The factor structure of the Tennessee Self-Concept scale among lower-class urban psychiatric patients

Abstract

PROBLEM Fitts@) has operationalized the idea of self-concept by his development of the Tennessee Self-Concept Scale (TSCS). It is assumed to be a multidimensional description of the self that includes self-perception, self-acceptance, behavior, and several measures of external reference: physical, moral, ethical, personal, family and social self. An additional 10 items taken from the L-Scale of the MMPI are purported to index self-criticism. Although ad hoc theorizing about the dimensions of self-concept is a legitimate and useful procedure for item selection, any procedure for the identification and measurement of attitudinal dimensions must, in the final analysis, be put to an empirical test. Factor analysis appears to be the most appropriate method for empirical identification of attitudinal dimensions.This study reports a factor analysis of the TSCS based upon the responses of lower-class urban psychiatric patients. Although the TSCS assumes invariant dimensions of self-concept across populations, we recognize that the factor structure may vary with disparate samples. METHOD Subjects and ProcedureThe TSCS, which consists of 100 self-descriptive statements, was administered individually on admission to 135 female and 117 male psychiatric patients, of whom 219 were inpatients and the remaining 33 were outpatients, at New York Medical College-Metropolitan Hospital Center. The ethnic distribution was : 117 White, 79 Negro, 54 Puerto Rican, and 2 other. A comparable Spanish form was administered to 9 non-English-speaking 8s. The mean age of the subjects was 34.4 years, SD 13.2. Method of AnalysisTwenty principal component factors were extracted from the item correlation matrix with unities in the diagonals. Following the suggested break in latent roots, 6 factors were rotated by means of a Varimax solution and accounted for 35.6% of the variance. Assignment to factors proceeded on the basis of the highest loading (provided it exceeded .30) and low loadings on the other factors. Some few exceptions were made to this general 'rule after the manner of what Cattell calls "watch making". The items assigned to each factor and their loadings are presented in Table 1.' RESULTSFactor 1, which is composed of 27 items and accounts for 14.5% of the variance, clearly represents positive self-evaluation. It contains expressions of satisfaction with one's self across wide areas of functioning, including social, moral, religious and interpersonal behavior. Also included in this factor are reports of competence and responsibility in various areas of living. It might be expected that scores on this factor would be significantly related to the personality construct which is commonly called "ego strength", i.e., freedom from symptoms with fairly adequate defenses, socially appropriate behavior, adaptability under stress, and the capacity for forming close interpersonal relationships.

Rencontres avec des parents de schizophrènes

“…Parfois, il peut se traduire de façon paranoïde par une extériorisation du blâme, par une projection inefficace de la culpabilité. Les recherches démontrent que les schizophrènes se culpabilisent souvent pour des peccadilles (Fitzgibbons et Cutler, 1972); leur dépendance au Bien-être social peut occasionner des hospitalisations répétées à cause de leur manque d'estime de soi (Serban et Gidynski, 1974). Nous avons noté la même tendance chez les parents de nos groupes.…”

A Contribution to the Construct Validity of the Tennessee Self-Concept Scale: A Confirmatory Factor Analysis

“…These procedures, legitimate for other purposes, do not address the question of whether the internal structure of the TSCS is as conceptualized by Fitts (1965). More germane to the present investigation, Bolton (1976), Fitzgibbons and Cutler (1972), Gable et al (1973), Lang andVernon (1977), and Vacchiano and Strauss (1968) have analyzed the TSCS item structure. Although these investigators have collected data from a variety of types of people and have used a variety ot tactor analytic methods, methodological problems are apparent in each of the previous studies.…”

“…The majority of the Fitzgibbons and Cutler (1972) subjects were psychiatric inpatients together with some outpatients. Lang and Vernon (1977) used psychiatric day patients aged 17 to 65 years old.…”

“…Fitzgibbons and Cutler (1972), Gable et al (1973), and Vacchiano and Strauss (1968) performed principal components analyses. Because principal components analysis assumes an instrument has pertect reliability, this represents an inappropriate choice of factor extraction method.…”

“…Despite the fact that Fitts (1965) designed the TSCS subscales to be correlated, Fitzgibbons andCutler (1972), Lang andVernon (1977), and Vacchiano and Strauss (1968) used an orthogonal varimax rotation procedure. More appropriately, Gable et al (1973) used oblimax and Bolton (1976) used oblimin.…”