Do Subject-Specific Self-Concepts Influence Relevant Students’ Achievements Across Racial Groups and Subject Matters? Findings Using TIMSS (Trends in International Mathematics and Science Study), PISA (Programme for International Student Assessment), and PIRLS (Progress in International Reading Literacy Study) Data

• In Event: Roundtable Session 32
  In Roundtable Session: Do Various Teaching Approaches and Students’ Self-Concepts Affect Achievement Across Racial Groups and Subjects Differently? Findings Using the Data From Large-Scale International Studies

Abstract

Subject specific self-concept is assumed to predict students’ achievements in that subject (Bandura, 1989). This assumption is reflected in the teaching standards aimed to reform teaching in various subjects in that development of student interests and values of learning should drive its teaching design and implementation. However, the empirical literature does not support this assumption in two situations (Pajares & Schunk, 2001). First, across-country comparison showed consistently that higher students’ self-concepts in a country were often negatively related to their academic performances or vice versa (Kennedy & Trong, 2007; Shen & Tam, 2008; Wang, 2007). This finding indicates that the direct positive and causal relationship between student self-concepts and achievements may not exist as the social, cultural, and school environments in each country may influence both self-concepts and achievements and confound the assumed relationship (Wang & Lin, 2008). Second, across-racial comparison in the U.S. also shows anecdotally that the lower performance racial groups often developed higher self-concepts (Graham, 1994; Pajares & Schunk, 2001). This study further verified the above finding from across racial comparison and examined whether and to what extent the alternative conception for the findings from cross-country comparison (Wang & Lin, 2008) can be sustained in the context of cross-racial comparison.

Using U.S. data sets from TIMSS 2007, PISA 2006, and PIRLS 2006, this study identified mathematics, science, and reading achievement scores of Caucasian, African, Hispanic, and Asian American students as dependent variables. The independent variables were student self-concept scores composted by combing all items related to self-concepts in the student survey for each racial group. The descriptive statistics and multiple regressions were conducted to check the relationship between independent and dependent variables within a racial group first and then across racial groups in mathematics, reading, and science.

This study found that in TIMSS and PISA data student self-concepts in mathematics and science were negatively related to the scores of Caucasian, African, Hispanic, and Asian American students but in PIRLS data student self-concepts were positively related to reading scores of Caucasian, Hispanic, and African American groups. The findings challenge the popular assumption of the positive and causal relationship between subject specific self-concepts and student achievement in relevant subject areas (Bandura, 1989) with paradoxical relationships between students’ self-concepts and their achievements found in mathematics, reading, and science within each racial group. It also showed that there was an inconsistent relationship between independent variables and dependent variables in each subject area in the cross-group comparison, which suggests that the social, cultural and school environments where each group was situated may influence both self-concepts and achievements and confounded the popular assumption of the relationship (Wang & Lin, 2008). The study calls for further conceptions and examinations about how the above environments mediate self-concepts and students’ achievement in each subject area within and across each racial group.

Authors

• Qingmin Shi, University of Nevada - Las Vegas

• Jian Wang, University of Nevada - Las Vegas

• Shaoan Zhang, University of Nevada - Las Vegas