Generality in the errors of estimation of body image.

An aspect of the considerable research on the association between eating disorders and discrepancies in body image estimation is the extent, if at all, that such discrepancies could be due to a more general tendency to show a positive or negative bias in estimation of the size of objects. One hundred female subjects, without known eating disorder, estimated the width of their own bodies (head, shoulders, waist, and hips), of another person's body (waist and hips), of four commercial packages, and of a cylinder. Errors in estimation, computed as algebraic differences from the true value, were computed, and all error means were found to be positive. Correlations between the error scores were all positive. Factor analysis of the correlations between the 11 error scores indicated a substantial general factor plus a bipolar factor, or alternatively two positively correlated factors referring to human body or physical measures. Results were interpreted as supporting the notion of a broad general tendency to enhance or diminish estimations of the width of familiar objects including estimations of the width of own body parts. The relevance of this finding to eating disorders was discussed.