Concerns regarding the potential negative effects of the psychopathy “label” on punishment outcomes may not be deserved

Colleen M. Berryessa, Rutgers University
Barclay Wohlstetter, Bucknell University (L3)

The current study, using a meta-analytic approach and moderation analysis, examines 22 studies reporting how psychopathic “labeling” influences perceptions on 3 punishment outcomes (dangerousness, treatment amenability, and legal sentence/sanction) for 2 types of experimental studies utilizing vignettes: (a) studies in which a defendant with a psychopathic “label” is compared to a defendant with no mental health diagnosis (psychopathic label vs. no label) and (b) studies in which a defendant with a psychopathic “label” is compared to a defendant with a different psychiatric diagnosis (psychopathic label vs. other psychiatric label). Results show statistically significant or marginally significant (p < .10) summary effect sizes, albeit of different strengths, for the three punishment outcomes studied (legal sentence/sanction: d = 0.17; dangerousness: d = 0.58; and treatment amenability: d= 0.30) for studies comparing a psychopathic label versus no label. Conversely, all summary effects sizes for the three punishment outcomes in studies comparing a psychopathic label versus other psychiatric label were both weak and nonsignificant (legal sentence/sanction: d _ 0.09; dangerousness: d _ 0.14; and treatment amenability: d = 0.02). This suggests a significant general labeling effect, but not a specific labeling effect, for psychopathy in these studies. Further, these results suggest that the lay public, but not those in the criminal justice system, may subscribe to both general and specific labeling effects for psychopathy when it comes to punishment. This has potential implications for jury sentencing in both capital and, in select states, noncapital cases.

psychopathy, label, meta-analysis, stigma, punishment

“The current study, using a meta-analytic approach, examines how psychopathic ‘labeling’ effects influence the perceptions of punishment in three areas: dangerousness, treatment amenability, and legal sentence/sanction. For each of these outcomes, studies involving two different scenarios were examined: (a) a defendant with a psychopathic ‘label’ compared to a defendant with no mental health diagnosis, which are studies said to represent a general labeling effect (psychopathic label vs. no label) and (b) a defendant with a psychopathic ‘label’ versus a defendant with a different psychiatric diagnosis, which are studies said to represent a specific labeling effect (psychopathic label vs. other psychiatric label).” (p. 12).

“Results show two statistically significant summary effect sizes, albeit of different strengths, for two of the punishment outcomes studied (legal sentence/sanction and dangerousness) for studies comparing an offender with a psychopathic label versus an offender with no label… Across studies, the psychopathic label did lead to significantly more support for punitive legal sanctions and reduced views regarding an offender’s potential for treatment, compared to an unlabeled offender; this not only supports literature that the psychopathic label may prompt more aggravated punishments based on stigma, but also that the psychopathic label may perpetuate views on lack of treatability. However, these effect sizes were weak; this may suggest that concerns related to how the stigma associated with psychopathy may lead to unjustly punitive outcomes for offenders, particularly related to less options for treatment, are substantiated as the label could significantly influence views, but may ultimately not result in extensive effects to perceptions of legal sanctions and treatment amenability of labeled offenders” (p. 19).

“The effect size for treatment amenability in studies comparing an offender with a psychopathic label versus an offender with no label, however, did have some interesting moderators. Particularly, the fact that an offender was a juvenile in the vignette moderated the relationship between the psychopathic label and potential for treatment amenability, while the effect was not significant for when the offender was an adult. Thus, although the effect size was moderately small, juvenile offenders with the psychopathic label were viewed as significantly less treatable, while the same views for adult offenders with a psychopathic label were not statistically different from an offender without a label. This is interesting, as there are many long-standing concerns about the negative effects of applying the psychopathy construct to youth… [L]abeling juvenile offenders with psychopathy, as well as callous and unemotional traits, may potentially prevent access to treatment opportunities that could have a tangible effect on curbing recidivism and antisocial behavior”(p. 19).

“However, sample type also moderated results related to treatability; the effect size for clinicians was both significant and moderately strong, compared to criminal justice actors, lay public and students, and students which were nonsignificant (except for lay public). Thus, clinicians more strongly viewed offenders with a psychopathic label as less amenable to treatment, compared to the other sample types. As clinicians would likely be the sample type with the most experience regarding whether or not individuals with psychopathy would be able to be successfully treated, this result suggests that the stereotype that persons with psychopathy are not amenable to rehabilitation may be warranted. It is also possible that this finding may represent clinician bias, or potentially, the fact that there still is little research on the treatment of psychopathy. Regardless, as psychopathy is an accepted clinical construct, this could potentially cause problems particularly for juvenile offenders with the psychopathic label and getting access to treatment” (p. 19).

“The effect size for dangerousness in studies comparing an offender with a psychopathic label versus an offender with no label was moderately strong. This result appears to be intuitive, as it would be unlikely that an individual with a psychopathic label would be viewed as less dangerous in comparison to someone without any label, given the stigma around psychopathy and dangerousness” (p. 20).

“Finally, the type of labeling effect moderated results for dangerousness; all but two studies either used a labeling effect (13 effect sizes) or both a labeling and criterion effect together (21 effect sizes). Interestingly, the label effect alone was significant but resulted in a small effect, whereas the use of both label and criterion effects resulted in an effect size over five times as large as the labeling effect alone. Previous literature has discussed that it is unclear whether the results of these studies reflect a labeling effect or a criterion effect. These results suggest that the combination of both label and criterion effects substantially influences outcomes related to dangerousness in studies comparing an offender with a psychopathic label versus an offender with no label; as use of the label alone produced only a small effect and much weaker than when the label and criterion effects were presented together, these data appear to show how impactful including the characteristics of psychopathy, along with the label, may be on perceptions of dangerousness” (p. 20).

“Conversely, summary effects sizes for the three punishment outcomes in studies comparing an offender with a psychopathic label versus an offender with another psychiatric label were both weak and nonsignificant. The effect sizes for legal sentence/sanction and treatment amenability were close to zero” (p. 20).

“The diagnostic label also moderated the results for the legal sentence/sanction outcome; principally, only the effect size for psychotic disorder was significant, and those for conduct disorder, antisocial personality disorders, and paraphilic disorder were all essentially zero. This suggests, at least related to the rendering of a legal sentence/sanction, that an offender with a psychopathic label is not viewed as significantly different from an offender labeled with conduct disorder, antisocial personality disorders, or paraphilic disorder. Conduct disorder, antisocial personality disorders, and paraphilic disorder all have been traditionally associated with negative stigmata, such as dangerousness, lack of treatability, and punitive judgments similar to psychopathy. Psychosis, on the other hand, has been associated with reduced perceptions of responsibility and punitiveness due to beliefs that diagnosed individuals are less culpable for their behavior because they may have ‘broken from reality’ when offending. This might help to explain why participants appeared to render more punitive legal sentences/sanctions toward an offender with a psychopathic label compared to being labeled with psychotic disorder” (p. 20).

“The effect size for dangerousness in studies comparing an offender with a psychopathic label versus an offender with another psychiatric label was moderated by age of the offender in the vignette, psychiatric label used in the vignette, the type of labeling effect used for psychopathy, and how a study included the psychopathic label” (p. 20-21).

“Most perplexing however, related to other psychiatric labels used in the vignette, was that the only significant effect was for antisocial personality disorders; this means that offenders with psychopathic labels were viewed as significantly more dangerous than offenders with antisocial personality disorders, but not more significantly dangerous than offenders with conduct disorder or psychotic disorder. This appears to indicate that psychopathy is not viewed alone and apart from at least some other mental disorders (at least psychotic disorder and conduct disorder) as more indicative of dangerousness” (p. 21).

“Finally, type of labeling effect moderated results; unlike the effect size for dangerousness in studies comparing an offender with a psychopathic label versus an offender with no label, whether the label effect alone was used in the study was significant, whereas when both label and criterion effects were used, the effect size was nonsignificant and close to zero. As discussed below, this may suggest that certain types of labeling are only effective for certain types of studies involving punishment outcomes and the use of the psychopathic label” (p. 21).

“These results appear to suggest that significant effects on punishment outcomes may be attributable to the general impact of any mental disorder diagnosis, rather than psychopathy specifically. This indicates that concerns about the potential negative effects of the psychopathy label on punishment outcomes may not be deserved for the psychopathy labeling particularly, but mental health diagnoses generally. However, although this is an interesting finding, it is important to point out that in real-life contexts in the courtroom, whether the bias is due to the psychopathic label or another label both still may produce negative outcomes on punishment. Therefore, regarding the ecological validity of this result, this finding does not necessarily affect whether or not negative views toward punishment occurs in cases involving psychopathic labeling but may elucidate why such negative views are elicited” (p. 21).

“Second, although sample or participant type did moderate several effect sizes, the summary effect sizes for criminal justice actors were not significant except for dangerousness in studies comparing an offender with a psychopathic label versus an offender with no label, which was also significant for most sample types studied. This is interesting, as there have been fears that psychopathic labeling and stigma may result in unjustly punitive outcomes in other practices in criminal justice related to punishment, such as probation and sexually violent predator hearings. Psychopathy assessments have been increasingly used within the criminal justice system to inform decisions in court apart from sentencing, particularly by judges and probation officers. These results suggest that the psychopathic label may not affect the perceptions of criminal justice actors regarding punishment as negatively as previously thought, and therefore in practice, it is possible that such labeling may not result in unjustly punitive outcomes for offenders due to the psychopathic label in criminal justice contexts related to punishment. However, recent research has found that the influence of psychopathy on judicial ‘restrictiveness’ in juvenile contexts may function through the perceived dangerousness of the offender. Therefore, although labeling might not indeed affect the perceptions of criminal justice actors as negatively as previously though, the fact that the dangerousness summary effect size for criminal justice actors was significant might have implications for the rendering of legal sanctions” (p. 21).

“Third, the summary effect sizes for the lay public were significant across all punishment outcomes, even for those of which did not have significant summary effect sizes. The one exception was dangerousness in studies comparing an offender with a psychopathic label versus an offender with no label. Although it is unclear on why this exception exists, the fact that the effect sizes for the lay public subgroups were almost entirely significant, even for the outcomes that did not have significant summary effect sizes, suggests that the lay public does indeed subscribe to both general and specific labeling effects for psychopathy when it comes to punishment. This has potential implications for jury sentencing in both capital and, in select states, noncapital cases; as previous literature has cautioned that misconceptions and stigma against psychopathy related to dangerousness, lack of treatment amenability, and violence, may prompt jurors to have a perverted view on what psychopathy actually is and whether it represents an aggravating factor for punishment, this meta-analysis suggests that such concerns on the potential punitiveness of jurors due to the psychopathic label may be justified, and that the potential prejudicial impact of such evidence on psychopathy might indeed potentially outweigh its probative value” (p. 21-22).

“Finally, this meta-analysis may suggest that certain types of labeling effects are only effective for affecting perceptions of punishment outcomes in particular types of studies on psychopathy. Particularly, for dangerousness in studies using a design comparing an offender with a psychopathic label versus an offender with no label, a psychopathic label appeared to have a substantial effect on increasing perceptions of dangerousness only when paired with the characteristics or symptoms of psychopathy (label and criterion effects used together). Conversely, in studies using a design that compares an offender with a psychopathic label versus an offender with another psychiatric label, the impact of the psychopathic label appeared to only have a substantial effect in this area when the label effect alone was used. This suggests that in these psychopathy studies, the general labeling effect related to dangerousness may have the most impact when label and criterion effects are both used, whereas the specific labeling effect related to dangerousness has the most impact when a labeling effect is used alone. Particularly related to the specific labeling effect, this finding could be related to the fact that many personality disorders in the DSM, including antisocial personality disorder and conduct disorder, have similar diagnostic features to psychopathy, and thus, when both label and criterion effects are used, participants may recognize the similarities between disorders’ features and be less likely to view the disorders as significantly different as it relates to dangerousness. However, when only the label effect is used, people might be making judgments based upon the stigma associated with psychopathy—as psychopathy is representative of the cold blooded evil serial killers on TV and in movies—rather than its clinical features, and, therefore, view a labeled offender as significantly more dangerous than offenders with the other disorders” (p. 22).

“[A]lthough we attempted to find unpublished studies by contacting researchers in the field who may have knowledge of unpublished literature and the fail-safe N analyses for significant effect sizes were robust and did not indicate publication bias, it is always possible that we may have missed studies that could have been included in this metaanalysis which may have affected results. Further, the number of studies that could be included in this analysis, and therefore k values used in the calculation of the effect sizes, was low. However, this could not be avoided, as this meta-analysis covers the known labeling psychopathy studies. Given the far ranging and conflicting results of these existing studies, a meta-analysis on these existing studies is beneficial regardless of the study number. In the same vein, one of the effect sizes for psychopathic label versus other psychiatric label (treatment amenability) did not contain at least 10 studies, which Higgins and Altman (2008) argue is the minimum number of studies, in most circumstances, needed in order to conduct moderation analysis. Therefore, although not found to be significant, the results of moderation analysis for treatment amenability should be taken with a grain of salt and might be skewed by the number of studies. As previously mentioned, this could not be avoided based on the number of existing studies, and although a limitation, this moderation analysis is still beneficial” (p. 22).

“Overall, [results] suggests a significant general labeling effect, but not a specific labeling effect, for psychopathy in these studies, and these results suggest that the lay public, but not those in the criminal justice system, may subscribe to both general and specific labeling effects for psychopathy when it comes to punishment and certain types of labeling effects are only influential for certain punishment outcomes, such as label and criterion effects on views of dangerousness. Future meta-analyses should regularly reevaluate the literature on these issues to examine if lay and criminal justice perceptions of psychopathy and punishment change moving forward” (p. 22).

As always, please join the discussion below if you have thoughts or comments to add!