A new scale for use in non-clinical research into disinhibitive eating

Summary-The DIS, a 17-item questionnaire, was developed for subject selection in non-clinical research into disinhibitive eating. The DIS was designed to possess a tight and clear subscale structure, lessened impact of concern for dieting and explicit reference to self-esteem. Four subscales tapped the following themes: failed restriction, binge eating, weight fluctuations and self-esteem. DIS and the often used Restraint Scale (RS) were tested psychometrically among 269 female students from the Netherlands. Cronbach alpha was 0.88 for the DIS and 0.84 for the RS; correlation between the scales was 0.78. Factor analysis of the RS yielded the traditional two factors concern for dieting-weight fluctuations, whereas the DIS had a somewhat clearer four factors structure in accordance with its a priori subscales. On both the DIS and RS, high scorers were often currently dieting, and had higher Body Mass Index than low scorers. As to selection, DIS and RS differed in that partly different groups of individuals were selected when applying equally stringent cut-off scores. It was concluded that some of the psychometric and conceptual problems with the RS are alleviated by the DIS and that a next step is to behaviourally validate the DIS in disinhibition research. Copyright 0 1996 Elsevier Science Ltd