Abstract
Subclinical sadism has received substantial attention in recent research as a trait that predicts a variety of malevolent behaviors. The objective of this study was to assess the ‘psychometric robustness and portability’ of the Assessment of Sadistic Personality (ASP). We examined the convergent and discriminant validity, and invariance of translated versions of the ASP within community samples of Polish and Italian individuals. The study included 568 individuals (340 women and 228 men) residing in Italy (Mage = 23.57, SDage = 2.55) and 556 individuals (411 women, 144 men, 1 other) residing in Poland (Mage = 23.48, SDage = 4.60). For cultural invariance purposes, data from a Canadian sample comprising 638 students were used. To establish convergent and discriminant validity, participants completed measures of sadism, the Dark Triad, the Big Five, interpersonal reactivity, and maladaptive traits described in the DSM-5. Across both samples, convergent and discriminant validity were supported. Configural and partial metric invariance were satisfied, and following implementation of alignment optimization, latent mean differences were evaluated between countries. Results of the study supported the psychometric qualities of the ASP across different cultures and languages, and the utility of the ASP as a valid measure extending beyond university samples.
Keywords:
Notes
1 We analyzed whether sadism and psychopathy form two distinct factors and found that despite being significantly correlated, they do indeed load on two separate factors. The detailed description of the analysis with relevant syntaxes are available on our OSF project site.
2 Clinical sadism and sexual sadism have been excluded from the meta-analysis because the main purpose of the meta-analysis was to inform our hypotheses. Confounding different types of sadism with everyday sadism may have given us different results as these are distinct constructs; thus giving us less pure estimates of relationships between everyday sadism and external correlates.
3 The Polish version of the SSIS was translated following the back-translation procedure by one of the authors of the paper for the purpose of the present study. The Italian version of the SSIS was translated following the back-translation procedure by the research group in criminological and forensic psychology led by Prof. Georgia Zara (Department of Psychology, Turin University).
4 We have also assessed the incremental validity of ASP sadism over psychopathy, as well as the incremental validity of the ASP over the SSIS in predicting the external measures used in our study. To assess this, we used multivariate regression and commonality analyses (which we interpreted when R2 was ≥ .20). We found that the ASP predicted (over the effect of psychopathy) agreeableness (11.83% uniquely explained variance by ASP; 65.74% in common with psychopathy), IRI empathic concern (8.51% unique, 45.26% common), IRI perspective-taking (44.14% unique, 46.14% common), and antagonism (44.07% unique, 46.17% common). Moreover, the ASP uniquely predicted (over the effect of the SSIS) agreeableness (32.78% unique, 66.52% common), IRI empathic concern (17.35% unique, 75.89% common), IRI perspective-taking (39.65% unique, 60.34% common), and antagonism (18% unique, 75.66% common). Our findings can be found on our OSF page.