Psychometric properties of the Stirling Eating Disorder Scales with bulimia nervosa patients.

BACKGROUND The Stirling Eating Disorder Scales (SEDS) are described as a well-validated measure of anorexic and bulimic cognitions and behaviors. However, there is little published research that demonstrates or tests the psychometric properties of the SEDS. The aim of this study was to assess the internal consistency of the SEDS with a clinical population at a specialist eating disorders service. METHOD This study was a cross-sectional component of a larger study (evaluating cognitive behavioural group treatment for bulimia nervosa). At assessment at a specialist eating disorders service, the SEDS was administered to a clinical sample of women with a diagnosis of bulimia nervosa (N=40). Cronbach's alpha coefficients were used to determine the internal consistency of the scales. RESULTS While the internal consistency of the overall SEDS was good, the existing SEDS subscales lacked internal consistency. Only one of the eight subscales (perceived external control) had an acceptable alpha (>.700). By removing up to three questionnaire items for each subscale that most impaired the alpha, only two further subscales were raised to an acceptable alpha. The sample's scores did not closely reflect the reported norms. DISCUSSION More studies are needed to determine the clinical and psychometric validity of the SEDS. However, the present findings indicate that the preliminary psychometric findings do not generalize to other adult clinical populations. Therefore, caution should be used when interpreting the results of the SEDS when working with adults with eating disorders.