Psychometric Properties and Dimensionality of the German Beck Scale for Suicide Ideation (BSS): Validation in a Sample of Psychiatric Patients

Abstract

The present study examines dimensionality, reliability and convergent validity of the German Beck Scale for Suicidal Ideation (BSS) in a clinical sample. 308 inpatients after suicide attempts/acute suicidality participated in the study (53,6% female). Of those, 224 completed the full BSS and self-report questionnaires assessing depression (DESC), hopelessness (BHS), interpersonal variables (INQ) and defeat (DS-d)/entrapment (ES-d). Dimensionality was investigated by confirmatory factor analysis (CFA, models with 1, 2, 3 and 5 factors) and convergent validity was investigated by correlational analysis (Pearson). In total, the 5-factor model achieved the best fit. Yet, model fit is comparable between all tested models without considering the RMSEA model. The multidimensional models result in similar subscales. Subscales reflecting passive death wishes (α>0,80), active suicidal ideation (α>0,73) and suicide-related behaviors (α>0,70) achieve acceptable internal consistency. The BSS sum score and the subscales assessing passive and active suididal thoughts correlate moderately positive with DESC, BHS, INQ and DS-d/ES-d (r between 0,25 and 0,66) while the behavior-related scale shows smaller (r between 0,14 und 0,27) or no associations (INQ, thwarted belongingness). The CFAs do not provide clear evidence for either an uni- or a multidimensional structure of the BSS. In the light of this finding, use of the BSS score is limited despite evidence supporting its reliability and convergent validity.