The Scale of Evil – Interjudge Reliability and Associations with Predictor Variables

Background: The concept of Evil has only rarely been the subject of empirical studies, which in turn requires a distinct definition of the concept, schemes for operationalization and development of measurement tools. The Scale of Evil (SoE) was constructed 1993 but has not yet been studied empirically. Aim: The main aim was to assess the reliability of the SoE. A second aim was to explore the correlation pattern between SoE and a range of other characteristics of perpetrators of crime. This might shed light on how Evil is construed in the minds of forensic professionals, and laymen. Method: 139 forensic psychiatric patients (stratified selection) were scored according to SoE by two independent raters. Psychopathy and HCR-15 (risk assessment) scores, as well as a data on a wide range of other individual characteristics were available for most of the subjects. Results: The interjudge reliability was very high with respect to rank order (tau=.94) as well as distribution of scores (almost identical). Among more than 100 individual characteristics, one variable displayed a particularly strong association with SoE scores: Lack of Compassion (tau=.49). As expected SoE correlated with PCL scores, but actually stronger with a diagnosis of Antisocial Personality Disorder, and item H1 (previous violence) and H8 (early behavioural problems) of the HCR-15. Significant associations were also obtained with many other variables, in line with expectations. Conclusions: SoE is a highly reliable scale. The pattern of associations with the other individual characteristics verifies the importance of psychopathic characteristics when scoring SoE, but SoE goes beyond being” another psychopathy scale”. The Compassion variable, with its roots in criminological theory, appears to be a key to understand how raters construe Evil and rate SoE. The strong association qualifies as a construct validation of the SoE scale