Success and failure in the measurement of restraint: notes and data.

OBJECTIVE: The clarification of the present confusion about the confounding of success and failure in the measurement of restrained eating.
METHOD: Close inspection of results regarding the Restraint Scale (RS) and the restraint scales of homogeneous measures of restrained eating, for example, the restraint scales of the Three-Factor Eating Questionnaire (TFEQ) and the Dutch Eating Behaviour Questionnaire (DEBQ).
RESULTS: Various studies have led to the distinction of two different types of confounding of success and failure in the measurement of restrained eating. The first confounding is associated with the RS and the inability of its scores to differentiate between inhibition and disinhibition of restraint in one person, as a result of variance across time within one person in food intake and binging. The second type of confounding applies to the homogeneous measures of restrained eating, for example, the restraint scales of the TFEQ and DEBQ. Here the confounding is associated with the inability of their restraint scores to differentiate between dieters with low versus high susceptibility toward failure of restraint, as a result of variance across persons in food intake and binging. DISCUSSION: In contrast to the first confounding, the confounding of successful and unsuccessful dieters in subjects with high restraint scores can be easily unconfounded, by using a two-factorial classification including both scores on "pure" measures of restraint and scales for tendency toward overeating. For future research, the distinction between successful and unsuccessful dieters is an important first step in clarifying why some dieters do overeat under certain conditions.