exploring potential va riants of psychopathy and their implications for risk assessment. Aggression & Violent Behavior. Vol 8(5), 513-546. Speilberger, C.D. (1985). Assessment of st ate and trait anxiety: Conceptual and methodological issues. The S outhern Psychologist, 2, 6-16.
114 Speilberger, C.D., Gorsuch, R.L., & Lushene, R.D. (1970). STAI Manual for the State Â–Trait Anxiety Inventory (Â“Self-Eva luation QuestionnaireÂ”). Palo Alto, CA: Consulting Psychologists Press. Steuerwald, B. L., Kossen, D.S. (2000). Em otional experiences of the psychopath. In C. B. Gacano (Ed.), The Clinical a nd Forensic Assessment of Psychopathy: A Practitioners Guide (pp. 111-136). Mahwa h, NJ: Lawrence Erlbaum Associates. Taylor, J.A. (1953). A personality scal e of manifest anxiety. Journal of Abnormal Social Psychology, 53, 303-320. Tellegen, A. (1978/1982). Brief Manual for the Multidimensional Personality Questionnaire. University of Minnesota. Tengstrom, A., Grann, M., Langstrom, N., Kullgren, G. (2000). Psychopathy (PCL-R) as a predictor of violent reci divism among criminal offenders with schizophrenia. Law and Huma n Behavior, 24(1), 45-57. Toch, H., Adams, K. & Grant, J.D. ( 1989). Coping: Maladaptation in prisons. New Brunswick, NJ: Transaction Publishers, xxxi, pp 288. Tryon, R.C. & Baily, D.E. (1970). Cluste r Analysis. New York: McGraw-Hill. Vincent, G. M., Vitacco, M. J., Grisso, T. (2003). Subtypes of adolescent offenders: Affective traits and antisocial beha vior patterns. Behavi oral Sciences & the Law, 21(6), 6985-712. Vitale, J.E., Smith, S.S., Brinkley, C.A ., Newman, J.P. (2002). The Reliability and Validity of the Psychopathy Checklist-Revi sed in a Sample of Female Offenders. Criminal Justice and Behavior vol. 29, no. 2, pp. 202-231.
115 Wallace, J.F., Schmitt, W.A.., Vitale, J.E., & Newman, J.P.(2000). Experimental investigations of information-processing defi ciencies in psychopath s: Implications for diagnosis and treatment In C. B. Gacano (Ed.), The Clinical and Forensic Assessment of Psychopathy: A Practitioners Guide (pp. 87-109). Mahwah, NJ: Lawrence Erlbaum Associates. Widiger, T.A., & Corbitt, E.M. (1993). Antisocial personality disorder: Proposals for DSM-IV. Journal of Persona lity Disorders, 7(1), 63-77. Widom, C.S. (1977). A methodology fo r studying no institutionalized psychopaths. Journal of Consulting & Clinical Psychology, 45(4), 674-68 Williamson, S. E., Hare, R., & Wong, S. (1987). Violence: Criminal psychopaths and their victims. Canadian Jour nal of Behavioral Sciences, 19, 454-462. Wilson, D. L., Frick, P.J., & Clements, C.B. (1999). Gender, somatization and psychopathic traits in a college sample. Journal of Psychopathology and Behavioral Assessment, 21(3), 221-235. Wilson, J.Q. & Herrnstein, R.J. (1985). Crime and Human Nature. New York, NY: Simon & Schuster, pp. 639. Zagon, I. K., & Jackson, H. J. (1994). Construct validity of a psychopathy measure. Personality & Indi vidual Differences, 17(125-135).
117 Appendix A Demographics What is you gender? ____Male ____Female What is your marital status? ____ Single ____ Divorced ____ Married ____ Separated What is your age? ___________ What is your race? (Check all that apply) ____ African American _____ Caucasian _____ Other (please specify) ________ What is your ethnicity? _____ Hispanic _____other (please specify) _________________ What is your average family income per year? _____$0-$9,999 ______$10,000-19,000 _____$20,000-29,999 ______$30,000-39,999 _____$40,000-59,999 ______$60,000 and higher Do you currently drink? ____ yes ____ no If you answered yes, what age did you start drinking? ______ How many drinks per week? __________ per month?___________ Do you currently use drugs? ____ yes ____ no Have you ever used drugs? ____ yes ____ no If you answered yes, what age did you start using drugs? _____ How much do you consume per week? ________ per month?______________ Have you ever been arrested? ____yes ____no How much time did you spend in jail?_______________ Have you been convicted of a crime? _____yes _____no Do you have any psychiatric/psychological history? ____ Inpatient hospitalization ____ Outpatient care or services ____ No psychiatric/ps ychological history
118 Appendix B LPS Directions: A number of statements are given belo w. Read each statement and then circle the answer that indicates your leve l of agreement with that statement (Rate statements according to the scale provide d below). There are no right or wrong answers. Do not spend too much time on any one statement 1. Success is based on survival of the fittes t; I am not concerned about the losers. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 2. For me, whatÂ’s right is what ever I can get away with. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 3. In todayÂ’s world, I feel justified in doing anything I can get away with to succeed. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 4. My main purpose in life is ge tting as many goodies as I can. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 5. Making a lot of money is my most important goal. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 6. I let others worry about hi gher values; my main concer n is with the bottom line Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 7. People who are stupid enough to ge t ripped off usually deserve it. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 8. Looking out for myself is my top priority. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly
119 9. I tell other people what they wa nt to hear so that they wi ll do what I want them to do. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 10. I would be upset if my success came at someone elseÂ’s expense. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 11. I often admire a really clever scam. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 12. I make a point of trying not to hur t others in pursuit of my goals. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 13. I feel bad if my words or actions cause someone else to feel emotional pain. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 14. Even if I were trying very hard to sell something, I wouldnÂ’t lie about it. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 15. Cheating is not justified becau se it is unfair to others. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 16. I find myself in the same kinds of trouble, time after time. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 17. I am often bored. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 18. I find that I am able to pur sue one goal for a long time. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly
120 19. I donÂ’t plan anything very far in advance. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 20. I quickly lose interest in tasks I start. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 21. Most of my problems are due to the fact that other people just donÂ’t understand me. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 22. Before I do anything, I carefully co nsider the possible consequences. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 23. I have been in a lot of shou ting matches with other people. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 24. When I get frustrated, I often Â“l et of steamÂ” by blowing my top. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 25. Love is overrated. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 26. I enjoy manipulating other peopleÂ’s feelings Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly
121 Appendix C BIS/BAS Scales Directions : A number of statements are given below. Read each statement and then circle the answer th at indicates your leve l of agreement with that statement (Rate statements according to the scale provided be low). There are no right or wrong answers. Do not spend too much time on any one statement. 1. If I think something unpleasant is goi ng to happen I usually get pretty Â“worked up.Â” Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 2. I worry about making mistakes. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 3. Criticism or scolding hurts me quite a lot. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 4. I feel pretty worried or upset when I think or know somebody is angry at me. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 5. Even if something bad is about to happe n to me, I rarely experience fear or nervousness. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 6. I feel worried when I think I have done poorly at something. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 7. I have very few fears compared to my friends. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 8. When I get something I want, I feel excited and energized. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly
122 9. When IÂ’m doing well at something, I love to keep at it. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 10. When good things happen to me, it affects me strongly. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 11. It would excite me to win a contest. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 12. When I see an opportunity for somethi ng I like, I get excited right away. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 13. When I want something g, I usually go all-out to get it. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 14. I go out of my way to get things I want. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 15. If I see a chance to get somethi ng I want, I move on it right away. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 16. When I go after something I use a Â“no holds barredÂ” approach Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 17. I will often do things for no other reason than that they might be fun. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 18. I crave excitement and new sensations. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly
123 19. IÂ’m always willing to try something new if I think it will be fun Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly 20. I often act on the spur of the moment. Disagree Disagree Agree Agree Strongly Somewhat Somewhat Strongly
124 Appendix D STAI Self-Evaluation Questionnaire STAI Form Y-1 Directions : A number of statements which people have used to describe themselves are given below. Read each statement and then ci rcle the appropriate number to the right of the statement to indicate how you feel right now that is, at this mo ment (Rate statements according to the scale provided below). There are no right or wrong answers. Do not spend too much time on any one statement but give the answer, which seems to describe your present feelings best. 1= not at all 2=somewhat 3=moderately so 4=very much so 1. I feel calm......................................................................................... 1 2 3 4 2. I feel secure...................................................................................... 1 2 3 4 3. I am tense.......................................................................................... 1 2 3 4 4. I feel strained.................................................................................... 1 2 3 4 5. I feel at ease...................................................................................... 1 2 3 4 6. I feel upset......................................................................................... 1 2 3 4 7. I am presently worrying over possible misfortunes.......................... 1 2 3 4 8. I feel satisfied.................................................................................... 1 2 3 4 9. I feel frightened................................................................................. 1 2 3 4 10. I feel comfortable.............................................................................. 1 2 3 4 11. I feel self-confident............................................................................ 1 2 3 4 12. I feel nervous..................................................................................... 1 2 3 4 13. I am jittery......................................................................................... 1 2 3 4 14. I feel indecisive................................................................................. 1 2 3 4 15. I am relaxed....................................................................................... 1 2 3 4 16. I feel content...................................................................................... 1 2 3 4 17. I am worried ..................................................................................... 1 2 3 4 18. I feel confused................................................................................... 1 2 3 4
125 38. I take disappointments so keenly that I can't put them out of my mind.. 1 2 3 4 39. I am a steady person........................................................................ 1 2 3 4 40. I get in a state of tension or turmoils as I think over my recent concern and interests................................................................ 1 2 3 4
126 Self-Evaluation Questionnaire STAI Form Y-2 Directions: A number of statements which pe ople have used to describe themselves are given below. Read each statement and then ci rcle the appropriate number to the right of the statement to indicate how you generally f eel (Rate statements according to the scale provided below). There are no right or wrong answers. Do not spend too much time on any one statement but give the answer which seems to describe how you generally feel. 1=almost never 2=sometimes 3=often 4=almost always 1. I feel pleasant.................................................................................... 1 2 3 4 2. I feel nervous and restless................................................................. 1 2 3 4 3. I feel satisfied with myself................................................................ 1 2 3 4 4. I wish I could be as happy as others seem to beÂ…Â…Â…Â…Â…Â…Â….. 1 2 3 4 5. I feel like a failure........................................................................Â…. 1 2 3 4 6. I feel rested....................................................................................Â… 1 2 3 4 7. I am "calm, cool, and collected"....................................................... 1 2 3 4 8. I feel that difficulties are pi ling up so that I cannot overcome them.... 1 2 3 4 9. I worry too much over something that really do esn't matter............ 1 2 3 4 10. I am happy........................................................................................ 1 2 3 4 11. I have disturbing thoughts................................................................ 1 2 3 4 12. I lack self-confidence....................................................................... 1 2 3 4 13. I feel secure...................................................................................... 1 2 3 4 14. I make decisions easily.................................................................... 1 2 3 4 15. I feel inadequate.............................................................................. 1 2 3 4 16. I am content..................................................................................... 1 2 3 4 17. Some unimportant thought runs through my mind and bothers me... 1 2 3 4 18. I take disappointments so keenly that I can't put them out of my mind.. 1 2 3 4 19. I am a steady person........................................................................ 1 2 3 4 20. I get in a state of tension or turmoils as I think over my recent concern and interests................................................................ 1 2 3 4
127 Appendix E Aggression Questionnaire Directions : Read each statement and then circle the appropriate number to the right of the statement to indicate wh ether the statement describes you or not. Rate statements according to the scale provided below. There are no right or wrong answers. Do not spend too much time on any one statement but give the answer, which seems to describe you best. 1= extremely uncharacteristic of me 2=mostly uncharacteristic of me 3=somewhat characteristic of me 4=extremely characteristic of me 1. When people annoy me, I may tell them what I thi nk of them........... 1 2 3 4 5 2. I have trouble controlling my temper................................ ............Â….. 1 2 3 4 5 3. I have threatened people I know...................................................Â….. 1 2 3 4 5 4. At times I feel I have gotten a raw deal out of life.......................Â….. 1 2 3 4 5 5. Once in a while I can't control the urge to strike another person...Â…. 1 2 3 4 5 6. My friends say that I'm somewhat argumentative..........................Â… 1 2 3 4 5 7. When frustrated, I let my irritation show......................................Â…. 1 2 3 4 5 8. I am sometimes eaten up with jealousy.........................................Â…. 1 2 3 4 5 9. I can't help getting into arguments when people disagree with meÂ… 1 2 3 4 5 10. I get into fights a little more than the average person..................Â…. 1 2 3 4 5 11. Some of my friends think I'm a hothead......................................Â…. 1 2 3 4 5 12. When people are especially nice, I wonder what they want.......Â….. 1 2 3 4 5 13. I tell my friends openly when I disagree with them....................Â….. 1 2 3 4 5 14. Given enough provocation, I may hit another person..................Â…. 1 2 3 4 5 15. I wonder why sometimes I feel so bitter about things.................Â…. 1 2 3 4 5 16. I flare up quickly but get over it quickly.....................................Â….. 1 2 3 4 5 17. If somebody hit me, I hit back.....................................................Â…. 1 2 3 4 5 18. I know that "friends" talk about me behind my back...................Â…. 1 2 3 4 5 19. I have become so mad that I have broken things.........................Â…. 1 2 3 4 5 20. I am an even-tempered person.......................................................... 1 2 3 4 5
128 21. Other people always seem to get the breaks................................. 1 2 3 4 5 22. I often find myself disagreeing with people.................................. 1 2 3 4 5 23. There are people who pushed me so far that we cam e to blows... 1 2 3 4 5 24. I sometimes feel that people are laughing at me behind my back... 1 2 3 4 5 25. I sometimes feel like a powder keg ready to explode................... 1 2 3 4 5 26. If I have to resort to violence to protect my rights, I will.............. 1 2 3 4 5 27. I am suspicious of overly friendly strangers............... ................... 1 2 3 4 5 28. Sometimes I fly off the handle for no good reason......................... 1 2 3 4 5 29. I can think of no good reason for ever hitting a pe rson.................. 1 2 3 4 5
129 Appendix F Aggressive Incidents Repor t form and Coding Form Appendix F Aggressive Form Think of the times when you were aggressive towards someone. Using the following definition for aggression, please describe at least three of your mo st aggressive actions (more space is provided on the back of the page). For each act include a brief description of the behaviors, the motivation, who it was against, and your age at the time of the incident. Aggression: any physical (e.g. shoving, hitting) or ve rbal (e.g. arguing, shouting, screaming) behavior carried out with the intention of delivering an unpl easant action to someone else (e.g. Family members, significant other, friend, st ranger, etc.). Aggressive acts can be in response to a provocation, including insults, threats, or other acts that cau se frustration or anger, or aggression can be to obtain a goal such as power, money, sexual gratification, or some other objective beyond inflicting injury on the victim. Aggressive example # 1: Relationship: Stranger Age: 19 I wanted tickets to a concert so I waited in line for several hours. When they opened the ticket counter up it got a little chaotic. I was worried that I would not get the tickets so I shoved someone in line for concert tickets in order to get to the front of the line. Aggressive example # 2: Relationship: Stranger Age: 24 I was driving on the highway and I was cut off by another car. I got angry and began to curse at the driver in the other car when they cut me off. Aggressive example # 3: Relationship: Player on opposite team Age: 15 I was playing in a team softball game and we down by one run. I was on third base and I kept thinking that I had to score no matter what in order for us to tie the game up. When the batter hit the ball, I ran and the catcher was in the way of home plate. I ran right into her and knocked her over to get to home plate and score. Aggressive example # 4: Relationship: girlfriend Age: 22 I was arguing with my girlfriend and we were both saying some pretty mean things. She started calling me Â“stupidÂ” and it really pissed me off. I grabbed her arm and she hit me with her other hand. I was so pissed that I hit her in the arm. Finally I just left and slammed the door.
130 1. Please describe your most aggressive incident (please continue on back of page if you need more space): Relationship: Age: a. What motivated you to act aggressively? What happened just before you acted aggressively? Was your aggressive behavi or in response to anything? Please explain. b. Did you plan or prepare for the aggressive action, or was it more spontaneous? Please explain. c. How did you feel at the time of the aggressi ve act? What, if any emotions did you experience?
131 2. Please describe your second mo st aggressive incident: Relationship: Age: a. What motivated you to act aggressively? What happened just before you acted aggressively? Was your aggressive behavi or in response to anything? Please explain. b. Did you plan or prepare for the aggressive action, or was it more spontaneous? Please explain. c. How did you feel at the time of the aggressi ve act? What, if any emotions did you experience?
132 3. Please describe your third mo st aggressive incident (p lease continue on back of page if you need more space): Relationship: Age: a. What motivated you to act aggressively? What happened just before you acted aggressively? Was your aggressive behavi or in response to anything? Please explain. b. Did you plan or prepare for the aggressive action, or was it more spontaneous? Please explain. c. How did you feel at the time of the aggressi ve act? What, if any emotions did you experience?
133 4. Please describe any other aggressive incidents: Relationship: Age: a. What motivated you to act aggressively? What happened just before you acted aggressively? Was your aggressive behavi or in response to anything? Please explain. b. Did you plan or prepare for the aggressive action, or was it more spontaneous? Please explain. c. How did you feel at the time of the aggressi ve act? What, if any emotions did you experience?
134 Instrumental Aggression: Someone who uses instrumental aggression acts to obtain a r eadily apparent goal such as power, money, sexual gratificati on, or some other objective beyond inflicting injury on the victim. Some examples of instrumental aggression include 1) verbally abusing or physically hurting someone to impress your frie nds; 2) in a basketball game, punching or hurting someone to gain control of the ball. Ph ysical or verbal aggr ession during rape or date rape is almost always instrumental. Instrumental aggression is initiated as a means to an end rather than as an act of retaliati on or self-defense. It is usually unprovoked and is not delivered out of rage or anger. In strumental aggression of ten involves planning or preparation. However, in some cases instrume ntal aggression can i nvolve relatively little planning. Other than the incidents described ab ove, have you ever engaged in any INSTRUMENTALLY AGGRESSIVE act. Please describe any additional instrumentally aggressive behaviors based on the format from above: a. What motivated you to act aggressively? What happened just before you acted aggressively? Was your aggressive behavi or in response to anything? Please explain. b. Did you plan or prepare for the aggressive action, or was it more spontaneous? Please explain. c. How did you feel at the time of the aggressi ve act? What, if any emotions did you experience?
135 On the Answer Sheet (p. 3), rate your aggressive behaviors on the specific aspects of aggression described below. Within the parenth eses are examples of each level, but these examples do not exhaust the possibilities. Ra te each aspect of aggression independently, regardless of scores on the other aspects. Planning How much did you plan or prepare for the aggressive action? Consider both the length of time involved in preparation and the amount of preparatory activity. 4 Â– Extensive Planning (detailed plan or preparation, rehearsal) 3 Moderate planning (contemplation of action for more than 24 hours) 2 Â– Some planning (action within 24 hours, some plan or preparation) 1 Â– Very little or no planning (act during argument or fight, no preparation) Assign a (1) to actions which are part of a contiguous event, such as a brief pause during an argument. Assign a (2) if there is a break in the argument where you leave the scene of an argument and return later in the day. Goal-Directedness How much were you motivated by an external incentive, goal, or objective beyond just responding to provocation or threat? Readily apparent goals include money, power, sexual gratification, or some other external goal or benefit. Do not include such goals as self defense, escaping harm, taking revenge for previous aggression, or acting out of frustration. 4 Â– unequivical goal-directedness 3 Â– Primary goal-directedness with presence of other motives 2 Â– Secondary goal-directedness, in presence of other primary motives 1 Â– No apparent goal-directedness (motive to injure victim, retaliate, defend) Provocation Did the victims actions provoke your aggression? Include provocation that occurred prior to the incident (e.g. prior abusive treatment or confrontation) 6 Â– Exceptionally strong provocation (repeated assault, severe abuse) 5 Â– Very strong provocation (assault) 4 Â–Strong provocation (breakup of a romantic relationship, threat of major life change) 3 Â– Moderate Provocation (serious argument or dispute, threat of assault) 2 Â– Mild provocation (insult, minor argument, confrontation with others) 1 Â– No apparent provocation Arousal How much arousal, especially anger, did you experience at the time of the aggressive act? Just code your mental state, not attitude towards the victim. 4 Â– Enraged, furious, described as Â“out of controlÂ” or Â“irrationalÂ” 3 Â– Angry, mad, extremely frightened 2 Â– Excited, very nervous, anxious 1 Â– Calm or tense at most Arousal at the (4) level is extraordinary, and should be short duration Relationship with Victim Code the degree of contact or closeness between you and the victim. Code based on duration and closeness of relationship. 5 Â–Very close relationship (immediate family member, romantic partner) 4 Â– Close relationship (friend, relative, dating partner, etc.) 3 Â– Specific relationship (co-worker, person in one of your classes, etc.) 2 Â– Acquaintance 1 Â– Stranger
136 ANSWER SHEET Rater 1 Please complete the following ratings of your aggressive acts. Circle the most appropriate level of each aspect of aggression based on the descriptions listed below. Aggressive Act Planning GoalDirectedness Provocation Arousal Relationship with Victim Act #1 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #2 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #3 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #4 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #5 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1
137 ANSWER SHEET RATER 2 Please complete the following ratings of your aggressive acts. Circle the most appropriate level of each aspect of aggressi on based on the descriptions listed below. Aggressive Act Planning GoalDirectedness Provocation Arousal Relationship with Victim Act #1 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #2 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #3 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #4 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #5 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1
138 ANSWER SHEET RATER 3 Please complete the following ratings of your aggressive acts. Circle the most appropriate level of each aspect of aggressi on based on the descriptions listed below. Aggressive Act Planning GoalDirectedness Provocation Arousal Relationship with Victim Act #1 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #2 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #3 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #4 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1 Act #5 4 3 2 1 4 3 2 1 6 5 4 3 2 1 4 3 2 1 5 4 3 2 1
139 Appendix G PPI PERSONALITY STYLES INVENTORY This test measures differences in personali ty characteristics among people-that is, how people differ from each other in their persona lity styles. Beginning on the next page, read each item carefully, and d ecide to what extent it is false or true as applied to you Then mark your answer in the space provide d to the left of each item using the scale provided below. 1) False 2) Mostly False 3) Mostly True 4)True Even if you feel that an item is neither fa lse nor true as applie d to you, or if you are unsure about what response to make, try to make some response in every case. If you cannot make up your mind about the item, select the choice that is closest to your opinion about whether it is false or true as applied to you. Here's a sample item. _____I enjoy going to movies. If it is true that you enjoy going to movies, place a 4 on the line to the left of the item, as shown below. 4 I enjoy going to movies. If it is mostly false that you enjoy going to movies, place a 2 on the line to the left of the item, and so one. Try to be as honest as you can, and be sure to give your own opinion about whether each item is fa lse or true as applied to you. _____1. With one smile, I can often make someon e IÂ’ve just met interested in getting to know me better. _____2. I like my life to be unpredic table, even a little surprising. _____3. Members of the opposite sex find me Â“sexyÂ” and appealing. _____4. I am very careful and cautious when doing work involving detail.
140 _____5. Physically dangerous activities, such as sky-diving or clim bing atop high places, frighten me more than th ey do most other people. _____6. I tend to have a short temper when I am under stress. _____7. Even when others are upset with me, I can usually win them over with my charm. _____8. My table manners are not always perfect. _____9. If IÂ’m at a dull party or social gathering, I like to stir things up _____10. I weigh the pros and cons of major decisions carefully before making them. _____11. Being rich is much less important to me than enjoying the work I do. _____12. IÂ’ve always considered myself to be something of a rebel. _____13. I sometimes worry about whether I mi ght have accidentally hurt someoneÂ’s feelings. _____14. I find it difficult to make small talk with people I do not know well. _____15. I think a fair amount about my long-term career goals. _____16. I would not mind wearing my hair in a Â“mohawk.Â” _____17. I occasionally forget my name. _____18. I rarely find myself being the cente r of attention in social situations. _____19. It might be fun to belong to a group of Â“bikersÂ” (motorcyclists) who travel around the country and raise some hell. _____20. I tell many Â“white liesÂ” _____21. I often hold on to old objects or le tters just for their sentimental value. _____22. I am a good conversationalist.
141 _____23. A lot of people in my life have tried to stab me in the back. _____24. I am so moved by certain experien ces (e.g., watching a beautiful sunset, listening to a favorite piece of music) that I feel emotions that are beyond words. _____25. I often find myself resenti ng people who give me orders. _____26. I would find the job of movie stunt person exciting. _____27. I have always been extremely cour ageous in facing difficult situations. _____28. I hate having to tell people bad news. _____29. I think that it should be against the law to seriously injure another person intentionally. _____30. I would be more successful in life had I not received so many bad breaks. _____31. It bothers me (or it would bother me) qu ite a bit to speak in front of a large group of strangers. _____32. When I am faced with a decision involving moral matters, I often ask myself, Â“Am I doing the right thing?Â” _____33. From time to time I really Â“blow upÂ” at other people. _____34. Many people think of me as a daredevil. _____35. It takes me a long time to get over embarrassing or humiliating experiences. _____36. I usually feel that people give me the credit I deserve. _____37. IÂ’ve never really cared much about societyÂ’s so-called Â“values of right and wrongÂ”. _____38. If someone mistreats me, IÂ’d rather tr y to forgive him or her than get even.
142 _____39. It would bother me to cheat on an ex amination or assignment even if no-one got hurt in the process. _____40. I become deeply upset when I see pho tographs of starving people in Africa. _____41. I rarely monopolize conversations. _____42. Making a parachute jump really frighten me. _____43. At times I have been envious of someone. _____44. I become very angry if I do not receiv e special favors or pr ivileges I feel I deserve. _____45. I often find myself worrying when a friend is having serious personal problems. _____46. I pride myself on being offbeat and unconventional. _____47. Keeping in touch with old friends is very important to me. _____48. I usually strive to be the best a t whatever I do. _____49. I almost always feel very sure of myself when IÂ’m around other people. _____50. I look down at the ground whenever I h ear an airplane flying above my head. _____51. I could make an effective Â“con ar tistÂ” if the situation required it. _____52. I wouldnÂ’t mind spending my life in a commune and writing poetry. _____53. I have had Â“crushesÂ” on people that were so intense that they were painful. _____54. I like to stand out in a crowd. _____55. IÂ’m not intimidated by anyone. _____56. Before I say something, I first li ke to think about it for a while.
143 _____57. I would enjoy hitch-hiking my way across the United States with no prearranged plans. _____58. I am a guilt-prone person. _____59. I bet that it would be fun to pilot a small airplane alone. _____60. When I want to, I can usually pu t fears and worries out of my mind. _____61. Never in my whole life have I wished for anything that I was not entitled to. _____62. I generally prefer to act first and think later. _____63. I am easily flustered in pressured situations. _____64. I often make the same errors in judgment over and over again. _____65. I always look out for my own interest s before e worrying about those of the other guy. _____66. I smile at a funny joke at least once in a while. _____67. People have often criticiz ed me unjustly (unfairly). _____68. I almost always promptly return it ems that I have borrowed room others. _____69. I sometimes have difficulty standing up for my rights in social situations. _____70. If I want to, I can influence other peop le without their real izing they are being manipulated. _____71. My opinions are always completely reasonable. _____72. I become embarrassed more easily than most people. _____73. When I'm in a frightening situation, I can "turn off" my fe ar almost at will. _____74. It bothers me greatly when I see someone crying. _____75. Frankly, I believe that I am mo re important than most people.
144 _____76. I frequently have disturbing thought th at become so intense and overpowering that I think I can hear claps of thunde r or crashes of cymbals inside my head. _____77. If I do something that causes me trouble, I'm sure to avoid doing it again. _____78. I often place my friends' needs above my own. _____79. I like having my vacations carefully planned out. _____80. People whom I have trusted have often ended up "double-crossing" me. _____81. I often become deeply a ttached to people I like. _____82. I've been the victim of a lot of bad luck in my life. _____83. I have at times eaten too much. _____84. I sometimes question authority fi gures, "just for the hell of it." _____85. When my life becomes boring, I like to take some chances to make things interesting. _____86. I tend to be "thin-skinned" a nd overly sensitive to criticism. _____87. I've quickly learned from my major mistake in life. _____88. When some one is hurt by something I say or do, I usually co nsider that to be their problem. _____89. I like to dress differently from other people. _____90. If I really wanted to, I could convi nce most people of just about anything. _____91. I get restless and dissatisfied if my life becomes too routine. _____92. I generally feel that life has treated me fairly. _____93. Ending a friendship is (or would be) very painful for me.
145 _____94. When I am under stress, I often see la rge, red, rectangular shapes moving in front of my eyes. _____95. I often do favors for people even when I know that I will probably never see them again. _____96. I have sometimes "stood up" a date or a friend because something that sounded like more fun came up. _____97. I haven't thought much about what I want to do with my life. _____98. Looking down from a high place gives me "the jitters." _____99. I feel that few people in my life have taken advantage of me. _____100. I can't imagine being sexually invol ved with more than one person at the same time. _____101. I'm never concerned about whether I'm following the "rules" in social situations; I just ma ke my own rules. _____102. I find it easy to go up to someone I've never met and introduce myself. _____103. I often feel very nostalgic when I think back to peaceful moments in my childhood. _____104. When I go to a restaurant, I care fully look over the menu before deciding what to order. _____105. Some people seem to have gone out of their way to make life difficult for me. _____106. I have always been completely fair to others. _____107. I get a kick out of star tling or scaring other people.
146 _____108. I generally try to pay attention wh en someone important speaks to me directly. _____109. I feel very bad about myself after telling a lie. _____110. I enjoy watching violent scenes in movies. _____111. I would not enjoy being a race-car driver. _____112. I am very careful about my ma nners when other people are around. _____113. I feel that very few pe ople have ever understood me. _____114. I'm hardly ever th e "life of the party." _____115. I have occasionally felt discouraged about something. _____116. I agree with the motto, "If you are bored with life, risk it." _____117. I am a squeamish person. _____118. I enjoy (or I would enjoy) participa ting in sports involving a lot of physical contact (e.g., football, wrestling). _____119. I do not enjoy loud, wild parties and get-togethers. _____120. I often push myself to my limits in my work. _____121. I am easily "rattled" at critical moments. _____122. In school or at work, I sometimes try to "stretch" the rules a little bit just to see how much I can get away with. _____123. On occasion, I've had to restrain myself from punching someone. _____124. I wouldnÂ’t mind belonging to a group of people who "drift" from city to city, with no permanent home. _____125. I have at times been angry with someone.
147 _____126. If I were growing up during the 1960's, I probably would have been a "hippie" (Or, I was a "hippie" during the 1960's). _____127. When a friend says hello to me, I generally either wave or say something back. _____128. While watching a sporting event on TV, I sometimes wince when I see an athlete get badly injured. _____129. I'm good at flattering important people when it's useful to do so. _____130. I sometimes become deeply angry wh en I hear about some the injustices going on in the world. _____131. I'm not very good at talking people into doing favors for me. _____132. Seeing a poor or homeless person walk ing the streets at night would really break my heart. _____133. When someone tells me what to do, I often feel like doing exactly the opposite just to spite them. _____134. I always tell the entire truth. _____135. I prefer rude, but exciting people to nice, but boring people. _____136. I can remain calm in situations that would make many other people panic. _____137. I usually enjoy seeing someone I don't like get into trouble. _____138. When I'm in a group of people who do something wrong, somehow it seems that I'm usually the one w ho ends up getting blamed. _____139. People are almost always impressed with me after they first meet me. _____140. I like to (or would like to) w ear expensive, "showy" clothing.
148 _____141. In the past, people who were supposed to be my "friends" ended up getting me in trouble. _____142. I might enjoy flying across th e Atlantic in a hot-air balloon. _____143. I don't take advantage of other people even when it's clearly to my benefit. _____144. I'm the kind of person who gets "stressed out" pretty easily. _____145. Sometimes I'm a bit lazy. _____146. I sometimes like to "thumb my nose" at established traditions. _____147. During the day, I generally see the wo rld in color rather than in black-andwhite. _____148. When I am doing something important (e.g., taking a test, doing my taxes) I usually check it over at l east once or twice to make sure it is correct. _____149. When I'm among a group of people, I rarely end up being the leader. _____150. To be perfectly honest, I usually try not to help people unless I think there's some way that they can help me later. _____151. Many people probably think of my political beliefs as "radical". _____152. I sometimes lie just to see if I can get someone to believe me. _____153. I have to admit that I' m a bit of a materialist. _____154. I think that it might almost be exciting to be a passenger on a plane that appeared certain to crash, yet somehow managed to land safely. _____155. In social situations, I sometimes act the same way everyone else does because I don't want to appear too different. _____156. Never in my whole life have I taken advantage of anyone.
149 _____157. I can hold up my end of a conversatio n even if the topic is something I know almost nothing about. _____158. I often tell people only the part of the truth they want to hear. _____159. When I'm with a group of people w ho are having a serious conversation, I occasionally like to say something wild or outrageous just to be noticed. _____160. I tend to get crabby a nd irritable when I have too many things to do. _____161. I'm sure that some people would be pleased to see me fail in life. _____162. I frequently find that the way that others react to my behavior is very different from what I had expected. _____163. Some people probably think of me as a "hopeless romantic." _____164. When a task gets too difficult, I don't mind dropping it and moving on to something else. _____165. I often get blamed for thi ngs that aren't my fault. _____166. I often lose my patience with peopl e to whom I have to keep explaining things. _____167. Some people have made up stories about me to get me in trouble. _____168. I occasionally have periods of several days or more dating which I am uncertain whether I am awake or asleep. _____169. I sometimes get myself into a state to tension and turmoil as I think of the day's events. _____170. To be honest, how much I like some one depends a lot on how useful that person is to me.
150 _____171. I have sometimes felt slightly hesita nt about helping someone who asked me to. _____172. I occasionally do something dangerous because someone has dared me to do it. _____173. I sometimes try to get others to"bend th e rules" for me if I can't change them any other way. _____174. I am a "freewheeling", spontaneous person. _____175. I sometimes become so involved in my daydreams or fantasies that I momentarily forget a bout everything else. _____176. Some people have told me that I make too many excuses for myself. _____177. I am an ambitious person. _____178. Fitting in and having things in comm on with other people my age has always been important to me. _____179. I quickly become very annoyed at pe ople who do not give me what I want. _____180. I have never felt that I wa s better than someone else. _____181. If I were a fire-fighter, I think that I might actually enjoy the excitement of trying to rescue someone from th e top floor of a burning building. _____182. I will sometimes break a promise if it turns out to be inconvenient to keep. _____183. People who know me well regard me as reliable, dependable, and trustworthy. _____184. I watch my finances closely. _____185. I think that I would make a very good actor.
151 _____186. I often put off doing fun things so that I can finish my work. _____187. I think that holding the some job for most of my life would be dull.
152 ABOUT THE AUTHOR Diana Falkenbach is currently an assist ant professor on facu lty in the psychology department of John Jay College of Criminal Ju stice. She received her BachelorÂ’s degree in psychology from Emory University, and a MasterÂ’s degree in Counseling from Georgia State University. Dr. Falkenbach co mpleted her Doctoral studies in clinical psychology at the University of South Florid a. She completed her internship at NYUBellevue Hospital Center and Kirby Psychiatri c Forensic Hospital, where she worked in the field of forensic psychology, conducting evaluations of dangerousness, competency and insanity with mentally ill prisoners. Her current area of research involves considering psychopathy and aggression w ith a specific focus on the subtypes of psychopathy. Dr. Falkenbach is particularly interested in the assessment of psychopathy via self-report measures, which allows her to study psychopathic traits in nonclinical samples and juveniles.