Behavioral and academic outcomes differ for students by race, ability, and gender within the K-12 public education system. Moreover, striking gaps exist at the intersection of race, ability, and gender, despite the similarity in severity and frequency of behavior between groups. Few studies, however, have examined the educational mechanisms that contribute to these gaps. Despite this, the scientific literature? shows that when educators have high expectations, students are more likely to be successful academically and behaviorally. Therefore, this study examines the inverse of this relationship by recognizing that biases likely influence behavior and academic student outcomes through expectancy bias for certain groups of students. The present study utilizes an intersectional framework of disability studies and critical race theory (DisCrit) to examine preservice educator expectations of behavior and academic outcomes of a hypothetical student at the intersection of student race, ability, and gender using a factorial vignette experimental design. Analyses consisted of factorial multivariate analyses of main and interaction effects including covariates for social desirability, tolerance, severity, and demographic characteristics. Results indicated significant and meaningful differences in expectations of behavior and academic experiences by race and ability. However, interaction effects were not detected. Implications and limitations of this study are discussed.

Socially salient information (such as stereotypes and expectancies) can be transmitted amongst individuals in a variety of subtle ways. One of these is the Linguistic Expectancy Bias (LEB), in which patterns of linguistic abstraction indirectly indicate a speaker's attitudes toward a target. The LEB is a common feature of human communication, but research on it has largely been limited to the laboratory; its presence in news media reports is not well-studied. In three studies, I investigate the operation of the LEB in the print media domain. In the first, published reports of NFL games between intercity rivals were analyzed to determine whether or not hometown teams receive more favorable linguistic treatment than hated rivals; results indicate no evidence of a systematic LEB effect. In the second, news reports about the 2004 Presidential election were examined for differential coverage based on the party membership of the candidates, with no evidence of linguistic bias discovered. In the third, participants were exposed to a description of a politician that varies in the levels of abstraction used to describe his actions and asked to form impressions of him. Linguistic bias was found to have a subtly paradoxical effect, such that bias against a candidate resulted in greater explicit and implicit liking for him. Implications for both the social psychology and political science literatures are discussed. / Psychology