Along with teachers' educational background, skills, and training, researchers have examined the effect of teachers' racial/ethnic background and gender on the achievement of those from a similar or different background, with mixed findings. Hanushek (1992) found that Black teachers were more effective than White teachers in improving scores of Black children in Baltimore. Further, it appears that Black teachers from low-income backgrounds were most effective in teaching low-income Black students. More recently, Dee (2001), in an analysis of data from Tennessee's Project STAR class-size experimental evaluation, found that having a teacher of the same race may improve academic performance among a sample of children and young adolescents. Students were randomly matched with teachers, so racial pairings were assigned independently. Results suggest that a one-year assignment to an own-race teacher increases achievement in math and reading scores by around three to four percentile points (Dee, 2001). Notably, these studies included samples of children who were not all adolescents. However, in an analysis of a national sample of high school students, an age range of focus in this review, Ehrenberg, Goldhaber, and Brewer (1995) found that although teachers' race, ethnicity, and gender appear to influence the subjective evaluations given to students, such as teacher expectations that the student would go to college and whether they would recommend the student for academic honors, their background characteristics did not influence student learning. Further, these differences in teachers' subjective evaluations did not seem to lead to bias in their objective scoring of students. These findings held after controlling for prior test scores, individual and family characteristics, school characteristics, 10th grade subject teacher, and the 10th grade subject class.