Racial Bias in Teachers' Evaluation of Students' Delinquent Behavior: Race Matters.

Using waves 1 and 2 of National Educational Longitudinal Survey (NELS) data we used Hierarchical Linear Modeling (HLM) to test for racial bias in teachers' evaluations of students' delinquent behavior. Approximately 24 percent of the variance in teachers' report of student delinquent behavior was explained by knowing the nesting by teacher (level 2), while only 8.46% of the variance in students self-report of delinquency was explained by knowing the teacher nesting. In order test for teachers' racial bias, we used HLM and regressed teacher report of students' delinquent behavior on ethnicity, self report of delinquency, parent report of student's delinquency, self report of arrests, SES, grades, student's gender, teacher's race, teacher's gender, and region of the country. All variable except for teacher's gender were statistically significant. We found strong evidence for racial bias in teacher report of student delinquency in that student's ethnicity was still statistically significant even after controlling for self-report of delinquency, parent report of delinquency, gender, SES, grades, teachers gender, teacher's ethnicity, and region of the country. Implications of our findings as well as suggestions for reducing teacher's racial bias are discussed. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]