PROBLEMThere is general agreement, both in self-concept theory and in experimental evidence, that a negative self-concept indicates stress and tension within the individual. In contrast, interpretations of the significance of a positive self-concept have varied. A large body of research, identified with Carl R. Rogers and the Chicago School, was based on the assumption that a positive self-concept indicated a well-adjusted individual. Both in theory and research design, this group stressed the importance of self-integration with the implicit assumption that the wellintegrated individual would also be integrated into his society.Rogers and Dymond (I1) have recently suggested that the positive self-concept may also indicate a highly defensive paranoid individual. Frenkel-Brunswik and Hart(6) offer further support of the theory that positive self-concept, for some individuals, may be compensatory in nature. Hart recognized the difficulty in differentiating between true self-esteem and inflated self-esteem. The possibility that high self-esteem can compensate for underlying negative attitudes toward self has important theoretical implications.The purposes of this study are to examine the heterogeneity of the positive self-concept group, and to determine whether homogeneous sub-groups are related to external criteria. A positive self-concept is defined as a small discrepancy between self and ideal-self ratings. It was hypothesized that quantitative differences exist within a group of individuals expressing positive self-concept.METHOD SeZ -Concept. The experimenter constructed a Self Description Inventory (SDI), a measure of sedconcept which is expressed as a discrepancy (D), and which provides in addition] self (S) and ideal-self (I) scores for each subject roviding a basis of identifying quantitative differences within the positive self-concept group. #he portion of the SDI used in this study consists of 60 items rated on a five point scale (a) as the individual describes himself and (b) as the individual would ideally like to be. The discrepancy between the two responses constitutes the score for each item; item discrepancies are totaled to yield a measure of self-concept (D). Reliabilities of the three SDI scales are shown in Table 1. Detailed information about item selection, derivation of TABLE 1. RELIABXLITKES OF SDI SCALES SDI Scale 1 N Splitrhalf I N Teahetest* Discrepancy 110 .93 215 .68 .90 215 .83 .90 215 .39Self Ideal-self *Two months elapsed time.self and ideal-self scores, and reliabilities of parts of the scales can be found in the author's doctoral dissertation (7). Positive Self-Concept. The distribution of SDT discrepancy scores was divided into quarters, The highest quarter (D4) was defined as the negative self-concept group. The lowest quarter (D1) was designated the positive self-concept group and was treated as a discrete sample. Self scores of D1 (positive self-concept) individuals were ranked and the distribution WBB divided into thirds. The lowest third of the self scores was designated sll t...