clinical tool, used to assess psychopathic personality disorder and indicate an offender’s risk for recidivism. Critically discuss the use of this instrument for diagnosis of psychopathy and assessment of risk.
Psychopathy is a clinical construct defined by a pattern of interpersonal, affective, and behavioural characteristics; some of these include superficial charm; shallow affect; lack of guilt, remorse, or empathy; manipulation; deception; impulsivity; irresponsibility; poor behavioural controls; and a range of antisocial and unethical behaviours, however, these are not necessarily criminal (Hare, 1993; Hare, 2003). Research has found that these features can become evident in early childhood and tend to be relatively stable during adolescence and into adulthood (Lynam, Caspi, Moffitt, Loeber, and Stouthamer-Loeber, 2007; Lynam, Loeber and Stouthamer-Loeber, 2008).
Hare (-) suggests that approximately one percent of the general population are psychopaths, but this is much higher in the prison population (approximately twenty percent) – check this, done from memory of lectures – also find citation.
Although a small percentage of psychopaths are present in the general population, Hare (1998) states that psychopaths commit their first crime earlier and commit offences at a higher rate than other criminals. While in prison or an institute, they tend to be uncooperative and disruptive, therefore they are hard to manage and most psychopaths do not respond to standard treatment programmes. Therefore, it is essential that psychopathy is measured accurately in order to gain insight into treatment methods and risk assessment with psychopaths.
This essay is going to focus on the PCL-R as this is the tool that is most frequently used to assess psychopathy and research has established high reliability and validity in prison and forensic psychiatric populations (Hare, 1991 cited in Gacono and Hutton, 1994). First, I will discuss the use of the PCL-R for diagnosing psychopathy; I will then critically discuss whether the PCL-R is useful for assessing the risk of psychopaths.
Hare (1980) created a research tool for measuring psychopathy; this was originally called the PCL, which was then later revised and published as the PCL-R (Hare 1991; see also Hart, Hare and Harpur, 1992). Hare (1998) has argued that to understand the nature of psychopathy, we must first be able to identify and measure the construct. If we are to make use of correctional and mental health resources, we have to be able to make educated choices about which treatments are suitable. Additionally, in order to protect society, we have to assess the risk of psychopathy for violence and recidivism reliably. Hare (1998) believes that the PCL-R achieves these purposes.
Psychopathy is measured using detailed file information alongside a semi-structured interview in which the PCL-R is used. The PCL-R is a clinical construct which has 20 items; each item has a 3-point scale whereby a score of 0, 1, or 2 is given depending on the criteria matched. An estimate of the degree to which someone corresponds with the prototypical psychopath is then given by the final score (ranging from 0 to 40). Hare, Harpur, Hakstian, Forth, Hart and Newman (1990) found that the PCL-R measures a unitary construct (high internal consistency), but also has a two-factor structure. The tool measures affective/interpersonal features (factor 1) and social deviance (factor 2); factor 2 is also significantly correlated with diagnoses of antisocial personality disorder (APD). However, it is important to note that the two labels cannot be used interchangeably as people can be diagnosed with either or both. Hare (1996a – from without conscience) states that if this is not recognised, conclusions can be misleading.
Now that a general background on psychopathy has been given and the basic way in which the PCL-R works has been explained, I will discuss studies that have focussed on whether the PCL-R is an accurate measure of psychopathy.
Research has confirmed that the PCL-R measures psychopathy; the tool has a stable factor structure (Hare, et al. 1990; Harpur, Hakstian and Hare, 1988) and has high interrater and test-retest reliability (Hare, et al. 1990; Schroeder, Schroeder and Hare, 1983 AND Cacciola, Rutherford and Alterman, 1990; Alterman, Cacciola and Rutherford, 1993). Therefore, on first site, it seems as though the PCL-R can accurately measure psychopathy. However, due to researchers focussing on North American Caucasian males in prison and forensic populations, the psychometric properties have only been established with this section of the population (Cooke and Mitchie, 1997; Harpur, Hare, & Hakstian, 1989). Therefore, I am going to find out whether the PCL-R measures psychopathy accurately in other groups; I will specifically look at cross-cultural and gender differences.
Researchers in the United States have included black inmates in several studies and have therefore applied the psychopathy construct to White and non-White populations without ever specifically looking at the generality of psychopathy across race. Kosson, Smith and Newman (1990) were the first people to address whether the psychopathy construct is valid in Black inmates, and whether the PCL/PCL-R was a valid measure of psychopathy in Blacks; three studies were conducted. The studies confirm that psychopathy is present in Black male inmates, however, the way in which psychopathy is expressed differs between race; there is a difference in distribution of scores and in underlying factor structure of the PCL/PCL-R. Cooke, Kosson and Michie (2001) compared PCL-R ratings of Caucasian and African American participants to determine whether the structure of psychopathy is the same for the two groups. They found no difference in factor structure across the two groups, but five of the twenty PCL-R items present themselves differently; the differences were significant but because they were small, they cancelled each other out. Therefore, this study shows that although there are differences between the two groups, the PCL-R can be used accurately to assess psychopathy in the African American population.
Recently, there has been investigation into whether the PCL-R can be applied to female samples. Salekin, Rogers and Sewell (1997) examined whether psychopathy can be applied to female offenders in the same way as it is with male offenders. They used three instruments to assess this, one of which was the PCL-R. It was found that the PCL-R is a reliable measure of psychopathy in female offenders, but the rates and severity of symptoms were lower. The PCL-R also revealed that the factor structure of psychopathy in women is substantially different to that of men. It was therefore concluded that the PCL-R is a good measure of psychopathy in women but some refinements could be made. There are also other studies which have found that the PCL-R is a reliable and valid instrument with female samples (e.g. Strachan, 1995). However, the conclusions made in all of these papers need to be acknowledged with care as they all raise crucial methodological issues. Focussing on the Salekin, Rogers and Sewell (1997) study described above, their sample size was very small (103 women) and their sample was not divided by race, therefore questioning whether the factor analysis can be taken as accurate.
Vitale, Smith, Brinkley and Newman (2002) also studied whether the PCL-R is a reliable measure of psychopathy in a female sample but acknowledged the methodological problems described above by examining 528 non-psychotic female offenders and also dividing race when analysing the results. They found that the PCL-R is a reliable measure of psychopathy in female offenders; the item-to-total correlations and internal consistency were similar to male offenders. However, the number of participants that scored over 30 was scarce; the reason for this could be that women have a lower base rate of psychopathy than men or that women express the construct differently. Therefore, to account for this, the cut-off score could be reduced or the PCL-R items could be modified to capture the construct successfully. Before this is possible, future research needs to examine item function and the cause of psychopathy.
Mention that they also looked at difference between Caucasian and African American females and found that there were few important differences between the races?
Cooke and Michie (1997) used an item response theory (IRT) approach to analyse the data of 2,067 participants. They found that the PCL-R suitably diagnosed psychopathic personality disorder and was accurate at measuring trait strength. It was also found that the PCL-R is remarkably consistent when used in different cultural groups and also different settings; the analysis found no evidence to suggest that the tool is biased because of a mental disorder or race. However, it is suggested that when diagnosing someone as a psychopath, the cut-off scores could be altered among subgroups and/or cultures.
DISCUSS THIS – ADD INFO!
MENTION DIFFERENT SCORES IN USA AND UK
Even though it can be seen from the above studies that the PCL-R is generally a reliable and valid measure of psychopathy among cultures and gender, the score should always be considered carefully. The PCL-R can be used to help make decisions in any context in which psychopathy is relevant, and it is not always the case that the instruments are properly administered and interpreted, or that the decision makers use the information judiciously. It can therefore be seen that if a PCL-R rating were not accurate, detrimental consequences could occur for a patient or offender, the criminal justice and mental health systems, and also society.
As a result of the PCL-R having significant predictive validity, it now plays a crucial part in creating new risk instruments. It can be used on its own or as a component of a risk instrument which in turn affects the evaluations of management and treatment, and decisions about sentencing and eligibility of parole.
Although the PCL-R was originally developed as a tool to identify psychopathy, research has generally found that the PCL-R is also useful as a risk assessment tool (CITATION). I will now discuss this issue further and evaluate the extent to which the PCL-R can be used as a risk assessment tool.
Harris, Rice and Quinsey (1993) found that the PCL-R is better at assessing non-violent and violent recidivism than other classic actuarial risk scales. However, most research has focussed on North American samples, therefore the results have only found that the PCL-R is useful for predicting violence and identifying recidivists in North American prison and forensic populations (Hart, 1998a). I will now focus on whether there is any evidence that the predictive validity of the PCL-R is the same with other samples (particularly European samples). Grann, Langstrom, Tengstrom and Kullgren (1999) assessed the predictive power of the PCL-R with Swedish offenders (with personality disorder) and found that the PCL-R was the best tool at predicting violent recidivism 2 years after release. From this study, it can therefore be seen that the PCL-R is as valid at predicting violent recidivism in forensic settings in Sweden as it is in North America.
RISK – page 105 of Hare (1998) The Hare PCL-R: Some Issues Concerning its use and misuse. ALSO look at labelling section in the same paper.
Harris (1993) violent recidivism of mentally disordered offenders
Gacono and Hutton (1994).
Hare (1991). Manual for the revised psychopathy checklist, Multi-Health Systems, Inc, Toronto, Ontario
Hare (1993). Without Conscience: The Disturbing World of the Psychopaths among Us
Losel, F. (1998). Treatment and Management of Psychopaths. In D. J. Cooke, A. E. Forth & R. D. Hare, Psychopathy: Theory, Research, and Implications for Society. Dordrecht, The Netherlands: Kluwer
Lynam, D., Caspi, A., Moffitt, D., Loeber, R., & Stouthamer-Loeber, M. (2007). Longitudinal Evidence That Psychopathy Scores in Early Adolescence Predict Adult Psychopathy. Journal of Abnormal Psychology, Vol. 116, 155-165
Lynam, D. R., Loeber R., & Stouthamer-Loeber, M. (2008). The stability of psychopathy from adolescence into adulthood: The search for moderators. Criminal Justice and Behaviour, Vol. 35, 228-243
Strachan, C. E. (1993). The assessment of psychopathy in female offenders. Unpublished doctoral dissertation, University of British Columbia, Vancouver, British Columbia, Canada.