(From Education Week)
It’s time to make grading bias-resistant and fairer
By Joe Feldman
January 23, 2019
Grades are used to make a number of significant decisions about students, including promotion, retention, college admissions, and scholarships. But parents, teachers, and school administrators concerned about providing equitable opportunities for every child are frequently stunned to learn that many common grading practices are outdated, inaccurate, and harmful to student success. In fact, grading policies—which appear to be an objective, fair, and accurate method to describe a student's academic performance—often increase achievement gaps by infusing grades with teachers' implicit biases or by rewarding or punishing students based on their families' resources.
Because grading is often not addressed in teacher preparation or in ongoing professional development, most teachers choose their own way to grade, guided by their best sense but uninformed by either research or best practices.
As a result, traditional grading practices result in grades that provide unclear and often misleading information to parents, students, and postsecondary institutions. Teachers often combine a range of unrelated student information into a grade, compressing a bucket of information into a thimble-sized container. When teachers collapse academic proficiency, soft skills, behavior, attendance, and effort into a single letter, it is impossible for anyone to discern the student's particular strengths and weaknesses in each of these aspects, rendering the grade vague, confusing, and even invalid.
Traditional grading practices are often also corrupted by implicit racial, class, and gender biases. It is well documented that schools' disciplinary actions often disproportionately punish African-American, Latino, low-income, and special education students. These same biases affect grading. When teachers grade students on subjectively interpreted behaviors in categories of participation or effort, their perceptions and judgments of those behaviors are influenced by that teacher's race, class, and gender. For example, in classrooms taught by white teachers, African-American students are typically rated as poorer "classroom citizens" than their white peers, and thereby are more likely to have a lower grade for those behaviors because of the teacher's biased perceptions.
"Traditional grading practices are often also corrupted by implicit racial, class, and gender biases."
Grades based on performance outside the classroom can also reproduce cycles of disparities, by rewarding or punishing students based on their income and resources. Students are more likely to complete homework if they have more resources (such as higher educated and English-speaking parents, internet access, and a safe, quiet space at home). Yet teachers often award points to students for their homework performance, which rewards or punishes students based on environmental factors outside of the student's control.
Finally, most teachers use grading practices based on calculations that depress student achievement and progress. Current practices conflict with contemporary understandings of growth mindset and how to encourage students to learn through practice and experimentation. An F early in a student's learning and an A at the end average out to a C, regardless of progress over time and final achievement. This mathematically unsound approach punishes students with early struggles, often those who enter the class with fewer resources and less prior academic success and have the most potential growth ahead of them.
Many teachers entered the profession to provide every student an opportunity to succeed, to minimize the achievement and opportunity gaps. Yet when we continue to use traditional grading practices, we inadvertently reproduce those gaps.
In partnering with urban, rural, and suburban public schools and districts, as well as elite private academies, we at the Equitable Grading Project—a research initiative of the equity-focused education consulting group I lead—have found that once teachers learn about the harms caused by our century-old grading practices, they are motivated to try alternative, more equitable grading practices.
In our review of academic research and observations of teachers' implementation, we have identified key aspects of what we call equitable grading, which is accurate, bias-resistant, and motivational:
Reflect growth and learning. Teachers should use a more proportionately structured 0-4 scale instead of a 0-100 point scale that is mathematically oriented toward failure, stop assigning a zero for missing work, and weight recent performance and growth instead of averaging performance over time. By allowing students to retake tests and projects (with the chance to replace previous scores), teachers can reward learning, support a growth mindset, and not disadvantage students who enter classrooms with weaker academic backgrounds.
Value knowledge, not environment or behavior. Instead of grading subjectively interpreted behaviors such as a student's "effort" or "participation," teachers should focus grades on required content or standards. Grades should not be used to reward compliance or homework completion, both of which invite implicit and institutional biases.
Lift the veil on how to succeed. Standards-aligned rubrics, simplified grade calculations, and standards-based scales and gradebooks make teacher's expectations explicit and facilitate students' understanding, ownership, and power over their grades.
Build soft skills—without including them in the grade. Teachers should employ a more expansive range of feedback strategies to incorporate peer and self-evaluation and build self-regulation instead of relying solely on grades as feedback.
These grading practices have a proven record of success. Last year, the Equitable Grading Project partnered with two independent evaluators to study what happens when teachers adopt these practices. In an analysis of 60 teachers in seven high schools in northern California, we found that the rate of students receiving D's and F's frequently decreases with these practices—and decreases more dramatically for vulnerable and underserved student populations. For example, among the teachers who worked with the Equitable Grading Project to implement equitable grading, the rate at which students eligible for free or reduced-price lunch received D's and F's decreased a statistically significant greater amount than students who were not eligible for free or reduced-price lunch.
Our analysis also found that equitable grading practices don't just reduce D and F rates, they are also more accurate. When teachers use equitable grading practices, they assign final grades that are more strongly correlated with students' standardized assessment scores on that content, and the effect is stronger for students who qualify for free and reduced-price lunch.
Finally, we found that equitable grading improves the learning environment for teachers. In surveys and independent case studies, teachers and students reported that classrooms are less stressful, teacher-student relationships are stronger and more trusting, and students are more motivated to learn after these interventions.
When we talk about equity, grading is rarely mentioned. By overlooking it, we unwittingly perpetuate the very disparities we dedicate ourselves as educators to correct. It is incumbent upon educators at every level—teachers, principals, district administrators, school boards, and state policymakers—to improve grading policies to ensure that they reinforce, not work against, our commitment to equity.