The article by Gordon et al. 1 presents a very understandable description of taxometric analyses. These writers are to be commended for taking this complex topic and making it comprehensible to those who are not experts in the field of taxometrics and classification research. Gordon et al. 1 conclude that the results of taxometric studies suggest the existence of three diagnostic groups, bulimia nervosa, anorexia nervosa, restricting type, and binge eating disorder. They recommend conceptualizing bulimia nervosa and binge eating disorder as categorical classes, but conceptualize anorexia nervosa as a dimensional or continuous entity. Wonderlich et al. 2 take a step back from the esoteric specifics of classification research and note that conclusions about the validity of any classification scheme will necessarily involve a series of value laden decisions.Both articles 1,2 report on recent scientific findings related to the validity of classification schemes of eating disorders. Wonderlich et al. suggest a comprehensive research strategy for validating eating disorder classification models. Both articles 1,2 observe that eating disorders involving binge eating can be conceptualized as qualitatively distinct types or classes and they conclude that psychopathological and personality features other than frank eating disorder symptoms, e.g., concern with body size/shape, binge eating, purging, and restrictive eating, are important aspects of eating disorder diagnoses.
Factor Analytic Studies of Eating Disorder SymptomsEight factor analytic studies of eating disorder symptoms 3 have yielded findings that are relevant to the conclusions of the Gordon and Wonderlich articles. All of the factor analytic studies that included measures of ''general psychopathology'' found a very strong and stable factor that reflected personality characteristics and ''neurotic'' symptoms, e.g., mood disturbances, anxiety, obsessive thinking. The factor analytic solutions indicated that this ''general psychopathology'' factor was separate and distinct from other factors that reflected symptoms generally regarded as specific to eating disorders, e.g., binge eating, restrictive eating, fear of fatness, and drive for thinness. These findings suggest that a diagnostic model that only specifies personality characteristics, e.g., the model of Figure 1d in the Wonderlich article, would be inadequate to capture a complete set of relevant features of the eating disorders. The results also suggest that classification models that ignore personality and general psychopathology features of eating disorders, e.g., the models of Figures 1a, 1b, and 1c, will also be inadequate. In my view, the personality features of eating disorders can be viewed as conditions that sometimes accompany the more specific behavioral and attitudinal features of eating disorders. A second consistent finding of the factor analytic studies was the presence of a strong and stable binge eating/bulimic behaviors factor. The fact that binge eating has also been identified as a late...