Eating restraint or disinhibition: Which factor better predicts eating in the laboratory
This study was a two-part investigation of the relative abilities of eating restraint and disinhibition to account for eating regulation and total eating. Restraint was measured by the restraint scales of the Three Factor Eating Questionnaire, Dutch Eating Behavior Questionnaire, and Eating Restraint Questionnaire. Disinhibition was measured by the disinhibition scales of the Three Factor Eating Questionnaire and Eating Restraint Questionnaire, as well as the Emotional and External eating scales of the Dutch Eating Behavior Questionnaire. It was hypothesized that disinhibition would predict eating regulation and total eating, while restraint would not. The Restraint Scale, the most commonly employed measure in eating restraint research, was also administered. The Restraint Scale contains subscales measuring Concern for Dieting and Weight Fluctuation. Its predictive ability was evaluated separately. During Phase One, participants completed a variety of self-report questionnaires, including the eating questionnaires cited above. During Phase Two, participants returned and engaged in two separate "taste-rating" conditions, counterbalanced for order. In one condition, participants were given a milk shake preload and then asked to taste and rate three flavors of ice cream. In the second condition, participants were given no preload and then asked to taste and rate three flavors of ice cream. Eating regulation (eating in the no preload condition - eating in the preload condition) and total eating (eating in the no preload condition + eating in the preload condition) were calculated. Four major findings emerged: (1) Disinhibition failed to predict either eating regulation or total eating; (2) Restraint predicted eating regulation, although only in one set of conditions; (3) The Restraint Scale predicted total eating, although only in one set of conditions; and (4) The order in which participants engaged in the eating conditions significantly impacted behavior. Restraint and disinhibition construct and measurement issues are addressed, as are experimental design and implementation issues. The nature of the relationship between the Restraint Scale subscales and total eating is investigated. Eating as a dynamic process is explored, as are the implications of this for future research.
Conger, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our