Sexual Sadism Among Sex Offenders in Switzerland.

Abstract

The Sexual Sadism Scale (SeSaS) was developed to assist in the diagnosis of sexual sadism, and it revealed adequate psychometric properties in prior research. This study cross validated the SeSaS in Switzerland using a sample of 179 male sex offenders. Specifically, the SeSaS conformed to a Mokken model of double monotonicity (scalability coefficient [H] = .46, coefficient of reproducibility [CR] = .89), indicating that it measures a unidimensional construct of sexual sadism with hierarchically ordered items. The reliability of the scale was acceptable to high (ρ = .80, λ2 = .75, κ = .88). In addition, the SeSaS was strongly associated with sexual sadism diagnoses based on mental health manuals (rpb = .60, odds ratio [OR] = 13.02, area under the curve [AUC] = 1) but not with recidivism. The results suggest that the use of the SeSaS may improve the validity and reliability of sexual sadism diagnoses, therefore playing a role in the assessment and management of sex offenders.