This article was downloaded by: [Cardiff University] On: 11 April 2011 Access details: Access Details: [subscription

number 930967547] Publisher Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK

Journal of Forensic Psychiatry & Psychology

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t714592861

A comparison of a self-report measure of psychopathy with the psychopathy checklist-revised in a UK sample of offenders

Sonja Copestakea; Nicola S. Grayb; Robert J. Snowdenc a Department of Psychology, HMP Leyhill, Wotton-under-Edge, Gloucestershire, UK b Department of Psychology, Pastoral Cymru, Ty Catrin, Cardiff, UK c Department of Psychology, Cardiff University, Park Place, Cardiff, UK Online publication date: 30 March 2011

To cite this Article Copestake, Sonja , Gray, Nicola S. and Snowden, Robert J.(2011) 'A comparison of a self-report measure

of psychopathy with the psychopathy checklist-revised in a UK sample of offenders', Journal of Forensic Psychiatry & Psychology, 22: 2, 169 182 To link to this Article: DOI: 10.1080/14789949.2010.545134 URL: http://dx.doi.org/10.1080/14789949.2010.545134

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

The Journal of Forensic Psychiatry & Psychology Vol. 22, No. 2, April 2011, 169182

A comparison of a self-report measure of psychopathy with the psychopathy checklist-revised in a UK sample of oenders
Sonja Copestakea, Nicola S. Grayb and Robert J. Snowdenc*
Department of Psychology, HMP Leyhill, Wotton-under-Edge, Gloucestershire, GL12 8BT, UK; bDepartment of Psychology, Pastoral Cymru, Ty Catrin, Dyfrig Road, Cardi, CF5 5AD, UK; cDepartment of Psychology, Cardi University, Park Place, Cardi, CF10 3AT, UK
Downloaded By: [Cardiff University] At: 08:34 11 April 2011
a

(Received 12 November 2010; nal version received 1 December 2010) Psychopathy is an extremely important construct in both forensic psychology and psychiatry. Recently, there have been attempts to produce self-report questionnaires of psychopathic traits. We examined the relationship between such a self-report measure of psychopathy [(Psychopathic Personality Inventory Revised (PPI-R), Lilienfeld, S.O., & Widows, M.R. (2005). Psychopathic Personality Inventory Revised. Lutz, FL: Psychological Assessment Resources, Inc.] and an assessment of psychopathy by a third party using the Psychopathy Checklist Revised (PCL-R; R.D. Hare, 2003. Hare Psychopathy Checklist-Revised (PCL-R) 2nd Edition. Toronto: ON: Multi-Health Systems.). A strong relationship (r .54) was found between the two measures of psychopathy in a sample of male oenders. However, the factors underpinning the PPI-R and the factors underpinning the PCL-R did not show any obvious correspondence. It is suggested that the PCL-R and the PPI-R measure dierent conceptualisations of psychopathy and that one cannot be used as a substitute for the other. Keywords: psychopathy; oending behaviour; PCL-R

Introduction Psychopathy is an important clinical construct in the eld of forensic psychiatry and psychology (Hare & Neuman, 2008; Monahan, 2006). The most commonly used measure of the construct is the Psychopathy ChecklistRevised (PCL-R; Hare, 2003). The PCL-R has strong psychometric properties (Hare, 2003), and high scores on the PCL-R are associated with a range of outcomes such as greater future criminality (e.g. Hart, Kropp, & Hare, 1988), greater institutional violence (e.g. Gray, Hill, McGleish, Timmons, MacCulloch & Snowden, 2003), and lesser treatment engagement and response (e.g. Hobson, Shine, & Roberts, 2000).
*Corresponding author. Email: snowden@cardi.ac.uk
ISSN 1478-9949 print/ISSN 1478-9957 online 2011 Taylor & Francis DOI: 10.1080/14789949.2010.545134 http://www.informaworld.com

170

S. Copestake et al.

The PCL-R was developed for use in forensic samples. As such it has been criticised for its heavy reliance on criminal behaviours as the indicators of the concept of psychopathy (e.g. Skeem & Cooke, 2010; but see also Hare & Neumann, 2010). The PCL-R relies heavily on good collateral information for its completion and should not be rated only on an interview with the person (PCL-R Manual, p. 19; Hare, 2003). Therefore, it takes a long time to complete (many hours depending on the amount of collateral information) and can only be performed if this information is available. Extensive training is required prior to administration of the PCL-R. Hence, for many purposes, such as exploration of the concept of psychopathy in non-criminal populations, the PCL-R cannot be coded due to a lack of collateral information or the cost (in both time to complete an individual assessment and training time for the researchers) makes this prohibitive. There have been several attempts to produce a self-report questionnaire of psychopathy (for a review and a critique of the individual measures and of self-reporting of psychopathy in general, see Lilienfeld & Fowler, 2006). Perhaps the most established of these is the Psychopathy Personality Inventory (Lilienfeld & Andrews, 1996). The PPI has undergone a revision to the PPI-R (Lilienfeld & Widows, 2005). This revision sought to maintain the strengths of the PPI whilst enhancing its psychometric properties and clinical and research utility (see Lilienfeld & Widows, 2005). This was done by lowering the reading level required, rewording culturally specic items and reducing its length to 154 items (from 187 items). A large volume of data has been accumulated on the PCL-R and there is strong evidence for two subfactors underpinning the global concept (Hare, 1991, 2003). Factor 1 is the Interpersonal and Aective factor and reects callous and unemotional traits. Factor 2 is the Lifestyle and Antisocial factor and reects the impulsive, reckless, and criminal behaviours of some individuals. More recently, these factors have been subdivided further into a four-facet model (Hare, 2003) comprising Facet 1 (Interpersonal), Facet 2 (Aective), Facet 3 (Lifestyle), and Facet 4 (Antisocial) see also Cooke and Michie (2001) and Neumann, Hare, and Newman (2007). Like the PCL-R, the PPI has been found to have at least two underlying factors (Benning, Patrick, Hicks, Blonigen, & Krueger, 2003): Fearless Dominance (often termed PPI-I), and Impulsive Antisociality (often termed PPI-II). This latter factor has been re-labelled as Self-Centered Impulsivity by Lilienfeld and Widows (2005). However, Neumann, Malterer, and Newman (2008) failed to replicate this two-factor structure in an incarcerated sample. Lilienfeld and Widows (2005) describe the PPI-R factors as follows:
In general, high scores on Fearless Dominance reect a tendency towards lack of anticipatory social and physical anxiety, low levels of tension and worry,

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

The Journal of Forensic Psychiatry & Psychology

171

low harm avoidance and high levels of interpersonal dominance. In general, high scores on Impulsive Antisociality reect a tendency toward selfcenteredness, ruthless use of others, brazen outing of traditional values, propensity to attribute blame to others for ones mistakes and reckless impulsivity (p.22).

These brief descriptions of the two factors of the PPI-R appear to map reasonably well to the two factors identied within the PCL-R. The PPI-R also contains a subscale called the Coldheartedness scale that captures the absence of strong aective responses believed to be at the heart of psychopathy. However, Benning, Patrick, Hicks, Blonigen & Krueger (2003) found that this scale did not load on PPI-I or PPI-II, and it was therefore left as a stand alone scale. Given that these two measures, the PCL-R and the PPI-R, both purport to measure the construct of psychopathy, it should be expected that they would be highly related to one another. Likewise, it would seem probable that the two major factors identied in the PPI would correspond closely to those of the PCL-R factors. Indeed, such an assumption has been made in recent research reports (Justus & Finn, 2007; Rilling, Glenn, Jairam, Pagnoni, Goldmsmith, Elfenbein et al., 2007). Surprisingly, there have been few studies of this important issue. Malterer, Lilienfeld, Neumann, and Newman (2010) recently explored this issue by correlating scores from the PPI with the PCL-R and the Psychopathy Checklist Screening Version (PCL:SV; Hart, Cox, & Hare, 1995), the latter of which is a shortened derivative of the PCL-R used for screening for psychopathy in both forensic and non-forensic settings.1 For the oender samples, they found modest correlations between the total scores on the two scales (.39.42), whilst a more in-depth analysis of the subfactors of each scale did not reveal any simple relationship between them. The authors concluded that the PPI and PCL, although overlapping in important aspects, exhibit notable dierences (page 10). Given the potential benets of a self-report measure of psychopathy, we thought it important to see if the results of Malterer et al. (2010) could be replicated in a dierent sample. However, whilst the major aim of the study was merely one of replication, we note some features that make our study unique: . Our sample is drawn from oenders in the UK and is the rst to examine these issues outside of North America. . The existing research has used the PPI and not the most up-to-date version, the PPI-R, which we have used. . Recent advances in the modelling of the factors underpinning psychopathy as dened by the PCL-R has suggested both three- and four-facet models (Cooke & Michie, 2001; Neumann et al., 2007). Hence, this study is the rst to examine how these facets are related to the factors of the PPI-R.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

172

S. Copestake et al.

Methods Participants Participants were 52 convicted, male oenders located at three dierent prisons. Twenty were located at HMP Long Lartin which is a high security prison, 25 of the participants were located at HMP Leyhill which is an open prison, and 7 of the participants were located at the Westgate Unit at HMP Frankland which is a specialist unit for oenders who meet the criteria for Dangerous and Severe Personality Disorder (DSPD). Of the total sample, 85% were serving life sentences and 15% were serving determinate sentences. The mean age of the sample was 38 years old (SD 9.7, range 2266 years). Most of the sample (77%) were of Caucasian origin, 18% were of Black Caribbean or Black African origin, and 5% were of Asian origin. Ethical approval for the study was provided by chartered forensic psychologists from both the Directorate of High Security Prisons and HMP Leyhill.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

Measures Psychopathy checklist Revised The PCL-R is a 20-item tool that has been designed to measure psychopathic traits via the assessment of le information, collateral reports and interviews with the oender. All participants PCL-R assessments in this study were completed either as part of a risk assessment, as most of them are life-sentenced prisoners, or to make a decision about whether an identied accredited oending behaviour programme would meet the oenders needs. PCL-R assessments were completed in line with HM Prison Service guidelines for the administration of the PCL-R. All PCL-R assessments were either completed and or supervised by psychologists who have successfully completed HM Prison Service PCL-R training and achieved inter-rater reliability through either the Darkstone or HM Prison Service certication process. As is recommended in the PCL-R manual (Hare, 2003), the vast majority of the assessments (54/57) were second scored and the consensus score of the two raters was used.2 The PCL-R scores for the remaining three participants were obtained from a database, and it was unclear whether these had been subjected to a second rater or not. The internal consistency of the total PCL-R score was acceptable (Cronbachs a .91) and similar to those published in the PCL-R manual, second edition (a .81.89). We also achieved acceptable internal consistency at the factor level (Factor 1: a .85; Factor 2: a .86) and at the facet level (Facet 1: a .84; Facet 2: a .73; Facet 3: a .74; Facet 4 a .76).

The Journal of Forensic Psychiatry & Psychology

173

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

Psychopathic personality inventory Revised The PPI-R is a 154-item self-report measure of psychopathy (Lilienfeld & Widows, 2005). Each question is answered on a four-point scale of 1 (False), 2 (Mostly False), 3 (Mostly True) and 4 (True). Participants were presented with written instructions about the PPI-R and how it should be completed. The PPI-R produces a global psychopathy score and scores for the PPI-I (Fearless Dominance), PPI-II (Self-Centered Impulsivity), and the Coldheartedness factors. The PPI-R also contains three validity scales that are designed to measure response style: Virtuous Responding measures impression management, Deviant Responding is designed to measure responses that reect faking bad, or deliberate attempts to sabotage the results, and Inconsistent Responding measures random or inconsistent responses, or reading diculty. Three participants had scores over the accepted cut-o (45) on the Inconsistent Responding scale. Analyses with or without these participants produced highly similar results. In the present article, all the analyses that involve results from the PPI-R excluded these participants. Previous work (Lilienfeld & Widows, 2005) has shown that the scales of the PPI-R have good internal consistency (a .78.92) and testretest reliability (r .82.95). Analyses of the present data showed high internal consistency for the global psychopathy score (a .93) and for each of the subscales (Fearless Dominance: a .87), Self-Centered Impulsivity: a .95, Coldheartedness: a .85). Results The descriptive statistics of the sample are shown in Table 1, and the correlations between the PCL-R measures and the PPI-R measures are shown in Table 2.

Table 1. Measure

Descriptive statistics for the PCL-R and the PPI-R. Mean 19.88 7.45 10.18 2.98 4.12 4.70 6.20 268.08 109.98 12.17 32.92 Standard deviation 8.69 4.31 4.93 2.44 2.14 2.61 2.93 44.29 18.83 33.00 9.36 Range 037 016 018 08 08 09 010 199410 65168 78218 960 Skew 7.14 .32 7.33 .73 7.02 7.23 7.75 .61 .39 .44 .45 Kurtosis 7.69 7.68 7.74 7.62 7.81 71.05 7.44 .40 1.08 7.25 .66

PCL-R total Factor 1 Factor 2 Facet 1 Facet 2 Facet 3 Facet 4 PPI-R total PPI-1 PPI-11 Coldheartedness

174
Table 2. factors. Measure

S. Copestake et al.
Correlations between PCL-R and PPI-R total scores and underlying Factor 1 .50** .19 .48** .43** Factor 2 .49** .21 .44** .46** Facet 1 .40** .05 .40** .46** Facet 2 .42** .04 .48** .48** Facet 3 .51** .25 .46** .42** Facet 4 .50** .35* .39** .31*

PCL-R .54** .19 .51** .42**

PPI total PPI-I PPI-II Coldheartedness

*p 5 .05, **p 5 .01.

PPI-R total score As expected, the total score of the PPI-R showed a signicant correlation with the PCL-R total score (r .54). This is regarded as a large eect size (Cohen, 1988) and accounted for 29% of the variance between the variables. PPI-R total score was signicantly associated with both Factor 1 and Factor 2 of the PCL-R (large eect size), and with all four facets of the PCL-R. PPI-I (fearless dominance) PPI-I was not signicantly correlated with the total PCL-R score. Crucially, for the aims of this study, PPI-I had only a weak and non-signicant correlation with Factor 1. At the facet level, the only facet to be related to PPI-I was Facet 4 (Antisocial), with a moderate eect size. PPI-II (self-centered impulsivity) PPI-II showed a strong correlation with the total PCL-R score. As expected, PPI-II was signicantly correlated with Factor 2 of the PCL-R, with a moderate to strong eect. However, the correlation of PPI-II with Factor 1 of the PCL-R was also signicant. Hence, it cannot be said that PPI-II is uniquely related to Factor 2. Due to the shared variance of Factors 1 and 2 of the PCL-R,3 the examination of zero-order correlations may be misleading. Therefore, we also performed partial correlations examining Factor 1 after removing the variance associated with Factor 2. The correlation between PPI-II and Factor 1 remained signicant (r .40, p 5 .01). Examination at the facet level showed that PPI-II was signicantly related to all four facets of the PCL-R, with approximately the same eect size (moderate) for all facets. Coldheartedness Coldheartedness was signicantly correlated with the PCL-R total score, factor scores and facet scores. All eects sizes were moderate.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

The Journal of Forensic Psychiatry & Psychology

175

Discussion The major aim of this article was to examine the relationship between the PCL-R and the PPI-R measures of psychopathic traits. Specically, we hoped to provide a replication study to that of Malterer et al. (2010). Total PPI score Malterer et al. (2010) found moderate correlations between the PCL and the PPI in their two oender populations (r .39 and .42).4 Our gure is a little higher than this (r .54) but is in good agreement with previous studies in North American populations (Edens, Poythress, Lilienfeld, & Patrick, 2008b; Poythress, Edens, & Lilienfeld, 1998). Hence, at a global level, there is a strong correlation between the two measures, suggesting that they are tapping similar constructs. Given recent evidence that psychopathy, even when measured using PCL scales, is a dimensional construct (Edens, Marcus, Lilieneld, & Poythress, 2006; Walters, Gray, Jackson, Sewell, Rogers, Taylo et al., 2007), there has been a move to use non-incarcerated people in the study of psychopathy (see, for example, Vanman, Mejia, Dawson, Schell, and Raine, 2003). Such people are far more readily available for research purposes and may have fewer problems, such as drug and alcohol use, that may confound studies of incarcerated oenders. The present results, and those outlined in this paragraph, would suggest that the use of the PPI-R as a global measure of psychopathy would be appropriate in such a community sample. The other major use of the PCL-R is in clinical/forensic decision making. Here, we would advise against using only self-report measures of psychopathy, such as the PPI-R. Other researchers have urged caution in accepting self-report in prison populations (e.g. Bagby, Rogers, & Buis, 1994; Hare, Forth, & Hart, 1989). This problem may be more acute for those with high psychopathy scores (Lilienfeld & Fowler, 2006; Spidel, , Greaves, & Yuille, in press). It is apparent that such individuals are Herve skilled at manipulating their projected image as evidenced by their ability to present themselves to decision makers in a manner that aids their conditional release (Porter, ten Brinke, & Wilson, 2009), gains more lenient sentences (Ha kka nen-Nyholm & Hare, 2009) and gains permission to appeal against these more lenient sentences. There are also reported examples of individuals who are able to manipulate their proles on self-report measures and even help others do the same (Hare, 1985). PPI-I score We were able to replicate Malterer et al.s (2010) nding that PPI-I and PCL Factor 1 show only weak correlations. Indeed, we also found that PPI-I was

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

176

S. Copestake et al.

not signicantly correlated to the total or factor scores of the PCL-R. However, we did nd a moderate association with PPI-I to the antisocial aspects of the PCL (Facet 4).5 This lack of association between PPI-I and the well-established PCL measures of psychopathy may seem puzzling, and warrants further investigation, given the ndings that PPI-I appears to have good construct validity (Cima & Raine, 2009; Edens, Poythress, Lilienfeld, Patrick, & Test, 2008a; Edens et al., 2008b). We shall return to the possible reasons for this lack of PCL and PPI-I associations later. PPI-II score We replicated Malterer et al.s (2010) nding that PPI-II and PCL Factor 2 are correlated. However, we also found that PPI-II was also correlated with PCL Factor 1 at about the same level of association. Signicant associations between Factor 1 and PPI-II are also notable in the data of Malterer et al. (2010). Thus, it does not appear to be the case that there is any unique association between PPI-II and Factor 2 of the PCL. At the facet level, the PPI-II was associated with all four facets with moderate eect sizes. Thus, whilst the PPI-II clearly appears to be measuring something that is related to the concept of psychopathy as dened by the PCL, it does not seem to tap any sub-dimension of the PCL better than any other sub-dimension. Coldheartedness Malterer et al. (2010) found only weak (but signicant) associations with the Coldheartedness scale and all PCL scales. The present study also found signicant correlations, but these have a moderate eect size. The Coldheartedness scale aims to measure an absence of feelings of guilt, empathy and loyalty, a lack of attachment to others, and a callousness to the feelings of others. As such, it would appear similar to the concepts underpinning Facet 2 (Aective) of the PCL. Facet 2 was signicantly correlated with Coldheartedness, however, so were all three other facets. Why are the factors of the PPI-R and the PCL-R so poorly associated? The PCL-R and the PPI-R were developed in dierent populations, with the PCL-R being developed from work with oenders, whilst the PPI-R was developed in community samples. These two groups might therefore interpret the questions of the PPI-R in a quite dierent manner. For instance, an item such as A lot of times, I repeat the same bad decisions, might have quite dierent connotations for the two groups. For many judgements about ourselves we must norm ourselves to others. In may well be that oenders tend to judge themselves against other oenders, whilst college students might judge themselves against other college students.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

The Journal of Forensic Psychiatry & Psychology

177

Hence, even though an individual may think he has repeated the same bad decisions several times, he might think he has done so less than many of his fellow inmates, and hence rated himself quite low on this item. This may explain some puzzling ndings in the literature. For instance, Malterer et al. (2010) show quite large dierences between their non-clinical sample (undergraduates Study 3) compared to their oender sample (Study 2) on the PCL:SV (means of 6.9 and 13.9, respectively). Given a standard deviation of around 5.0, this dierence could be described as an eect size (Cohens d) of around 1.4 (very large). On the other hand, these two samples had scores of 380 and 387 on the PPI (with SDs of around 40), thus producing a very small eect size (d .15). We also note that the present sample (which have high PCL-R scores overall) has a mean PPI-R score of 269, which is actually lower than the scores for the community sample of males age 3039 (mean 290; Lilienfeld & Widows, 2005). Hence, it seems unlikely that these two populations are actually interpreting these questions in a similar manner. If this speculation is correct, this might also go someway to explain why the two-factor solution to the PPI derived from a community sample by Benning et al. (2003) could not be replicated in an oender sample by Neumann et al. (2008). Another possibility for the weak associations between the PPI and PCL factors may arise due to the two scales having somewhat diering conceptions of psychopathic traits. Again, the fact that the PCL and its derivatives were developed and tested on mainly forensic populations, whereas the PPI was developed and tested mainly on community samples, may underpin the diering conceptions. These two samples may well dier with respect to their psychopathic characteristics. For example, one might expect that the former might contain a relatively high sample of unsuccessful psychopaths, as the vast majority of this research examines those in secure settings. On the other hand, the successful psychopath is more likely to appear in non-forensic settings. Hence, the conceptualisation of psychopathy derived from these dierent samples may well have dierences. Patrick and colleagues (Patrick 2010; Patrick, Fowles, & Krueger, 2009) have developed a model of psychopathy that has three underlying components: boldness, meanness and disinhibition. It has been suggested (Patrick, 2010) that the PCL-R tends to measure a combination of factors such as meanness (Factor 1) and externalising behaviour (Factor 2) that is consistent with some previous historical conceptions of psychopathy (e.g. McCord & McCord, 1964). In particular, Factor 1 of the PCL-R mainly taps the concept of meanness (with some elements of boldness), whilst Factor 2 mainly taps disinhibition (with some elements of meanness). As such the PCL-R does not seem cover the concept of boldness particularly well. Indeed, some researchers (e.g. Newman & Schmitt, 1998) have often subdivided groups that score highly on the PCL-R further on the basis of low vs. high anxiousness (a concept strongly related to boldness).

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

178

S. Copestake et al.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

The PPI, however, appears to emphasis the concept of boldness in the form of the Fearless Dominance scale (Patrick, 2010). Therefore, the present results showing little association between PCL-R Factor 1 and PPI-R Fearless Dominance are consistent with this notion that they are measuring dierent concepts. We note, however, that there was one signicant association between the PPI-R Fearless Dominance and the PCL scales in the form of Facet 4 (Antisocial). It seems possible that the boldness could lead to the expression of antisocial behaviour that might be inhibited by someone not so bold. A nal possibility relates to the nature of psychopathy itself (Lilienfeld & Fowler, 2006). Several authors (e.g. Cleckley, 1988) have commentated on their lack of insight into their personalities and, a related, inability to see themselves as others do. Cleckley (1988) further noted that psychopaths appear to suer from a semantic aphasia. He notes that it must be dicult for a person who has never experienced a particular emotion to report on its presence or absence. As an analogy one can imagine asking a colour-blind patient to report on the redness of an apple. Whilst they may be able to copy or mimic the response of other people when they use the word red, this is clearly dierent to a person who can perceive this colour. Hence, if the psychopath is not able to feel some particular emotion (such as empathy for another who has been injured) he may well mistake whatever feelings he is having (such as excitement) due to the labelling he hears from others. He may well then report on his levels of emotion (quite honestly) but does not appreciate that he is not reporting on the same emotion as other people. In such a condition, it may well be the interpersonal, rather than behavioural, components of psychopathy that the psychopath is not able to self-report accurately. Hence, self-report measures may struggle to capture these components of psychopathy producing a dierent factor structure than those that supplement selfreport with collateral information and third party information, such as the PCL-R. Limitations The ndings from the present study are limited in several ways. First, our sample was relatively small. However, the aim of this study was to establish the magnitude of the relationships between the PPI-R and the PCL-R in a UK oender sample, rather than to detect the presence of small correlations (where statistical sensitivity is of importance). Of course, a larger sample would give us greater condence in the eect sizes we report, and the ability to compare correlations in a statistically sensitive manner. Second, our sample consisted purely of male UK oenders, and may not generalise to other samples (e.g. female oenders). Likewise there have been, to date, no published data that have conrmed (or refuted) the factor structure of the

The Journal of Forensic Psychiatry & Psychology

179

PPI-R that has emerged from work in N. America. Our current sample size was not large enough to perform such an analysis. We also note the very low proportion of ethnic minorities in our sample. Given that PCL-R scores are often taken as part of a clinical routine in oenders and that the PPI-R is relatively easy to administer, it is hoped that these limitations can be overcome by future studies. Conclusions Our results appear to replicate those of Malterer et al. (2010) in most major respects and extend their ndings to a UK population. It appears that at the level of global or total psychopathy score, there is a reasonable correspondence between the measures. This is impressive in that (1) the two measures do not share method variance (the PPI-R is a self-report questionnaire, whilst the PCL-R is a clinical checklist based on collateral review and interview) and (2) the possible worries about dissimulation in this particular population (see Introduction). Hence, there may be instances (such as research in community samples) where the PPI-R would be the easier instrument to use. However, at a factor level, we did not nd that the factors of the PCL-R and PPI-R show equivalence. We suggest that this maybe due to the two instruments having either a dierent conceptualizations of psychopathic traits, or an inability of psychopaths to report on certain aspects of their emotions or interpersonal behaviours. With respect to the former idea, it has been suggested that the concept of boldness is not well represented in the PCLR, but is in the PPI-R. Several studies have used the factors of the PPI as synonymous to those of the PCL-R (e.g. Justus & Finn, 2007; Rilling et al., 2007). We suggest that this is inappropriate and that each instrument provides a dierent conceptual basis for the subdimensions of psychopathy. Acknowledgement
We would like to thank an anonymous reviewer for suggesting possible reasons (e.g. semantic aphasia) for our results.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

Notes
1. 2. 3. 4. For short-hand, we will refer to these instruments as the PCL under the assumption that they are measuring the same concepts at both total and factor scores (Guy & Douglas, 2006). Unfortunately, the individual scores from the two raters were not available to us to perform an analysis of inter-rater reliability. In the present sample, the correlation between Factor 1 and Factor 2 was r .49 (p 5 .001). Malterer et al. (2010) also present data from an undergraduate sample, but this is not considered here as we only had an oender sample.

180
5.

S. Copestake et al.
Data on the associations between the facet scores of the PCL-R (and PCL:SV) and the PPI were kindly supplied to us by Malterer et al. (2010). The results do not seem to form a consistent pattern, either across the studies within Malterer et al. (2010) or with the present data. Namely, Malterer et al. (2010) did not nd that Facet 4 was most associated with PPI-I. However, we note that they did not present this data in the original paper due to low levels of reliability. Clearly, our nding of an association between Facet 4 and PPI-I may be a Type 1 error, and further study is needed before this result should be accepted.

References
Bagby, R.M., Rogers, R., & Buis, T. (1994). Detecting malingering and defensive responding on the MMPI-2 in a forensic inpatient sample. Journal of Personality Assessment, 62, 191203. Benning, S.D., Patrick, C.J., Hicks, B.M., Blonigen, D.M., & Krueger, R.F. (2003). Factor structure of the Psychopathic Personality Inventory: Validity and implications for clinical assessment. Psychological Assessment, 15, 340 350. Cima, M., & Raine, A. (2009). Distinct characteristics of psychopathy relate to dierent subtypes of aggression. Personality and Individual Dierences, 47, 835 840. Cleckley, H. (1988). The mask of sanity. St. Louis, MO: Mosby (original published in 1941). Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cooke, D.J., & Michie, C. (2001). Rening the construct of psychopathy: Towards a hierarchical model. Psychological Assessment, 13, 171188. Edens, J.F., Marcus, D.K., Lilienfeld, S.O., & Poythress, N.G. (2006). Psychopathic, not psychopathic: Taxometric evidence for the dimensional structure of psychopathy. Journal of Abnormal Psychology, 115, 131144. Edens, J.F., Poythress, N.G., Lilienfeld, S.O., & Patrick, C.J. (2008b). A prospective comparison of two measures of psychopathy in the prediction of institutional misconduct. Behavioral Sciences and the Law, 26, 529541. Edens, J.F., Poythress, N.G., Lilienfeld, S.O., Patrick, C.J., & Test, A. (2008a). Further evidence of the divergent correlates of the psychopathic personality inventory factors: Prediction of institutional misconduct among male prisoners. Psychological Assessment, 20, 8691. Gray, N.S., Hill, C., McGleish, A., Timmons, D., MacCulloch, M.J., & Snowden, R.J. (2003). Prediction of violence and self-harm in mentally disordered oenders: A prospective study of the ecacy of HCR-20, PCL-R and psychiatric symptomology. Journal of Consulting and Clinical Psychology, 71, 443451. Guy, L.S., & Douglas, K.S. (2006). Examining the utility of the PCL:SV as a screening measure using competing factor models of psychopathy. Psychological Assessment, 18, 225230. Ha kka nen-Nyholm, H., & Hare, R.D. (2009). Psychopathy, homicide and the courts. Working the system. Criminal Justice and Behavior, 36, 761777. Hare, R.D. (1985). A comparison of procedures for the assessment of psychopathy. Journal of Consulting and Clinical Psychology, 53, 716. Hare, R.D. (1991). The Hare Psychopathy Checklist Revised. Toronto: MultiHealth Systems. Hare, R.D. (2003). The Hare Psychopathy Checklist Revised (PCL-R) (2nd ed.). Toronto: Multi-Health Systems.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

The Journal of Forensic Psychiatry & Psychology

181

Hare, R.D., Forth, A., & Hart, S.D. (1989). The psychopath as prototype for pathological lying and deception. In J.C. Yuille (Ed.), Credibility assessment (pp. 2549). Boston, MA: Kluwer. Hare, R.D., & Neumann, C.S. (2008). Psychopathy as a clinical and empirical construct. Annual Review of Clinical Psychology, 4, 217246. Hare, R.D., & Neumann, C.S. (2010). The role of antisociality in the psychopathy construct: Comment on Skeem and Cooke (2010). Psychological Assessment, 22, 446454. Hart, S.D., Cox, D.N., & Hare, R.D. (1995). The Hare Psychopathy Checklist Screening Version (PCL:SV). Toronto: Multi-Health Systems. Hart, S.D., Kropp, P.R., & Hare, R.D. (1988). Performance of psychopaths following conditional release from prison. Journal of Consulting and Clinical Psychology, 56, 227232. Hobson, J., Shine, J., & Roberts, R. (2000). How do psychopaths behave in a prison therapeutic community? Psychology Crime and Law, 6, 139154. Justus, A.N., & Finn, P.R. (2007). Startle modulation in non-incarcerated men and women with psychopathic traits. Personality and Individual Dierences, 43, 2057 2071. Lilienfeld, S.O., & Andrews, B.P. (1996). Development and preliminary validation of a self-report measure of psychopathic personality traits in noncriminal populations. Journal of Personality Assessment, 66, 488524. Lilienfeld, S.O., & Fowler, K.A. (2006). The self-report assessment of psychopathy. In C.J. Patrick (Ed.), Handbook of psychopathy (pp. 107132). London: The Guilford Press. Lilienfeld, S.O., & Widows, M.R. (2005). Psychopathic Personality InventoryRevised. Lutz, FL, USA: Psychological Assessment Resources, Inc. Malterer, M.B., Lilienfeld, S.O., Neumann, C.S., & Newman, J.P. (2010). Concurrent validity of the Psychopathic Personality Inventory with oender and community samples. Assessment, 17, 315. McCord, W., & McCord, J. (1964). The psychopath: An essay on the criminal mind. Princeton: Van Nostrand. Monahan, J. (2006). A jurisprudence of risk assessment: Forecasting harm among prisoners, predators, and patients. Virginia Law Review, 92, 391435. Neumann, C.S., Hare, R.D., & Newman, J.P. (2007). The super-ordinate nature of the psychopathy checklist-revised. Journal of Personality Disorders, 21, 102 117. Neumann, C.S., Malterer, M.B., & Newman, J.P. (2008). Factor structure of the Psychopathic Personality Inventory (PPI): Findings from a large incarcerated sample. Psychological Assessment, 20, 169174. Newman, J.P., & Schmitt, W.A. (1998). Passive avoidance in psychopathic oenders: A replication and extension. Journal of Abnormal Psychology, 107, 527532. Patrick, C.J. (2010). Conceptualizing the psychopathic personality: Disinhibited, bold, or just plain mean. In D.R. Lynam & R.J. Salekin (Eds.). Handbook of child and adolescent psychopathy (pp. 1548). New York: Guilford Press. Patrick, C.J., Fowles, D.C., & Krueger, R.F. (2009). Triarchic conceptualization of psychopathy: Developmental originas of disinhibition, boldness, meanness. Development and Psychopathology, 21, 913938. Porter, S., ten Brinke, L., & Wilson, K. (2009). Crime proles and conditional release performance of psychopathic and non-psychopathic sexual oenders. Legal and Criminological Psychology, 14, 109118.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

182

S. Copestake et al.

Downloaded By: [Cardiff University] At: 08:34 11 April 2011

Poythress, N.G., Edens, J.F., & Lilienfeld, S.O. (1998). Criterion-related validity of the Psychopathic Personality Inventory in a prison sample. Psychological Assessment, 10, 426430. Rilling, J.K., Glenn, A.L., Jairam, M.R., Pagnoni, G., Goldsmith, D.R., Elfenbein, H.A., et al. (2007). Neural correlates of social cooperation and noncooperation as a function of psychopathy. Biological Psychiatry, 61, 1260 1271. Skeem, J.L., & Cooke, D.J. (2010). Is criminal behaviour a contral component of psychopathy? Conceptual directions for resolving the debate. Psychological Assessment, 22, 433445. , H., Greaves, C., & Yuille, J.C. (in press). Wasnt me! A eld Spidel, A., Herve study of the relationship between deceptive motivations and psychopathic traits in young oenders. Legal and Criminological Psychology. Vanman, E.J., Mejia, V.Y., Dawson, M.E., Schell, A.M., Raine, A., & Lilienfeld, S.O. (2003). Modication of the startle reex in a community sample: Do one or two dimensions of psychopathy underlie emotional processing? Personality and Individual Dierences, 35, 20072021. Walters, G.D., Gray, N.S., Jackson, R.L., Sewell, K.W., Rogers, R., Taylor, J., Snowden, R.J., et al. (2007). A taxometric analysis of the psychopathy checklist: Screening version (PCL:SV): Further evidence of dimensionality. Psychological Assessment, 19, 330339.