Students' attitudes toward teachers' ethnicity, bilinguality, and gender.

The ethnic, linguistic, and gender composition of public school students differs greatly from that of teachers. According to recent figures from the National Center for Education Statistics (1998), in the fall of 1995, the percentage of ethnic minority students reached 35.2% of the total enrollment nationwide. Similarly, the number of Limited English Proficient (LEP) students grew by 1 million in the last 10 years to include 5.5% of all public school students (August & Hakuta, 1997). These numbers translate to at least one LEP student present in approximately 15% of all classrooms (Fleischman & Hopstock, 1993). In contrast, kindergarten through 12th-grade teachers continue to be predominantly White (90.7%) and female (74.4%) (National Center for Education Statistics, 1998), with only 42% of teachers of LEP students sharing a non-English language with their students (Fleischman & Hopstock, 1993). No immediate change in these trends is expected in the immediate future (National Center for Education Statistics, 1998).

In the fall of 1996 in California, the state in which this study took place, the ethnic composition of public school students was 40.5% Hispanic;(1) 38.8% White; 8.8% Black; 11.1% Asian, Filipino, and Pacific Islander; and 0.9% Native American (California Department of Education, 1998). Regional distribution patterns among minority teachers tend to reflect those of students, yet the magnitude of the disparity between proportions of ethnic minority students and teachers persists even in San Francisco, the county with the highest proportion of minority students (87.2%) and teachers (43.4%) (California Department of Education, 1998). In 1997, California public schools also served 1.3 million LEP students, or 30% of the national total (August & Hakuta, 1997). As with teachers nationwide, in the fall of 1996, California teachers were also mostly White (78.8%) and female (71.6%) (California Department of Education, 1998).

Differences between students' and teachers' backgrounds have prompted scholars and policy makers to call for greater teacher diversity in public schools (King, 1993; Nieto, 1996; Sleeter, 1993). These calls stem from a variety of concerns regarding the significance that background differences have on the teachers' and, especially, the students' lives. A particular view describes teachers' and students' background differences as contributing to create cultural discontinuity between the students' communities and schools (Baratz & Baratz, 1970; Nieto, 1996; Ogbu, 1978, 1995; Ogbu & Matute-Bianchi, 1986). According to this view, cultural discontinuity manifests itself in the classroom as clashing discourse patterns (Cazden, 1986; Heath, 1983; Philips, 1983), conflicting practices (McDermott, 1987; Simpson & Erickson, 1983; Spindler & Spindler, 1990; Tharp, 1989), and discordant values (Sue & Padilla, 1986) and beliefs (Metz, 1990; Tewell & Trubowitz, 1987. Vogt, Jordan, & Tharp, 1987).

Other researchers have assessed the significance of either ethnic or gender differences between students and teachers in various manners, On one hand. there is substantial evidence of teachers' biases against either minority students (Graybill, 1997; Sheets, 1996; Tettegah, 1996), female students (Acker, 1983; Shepardson & Pizzini, 1992; Xue Lan, 1996), or both (Avery & Walker, 1993). On the other hand, Ehrenberg, Goldhaber, and Brewer (1995) found no observable effects of matching student-teacher ethnicity or gender on students' academic achievement, although ethnicity and gender did influence teachers' subjective evaluation of students.

To better understand the significance of student-teacher background differences, in this study I will examine student attitudes toward teachers. I will offer a conceptual framework and argue for the need to understand the problem from the students' perspective, focusing on ethnicity, bilinguality, and gender. After describing the study and presenting the results, I will conclude by discussing the practical and theoretical implications of the findings.

Conceptual Framework

An underlying assumption behind this study is that interactions between students and teachers play a crucial role both on the students' academic future and on the teachers' sense of self-efficacy. The quality of student-teacher interactions has a profound and lasting effect on the students' socialization process in schools (Garcia, 1994). Furthermore, differences between students' and teachers' backgrounds influence the ability of teachers to motivate or to be role models for students (Barnhardt, 1982).

A second assumption in this study is not only that prejudices and social stereotypes influence teachers' perceptions of and behavior toward students (Eccles & Jussim, 1992; Pollard, 1989; Rist, 1970; Rosenthal & Jacobson, 1992; Rubowitz & Maehr, 1973) but also that similar attitudes influence students' perceptions of and behavior toward teachers. This assumption emerges out of attitude theory, which defines the relationship between attitudes and perception as circular (Allport, 1935), and from research documenting mutual influences between people's attitudes (Condon & Crano, 1988; Hays, 1984; Kenny & La Voie, 1982). Thus, attempts to understand the relevance of background differences between students and teachers must include more than just the teachers' perspectives.

Attitudes, Behavior, and Group Membership

Attitudes consist of systems made up of three different and interconnected components: affective, cognitive, and behavioral. The relationship among these three components is changeable and not necessarily balanced (McGuire, 1985). Social norms and behavioral consequences (Fishbein & Ajzen, 1975), as well as the concrete or abstract nature of attitudes (Mickelson, 1990), mediate attitudes' influence on behavior. Hence, abstract attitudes toward general aspects of school and education tail to predict student academic achievement (Berk, Rose, & Stewart, 1970; Lunn, 1972; Ogbu, 1978).

Group membership and identity influence attitudes toward others. Research has shown that individuals tend to like others who are perceived as being sympathetic (Condon & Crano, 1988), similar (Newcomb, 1961), or consensual (Byrne & Nelson, 1965). Attitudes toward in-group members tend to be more positive and complex than those toward out-group members (Linville, 1982; Linville & Jones, 1980), who often are perceived stereo-typically (Deaux & Lewis, 1984).

The salience of similarities and differences with other people and the extent to which individual attributes are stereotypical contribute to define an individual's group affiliation. Prejudices, social …