In this dissertation, I explore the relations between student race/ethnicity, nativity, gender, and exceptionality, and I examine how school context shapes those relations. I argue that exceptionalities are socially constructed, and that different categories of exceptionality have different meanings and statuses.

In the first empirical chapter, I use a factorial vignette survey to test for race/ethnicity, nativity, and gender effects in how teachers respond to children indicating academic and behavioral differences in the classroom. I found that boys of color were more likely to be suspected of exceptionality when they exhibit behavioral challenges. I argue that the category of emotional disorder, in particular, may maintain or exacerbate racial/ethnic inequalities in education.

In the second empirical chapter, I use a dataset of Wisconsin public school students to test whether and how school racial/ethnic composition moderates the relationship between individual-race/ethnicity and placement in special education across the more subjective categories of disability. I find that white students who attend schools with more peers of color have increased risk of identification with more-advantaged disabilities, while students of color that attend schools with more same-race peers have decreased risk of placement in special education. These findings support explanations of frog-pond effects driven by racial/ethnic bias and also the mechanism of racial/ethnic competition.

In the final empirical chapter, I use school-level racial composition data and an experimental survey design to examine whether a student's race/ethnicity, nativity, and gender affect teacher decisions to refer for exceptionality testing differently in schools with different racial/ethnic compositions. I find that in schools with more white students, teachers are less likely to perceive academic challenges as disability when evaluating a boy of color. While white boys experience decreased likelihood of referral for behavior problems in schools with more white students, boys of color experience no such decrease. Conversely, when they work in schools with more white students, teachers are more likely to refer girls of color when they have academic challenges, and less likely to refer when they have behavioral challenges. I argue that these findings suggest some support for contextual effects explained by racial-bias-driven frog- pond effects.