As an experiment, more than 1,500 teachers were given a piece of writing from an imaginary 2nd grader. Both of the papers were identical—full of large print and youthful misspellings—the only difference was that one piece of writing made reference to a brother named “Dashawn,” and the other talked about his brother “Connor.” The name choices come from a list of racially distinctive names; Connor is more likely to refer to a white child, while Dashawn is more likely to be the name of a Black child.
About 35 percent of teachers judged the “Connor” paper as being at or above grade level, compared with nearly 31 percent of those who gave the “Dashawn” paper the same marks. David M. Quinn, an assistant professor of education at the University of Southern California, suggests that teachers were influenced by the subtle signals within the paper that one child was Black and the other white.
But when teachers were given a grading rubric—a set of performance criteria to use in evaluating the student writing—that disparity disappeared. About 37 percent of teachers said that the writing samples met the guidelines intended to measure how well the imaginary student was able to recount an event.
Giving teachers a set of specific criteria in order to judge a child’s work, rather than forcing them to rely on a possibly more vague sense of grade-level performance, could be a useful tool in reducing the impact of racial bias, said Quinn, in an article for the journal Education Next published Nov. 2. Quinn’s research was also published in the journal Educational Evaluation and Policy Analysis in June.
The differences in grading were driven primarily by white teachers and by female teachers; minority teachers and male teachers did not show the same grading disparities in this research, Quinn said. And, the disparities showed up regardless of a teacher’s implicit or explicit bias, as measured by another part of the evaluation. That could be because the sample size for the experiment was too small to tease out real differences among teachers, or because implicit bias testing may not be a valid measure of individual attitudes, Quinn wrote.
It’s important to note that this study couldn’t measure how generalizable these results would be, Quinn said. Teachers may feel differently towards students they know, rather than the imaginary students in this experiment. Also, the impact of rubrics may be different depending on the subject matter and the assignment being graded.
But the findings offer a “proof of concept” that grading guidelines may be helpful, Quinn said. “Rubrics will not be the only answer, but may have a role to play in broader-based racial equity efforts. Regardless, all teachers should be reflective and honest about potential implicit or explicit stereotypes they hold, and should actively work to overcome them.”
Images of writing by David M. Quinn
Ruler image by Getty