Considering Equity in Grading in Computer Science


Dr. Kristin Stephens-Martinez teaches Computer Science 216, Everything Data; this large course introduces over 200 students to working with data – both fundamental concepts and computational tools – and finishes with a capstone team project working with real data. She says, “I’ve been on a journey for a while in trying to figure out how to make my grading more equitable.” In the most recent offering of CS 216, Spring of 2022, Dr. Stephens-Martinez made some dramatic changes in the grading scheme and introduced mini-exams. These changes were motivated in part by multiple conversations on her podcast, CS-Ed Podcast, in which she interviews various guests about teaching and equity in computer science. In particular, she was inspired by her conversation with Joe Feldman, the author of Grading for Equity, and later interviews with Brett Wortzman and Kevin Lin.

Dr. Stephens-Martinez focused on building flexibility and compassion into her course assessments. For example, there were four mini-exams, and students could retake each mini-exam. In these mini-exams, instead of allocating specific points to parts of questions, each question was graded with a four-category rubric, with each category having a specific number of points allocated. Only blank responses received a 0. Students did initially find the 4 category rubric a bit confusing, as grading didn’t point out exactly where they could have received a higher grade. They wanted more feedback, but by mid-semester, most students strongly agreed that grading was reasonable.

E (Exemplary) Work that meets all requirements and displays full mastery of all learning goals and material.
S (Satisfactory)Work that meets all requirements and displays at least partial mastery of all learning goals as well as full mastery of core learning goals.
N (Not yet)Work that does not meet some requirements and/or displays developing or incomplete mastery of at least some learning goals and material.
U (Unassessable)Work that is missing, does not demonstrate meaningful effort, or does not provide enough evidence to determine a level of mastery. 
Scoring rubric for sections of mini-exams

TAs were trained to grade specific questions on this 4 point scale by looking at examples graded by Dr. Stephens-Martinez.

The mini-exam and other practices Dr. Stephens-Martinez used in the course are examples of choices that support all students, described as 3 pillars of Grading for Equity: accurate, bias-resistant and motivational.

  • The mini-exam is not graded on points (0-100) but each question is graded on a 0-4 scale, with a minimum of 1 point if the student wrote anything relevant. This is an example of accurate and motivational grading as described in Grading for Equity. In the mini-exam logistics, Dr. Stephens-Martinez provided a translation from the rubric to a percentage score to help students understand their grade. She cautions that a 4 point scale does not allow partial credit, and thinks that allowing retakes helped students accept the 4 point scheme.
  • For many assignments other than the mini-exams, if a student scored above a certain percentage (80-90%, depending on the assignment), they got full credit, which is an example of grading based on standards rather than points, which is motivational.
  • Students had the option to retake any of the mini-exams without making special arrangements, and could decide which mini-exam(s) to retake, which is an example of bias-resistant grading – grades are based on evidence, not reflective of a student’s environment or the timing of the work.
  • Students could substitute a related grade for a missed mini-exam, another example of motivational grading.
  • There was a relatively liberal (and well-documented in her syllabus) late policy for most assignments, also an example of both bias-resistant grading as well as being motivational for students.
  • There were many chances to practice and get feedback before a final grade is assigned, a practice supported by all three pillars of Grading for Equity.

Dr. Stephens-Martinez found that educating students on why and how she was using this grading scheme was very important. As the grading scheme for the entire course was complex, and scores higher than a certain percentage got full credit, she provided a Google sheet with embedded formulas to calculate grades, and gave directions for students to make their own copy to track their own grades, an example of using grading to be motivational. Further, students were given many resources for help, including daily office hours with TAs both online (Zoom) and in person, and the online class discussion forum available as a lecture back channel to encourage use of the discussion forum for creating community.

For more on grading, read Grading for Equity by Joe Feldman. Duke community members can access an online version of the book via the Duke University Libraries.

CS-Ed Podcast Hosted by Dr. Kristin Stephens-Martinez at Duke University

More from Dr Stephens-Martinez

Read an explanation of the logistics for mini-exams with policies spelled out and the syllabus with other examples of flexibility.  

Listen to, or read transcripts of her podcast. She was particularly motivated by this episode “Grading for Equity with Joe Feldman” and her rubric design was heavily inspired from her discussion in a two part episode series: “Alternative Grading with Brett Wortzman and Kevin Lin Part 1” and “Alternative Grading with Brett Wortzman and Kevin Lin Part 2.”

More on Grading from our blog:

Symposium Spotlights: Seven (Feasible) Ways to Beat the Grading Grind, Amy Kenyon, 06/16/2022

Alternative Strategies for Assessment and Grading, Amy Kenyon, 03/09/2022