Relation of dietary restraint to bulimic symptomatology: The effects of the criterion confounding of the restraint scale

Studies indicate that various measures of dietary restraint show inconsistent relations to bulimic symptomatology. The present study tested the assertion that this is because the scales differ in the extent to which they reflect the success or failure of dietary efforts. This study also tested the competing hypothesis that criterion confounding of the Restraint Scale produced the inconsistent findings. Data from 117 undergraduates indicated that both the Restraint Scale and the Dutch Restrained Eating Scale were correlated with bulimic pathology, but that the magnitude to the relations were higher for the former. These results supported the dietary success-failure hypothesis, however, this difference disappeared when the two disinhibited eating items from the Restraint Scale were deleted. Thus, the inconsistent findings in prior research appear to be due to the criterion confounding of the Restraint Scale. Implications for the measurement of restraint are discussed. Copyright (C) 1997 Elsevier Science Ltd.