The path to becoming anti-racist is long, challenging, and for teachers, essential. Tackling one's mindset and implicit biases takes hard work, and even trainings intended to help people do this work may not lead to changes in behavior.
A new study from David Quinn (University of Southern California) explores whether there's a way to change how teachers' biases affect students, even without changing their underlying beliefs. He asks, do teachers show racial bias when grading student work, and can more specific grading criteria mitigate this bias?
The answer to both questions: Yes. This is a promising development given that changing individuals' biases takes time and students are currently being subjected to teachers' biases (even if unintended).
To dig into this, Quinn recruited teachers for an online survey, asking them to evaluate whether a writing sample was on grade level or not (with no further definition of what "on grade level" looks like). The teachers were randomly given a writing sample that referred to the student's brother and friend using either common Black male names or common white male names, thus implying the race of the writer. The samples were otherwise identical.
In this experiment, teachers were nearly 5 percentage points less likely to rate the Black student's version as "on grade level" compared with the white student's version. When examining findings by subgroup, these results were driven by female teachers (7 points less likely to rate the Black student's writing on grade level) and by white teachers (8 points less likely). There was no apparent bias in the grading by Black and Latinx teachers, and surprisingly, no apparent bias in grading from male teachers.
In contrast, when teachers were given a structured rubric that offered clearly defined performance criteria for evaluating the writing sample, teachers' ratings were virtually identical for both the Black and the white student work samples.
The biggest surprise in the study was that teachers' bias in grades were unrelated to whether they had implicit or explicit biases. This gives further credence to the theory that anti-racist trainings may not be enough to change people's behavior.
Creating more defined grading rubrics certainly isn't enough to remove bias from the classroom, which must remain the ultimate goal, but it could be an important step. Teacher preparation programs and school district leaders can support this effort by teaching how to develop and use grading rubrics. And compared with some of the alternative approaches to reduce bias in grading (like reducing the workload and distractions teachers face while grading student work!), establishing structured grading rubrics may be far more feasible.