James P. O'Connor

Some Patterns of Hypochondriasis

WITHIN the past few years there have been several factor analytic studies of the items of the original clinical and more recently derived scales of the Minnesota Multiphasic Personality Inventory (1, 2, 3, 5, 6, 7). These investigations were prompted by concern over the obvious lack of homogeneity of items comprising the various scales. Even the most casual inspection of the content of the MMPI scales suggests that many, if not all, are functionally complex. Subsequent factor analyses have confirmed these suspicions. The scales studied thus far have been demonstrated to be groupings of items each of which contains more than one independent psychological dimension. If the structure of a scale is complex, the psychological meaning of the total scale score is ambiguous and its predictive and diagnostic value may be seriously reduced. One solution to this problem appears to be the regrouping of test items into homogeneous clusters which closely approach unitary significance. It was with the hope of contributing toward this end that the present study was conducted. More specifically, our purpose was to isolate and identify, as far as possible, the factorial dimensions of the MMPI hypochondriasis scale. Before this study was completed, Comrey (1) published the results of a factorial study of the Hs scale on 360 cases mixed with respect to diagnosis, sex, and educational level. With the aid of an electronic computer, the author extracted 12 centroid factors which