Evidence for the concurrent and construct validity of the Bulimia Test (BULlT) was examined. Female university undergraduates (n = 2 73) were administered the BU-1IT and multiple questionnaires. As expected, the BUllT correlated most highly with tests of theoretically related constructs, although a moderate relationship with response set measures was found. In a sample of bulimic women, daily reports of vomiting, laxative use, and binge eating were compared with BULlT scores. In addition, changes in BUllT scores following treatment were compared with changes following nontreatment. Results suggested that the 5ULlT could be a valuable measure for use in intervention studies, but should not yet be used instead of daily reports of behaviors.
The Bulimia Test (BULIT) was developed in 1984 (Smith & Thelen) to distinguish between bulimic and non-eating-disordered individuals, between subtypes of bulimics based on symptomatology, and to assess severity of symptoms, for example, between current and treated bulimics. In a series of initial studies (Smith & Thelen, 1984), the BULIT successfully distinguished between a bulimic sample and a normal, university sample and demonstrated high specificity and moderate sensitivity as a screening device. The scale correlated highly with the Eating Attitudes Test (EAT) (r = 0.65; Garner & Garfinkel, 1979) and with the Binge Scale (r = 0.93; Hawkins & .
Despite these positive initial results, further studies extending the BULIT's development have primarily comprised factor analytic studies (Stein & Brinza,