|dc.description.abstract||In comparison to white students, African-American students typically display relatively high academic self-concept despite having relatively low achievement. The literature suggests that one reason for this phenomenon is black students’ tendency to discount objective performance measures (e.g., grades and test scores) when constructing their academic self-beliefs. This dissertation asks why black students are more prone to discounting performance measures when self-evaluating their own academic abilities. Guided by the sociology literature on racial mismatch, and the social psychology literature on performance discounting, this thesis argues that black students’ tendency to discount performance measures is structurally patterned in terms of racial mismatch between teachers and students. The literature on racial mismatch has seldom considered its impacts on the discounting process, while the literature on discounting processes has given little evidence of how these theories apply in real school and classroom settings. This dissertation addresses these two shortcomings. There is also relatively little knowledge about how racial mismatch effects vary across different school environments and across different measures of performance. This study tests how discounting processes are moderated by broad school contexts, and conducts separate analyses for test scores, teacher ratings, and parental ratings. Finally, racial disparity in performance discounting has almost exclusively been described in terms of between-person associations based on an external frame of reference (e.g., peer comparisons, etc.), and little is known about the disparity in terms of within-person evolution based on an internal frame of reference (i.e., comparison to one’s own prior test scores). Furthermore, there is little knowledge of how racial disparity in performance discounting may grow or decline over the course of childhood. This study explores the discounting process from a developmental perspective. It uses three waves of data from a national survey, the Early Childhood Longitudinal Study (ECLS-K).
This dissertation found that the performance discounting is significantly greater for blacks than for whites, across broad school context and achievement ratings. This racial difference, however is not explained by racial mismatching of students and teachers uniformly for all schools. The results suggest that the effects of racial mismatch may be moderated by the school’s racial composition. In all other school settings, racial mismatch had no relationships to the racial disparity in performance discounting. The study also found that performance discounting weakens over time, but that the racial disparity remains similar throughout schooling. This developmental trend is not explained by racial mismatch. Finally, the study found that, though performance discounting is greater for blacks than for whites at any point in time, blacks’ academic self-concepts respond to changes in their own performance just as strongly as do whites’. Implications for relevant issues and policy are discussed.||