Self-efficacy, outcome expectations, and outcome value: prediction of weight-loss program outcomes

This study was planned to assess the usefulness of Bandura's self-efficacy theory for prediction of attrition among women enrolled in a weight-loss program. It was hypothesized that clients who scored higher in self-efficacy, outcome expectancies, and/or outcome value would be more likely to continue in the program than those with lower scores. Another purpose of this study was to investigate the utility of these same expectancy concepts for prediction of weight loss among the clients. Again, it was hypothesized that persons who scored high on the expectancy measures would be more likely to lose more weight and body mass than those with lower scores.

Subjects were 42 female clients enrolled in a 12-week comprehensive outpatient weight-loss program. Both weight and body mass index (BMI) change were assessed. Attrition was measured in terms of completion of the 12-week program, and in terms of continuation with a follow-up treatment group at termination of the 19-week study.

The hypotheses regarding attrition were partially supported: clients who dropped out of the 12-week program had lower initial self-efficacy scores than did continuers, and those who were still attending treatment groups at the conclusion of the study initially scored higher on outcome value than did noncontinuers. However, outcome expectancies failed to predict attrition. Clients with relatively smaller BMIs were more apt to drop out of the 12-week initial treatment group, and those who were continuing with further treatment groups at the conclusion of the study were significantly younger than all noncompleters.

Contrary to predictions, total outcome expectations were related to lower weight losses during the 12-week program, and relatively greater positive outcome expectations were related to lower weight losses during the entire 19 weeks of the study. Self-efficacy and outcome value were not related to weight and BMI loss. The most powerful predictor variable of weight loss at follow-up was initial weight. Those who initially were the heaviest lost the most weight and body mass, and they tended to score lowest on initial social desirability. Another factor which was associated with weight and body mass loss during treatment was frequency of previous dieting attempts within the last three years: clients who reported having made relatively fewer previous dieting attempts lost more weight and body mass than did clients who reported having made more attempts. In addition, clients who reported having made relatively fewer previous dieting attempts had lost more weight at follow-up. No other factors were significantly related to weight or BMI loss, either at post-treatment or at follow-up. Weight/BMI loss was not related to these variables: age, marital status or having children, reporting a health problem related to weight loss, having had lost weight quickly or slowly in previous weight control attempts, previous participation in formal weight loss programs, or attitudes of husband/significant other or clients' children toward clients' involvement in the treatment program.

It is concluded that self-efficacy expectancies and outcome value expectations appear to be more useful for prediction of attrition from a weight-loss treatment program than for prediction of weight or BMI loss. Self-efficacy and outcome value measures may be found to be useful in screening for clients who are most at risk for terminating weight-loss treatment, in order that early interventions may be made. All in all, it appears that the effect strength of self-efficacy concepts tends to decrease as predictor tasks increase in complexity. There is an indication that a history of prior dieting attempts impedes weight loss during initial treatment programs. In addition, there is support for recent claims that longer weight-loss treatment may result in larger weight losses. Clients who continued with the program lost an average of 21.6 pounds after 19 weeks.