Self-prediction of the effectiveness of treatments for weight-loss
The present study investigated the accuracy of self-prediction using weight-loss as a demonstration. It was hypothesized that subjects would be able to accurately predict which of three treatments would be most effective, moderately effective, and least effective for them in regard to losing weight. In addition, subjects weight-loss predictions (idiographic approach) were contrasted with the mean weight-loss of the group (nomothetic approach) in order to determine which approach would yield a more accurate estimate of actual weight-loss. Forty-eight subjects were presented with detailed descriptions of the three weight control treatments (cognitive ecology, physical activity, and self-monitoring treatments), and were asked to predict how much weight they thought they would lose in each treatment. Subjects then participated in a ten week weight control program consisting of a baseline week and the three, three-week treatments mentioned above. The treatments were presented in three different orders, in order to assess the presence of order effects. Subjects were randomly assigned to one of the three orders in which the treatments were presented. The data revealed that order effects did not appear to be present; that overall, the program resulted in significant weight-loss for subjects; that each of the three treatments resulted in weight-loss; and that the self-monitoring treatment was the most effective, the physical activity somewhat less effective, and the cognitive ecology treatment was the least effective. Subjects failed to accurately report the relative effectiveness of treatments for them, and the nomothetic approach was superior to the idio-graphic approach as an estimate of subjects1 actual weight-loss. Subjects' predicted weight-loss was significantly greater than their actualweight-loss. Several factors which may have influenced subjects to inflate their weight-loss predictions were discussed: 1) faulty information, 2) demand characteristics, and 3) the manner in which the predictions were elicited. The need for further research on factors which influence subjects' predictions was suggested.