Características psicométricas de una adaptación española del Test de Bulimia (BULIT)

INTRODUCTION: The high prevalence of bulimia nervosa among eating disorders and limited repertoire of assessment tools available for these disorders in Spanish have resulted in special interest regarding the adaptation of a specific instrument such as the Bulimia Test (BULIT). METHODOLOGY: A total of 119 outpatients, most of them women (97.5%) seen in a community mental health center, were assessed with the Spanish adaptation of the Bulimia Test (BULIT) and by Eating Attitudes Test (EAT-40). Reliability (internal consistency and temporal reliability) and validity (convergent and discriminant) were studied. RESULTS: Cronbach's alpha coefficient (alpha: 0.93) shows high internal consistency while the temporal reliability is supported by the high correlation observed (r=0.83) between applications conducted in an interval of 8-10 weeks. In relationship with convergent validity, moderate correlations were observed with the global score on EAT (r=0.38) and substantial correlation with the EAT Bulimia subscale (r=0.69). By using an 88 cutoff on the BULIT, the test properly classified 90% of the bulimic subjects and 100% of those without eating disorder. CONCLUSIONS: The results firmly supported the reliability and validity of this Spanish adaptation, stressing its utility as a screening device to detect current or incipient cases of bulimia and to assess severity of bulimic symptoms.