Race, representation, and writing assessment: Racial stereotypes and the construction of identity in writing assessments

Recent research in social psychology has suggested that negative racial stereotypes can have a negative effect on student academic achievement. Although such research has provided experimental evidence on the effect of stereotypes in a variety of different testing contexts, there has been limited research on the effect of racial stereotypes in writing assessments. The purpose of this study was to understand how stereotypes about racially-ordered performance shape students\u27 expectations about the context and consequences of a Placement Exam. The theoretical orientation for this project drew upon Claude Steele\u27s research on stereotype threat, Norman Fairclough\u27s approach to Critical Discourses Analysis, and rhetorical analysis. First-year college students taking a Writing Placement Exam were given a survey that included questions about racial stereotypes and audience expectations. Of the 38% of students who completed the survey (n=1,195), 38 students were finally selected to be interviewed. The results of this research suggest that racial stereotypes do have an impact on students\u27 expectations of writing assessment contexts. Students of color, in particular, expected that their writing performance would be read along racial lines and that they might be graded according to racial expectations for performance. The results also showed that approximately 70% of students in every racial group expected readers to be white. In interviews students said that they based their expectations on racial demographics, stereotypes about college professors, and past educational experiences in English classes. An analysis of student essays was then conducted to determine if students\u27 expectations about stereotypes and readership were evidenced in their writing. A discourse analysis revealed that the primary identity adopted in student essays on the exam prompt patriotism was a moderate, unraced American identity, i.e., a cohesive us. A rhetorical analysis of three essays, written by students who strongly felt that their essays would be graded differently if the reader knew their race, revealed that students used various rhetorical moves to anticipate a potentially hostile audience. The results of this research suggest that racial stereotypes can have an effect in writing assessments and how students rhetorically approach writing assessments. The findings suggest new considerations in test design, test use, classroom interventions, and institutional change