Key points are not available for this paper at this time.
S ummary . The purpose of this study is to relate student self‐concept in seven different areas to teacher ratings of student self‐concept in the same areas and to academic ability. The results from three diverse samples are considered (total N = 958 students). Student‐teacher agreement on student self‐concept varied from modest (r = 0.24 for Relationship With Parents) to good (r = 0.52 for Mathematics Self‐concept) for the seven self‐concept dimensions (mean r = 0.40), was higher in academic areas than non‐academic areas, and was particularly good for academic self‐concept in the high SES sample (r = 0.74). A multitrait‐multimethod analysis of this student‐teacher agreement demonstrated the distinctiveness of the seven dimensions and argued for the multidimensionality of self‐concept. Academic abilities (reading achievement, IQ, and teacher ratings of ability in different areas) tended to be uncorrelated with non‐academic dimensions of self‐concept, moderately correlated with the academic dimensions, and most highly correlated with the academic dimension of self‐concept that most closely matched the particular academic ability. These correlations were largest in the high SES sample where several correlations exceeded 0.7. Further discussion suggested possible reasons why the results of this study provide better support for the construct validity of self‐concept than is typically reported.
Marsh et al. (Tue,) studied this question.