Selbstkonzepte von Schülern der Klassenstufen 3 bis 6: Messung und Validierung der multidimensionalen Struktur

Abstract

In the field of research on the multidimensional self-concept of preadolescent children it has remained an unresolved issue, whether academic self-concept is not only domain-specific but can be additionally differentiated into competence and affect components. A separation between competence and affect components of academic self-concept would not have been theorized by the classical self-concept definition of Shavelson, Hubner, and Stanton (1976) but is consistent with the assumptions of numerous theories in the field of research on achievement motivation. In a confirmatory factor analysis, Marsh, Craven, and Debus (1999) provided first empirical evidence of the separation between competence and affect components of academic self-concept. However, the study of Marsh et al. (1999) is exclusively based on the within-network approach to the study of construct validation. As such, the present study aimed to replicate the findings of Marsh et al. (1999) and to extend them by between-network studies. As existing German self-concept measures are not suited for examining the differentiation of academic self-concept into competence and affect components, a German version of the Self Description Questionnaire I (SDQ I; Marsh, 1990c) was constructed and empirically evaluated. A confirmatory factor analysis on this instrument replicated the findings of Marsh et al. (1999) with a German sample and thus, provided additional within-network evidence of the separation between competence and affect components of academic self-concept. In addition, as the competence and affect components were found to be differentially related to academic achievement, this study offered first between-network evidence of the separation between competence and affect components of academic self-concept. The competence component was found to be more highly correlated to academic achievement than the affect component both within and between math and verbal do mains. As such, by proofing the separation between competence and affect components of academic self-concept by using both the within-network and the between-network approach of the study to construct validation, the competence-affect separation of academic self-concept can be seen as a rigorously validated feature of academic self-concept. By constructing and evaluating a German version of the SDQ I, the present study also contributes to the field of self-concept measurement and gave insight into further possibilities of advancing and refining the SDQ I instrument.