Best practices for academic grading

No single approach to grading will work for all instructors and educational contexts. There are, however, several principles that can help promote transparency, equity, and a greater focus on student learning in any academic environment:

• Intentional design: in designing graded assessments it is important to be mindful of which specific dimensions of student work we want or need to assess. For example, timeliness of submissions or the level of formal polish might be central to the learning goals of the course–but they might also be entirely extraneous. In the latter case, taking off points for late work or surface errors might result in a grade that does not accurately reflect mastery of the course learning goals themselves. Designing any graded assessment, from multiple choice tests to open-ended creative portfolios, should begin with clearly articulating, for ourselves and for our students, what exactly will be assessed, how, and why. Clarifying goals and criteria also helps with making assessments more fair and consistent, as the object of the assessment.
• Transparency: It is essential to clearly articulate the grading schema for the course on the syllabus, as well as provide a mechanism for the ongoing monitoring of progress, so that students understand how they are doing in the class at any point in the term. For example, gradebooks embedded in learning management systems (e.g. Blackboard) can be set up ahead of time to calculate grade percentages as the term progresses. Better yet, robust feedback loops augmented by student self-reflection provide an organic, learning-centered structure for monitoring progress throughout the term. Finally, given the ongoing revolution in the field of generative AI, it is important to provide clear guidelines on allowed and disallowed uses of AI tools when completing course assessment tasks.
• Rubrics: grading rubrics can help instructors clarify their assessment criteria and help learners understand what exactly is being assessed and how. Rubrics are also helpful in articulating assessment priorities. For example, a final project rubric might list not only the desired qualities of the final product, but also process-oriented components like thoughtful revision, communication skills, or reflection. Finally, rubrics can help students in the process of self-assessment, especially when co-constructed in collaboration with the learners themselves.
• Growth mindset: like any aspect of course design, assessments can help promote (or undermine) learning-centered, process-oriented, student-friendly environments. For example, opening the course with a series of easier assessments before building up to more challenging ones can help reassure students experiencing belonging uncertainty and shore up the motivation required for succeeding in the class. Building in options for penalty-free assignment revisions and/or test re-takes allows students a chance to make-up for early mishaps, not to mention the invaluable opportunity to learn from past errors to improve future learning. Other methods of rewarding progress might include adding points for meaningful revision into grading rubrics or weighing end-of-term work heavier than earlier work in course grading schemas. Far from being purely extrinsic evaluatory mechanisms, our grading systems can play a key role in fostering intrinsic motivation and a growth-mindset approach to learning.
• Choice: offering students choices in how they demonstrate mastery of course learning goals can boost motivation and create opportunities for authentic engagement with course material. Consider thinking outside the box to develop (perhaps together with your students?) interesting and innovative mechanisms for assessing learning in your course.
• Inclusion: rather than deploying the same generic tests or assignments in every class, consider tailoring aspects of your graded assessments (topics, examples, case studies, names, places, etc.) to the particular interests and identities of your students. Project-based learning (PBL) lends itself particularly well to this type of tailored assessment, but small customizations can be introduced into any assessment sequence once instructors get a good sense of who is in their class.

Check out these related teaching tips on inclusive grading and grading anxiety [Requires sharepoint login]

Specifications grading

One popular alternative to traditional grading is specifications grading. The core of this approach is articulating detailed criteria (specifications) for what constitutes a successful attempt for each assessment task. The specifications can be general (“discussion post cites two class readings” or “80% of the problems have been answered correctly”) or very detailed (“the lab report meets all ten criteria listed on rubric”). The point is that, rather than assigning a grade or percentage value on a scale, the instructor simply deems the attempt acceptable or not acceptable according to the specifications listed. Instead of asking, “is this a B- or a C+ reflection paper?” the grader asks, “Did this reflection paper meet the specifications or not?” This binary approach mirrors more accurately professional practices (paper accepted or not accepted for publication; project funded or not funded). It also forces instructors to articulate very clearly what criteria (specifications) are required for successful completion of each task. Since the bar is typically set at around a B level performance, the specifications approach also has the advantage of ensuring a high level of rigor for every student. Finally, the pass/fail method motivates students to revise submissions that did not meet specifications criteria in order to earn credit for their work.

The final grade in a specifications grading system is calculated based on the total number of successful attempts, tallied at the end of the term or unit. In a classic specs grading model, assessment tasks are bundled together and organized into tracks that allow the student to choose which final grade they want to earn. For example, for a final grade of C, the student might need to successfully complete ten problem sets, four quizzes, and two reflections. For a grade of B, they might need all of the elements required for a C, plus an additional four problem sets, two quizzes, and two reflections. For a grade of A, they might need all the elements needed for a B plus a final paper and presentation, and so on. Crafting a specification grading schema requires a lot of intentional design: what should be the required minimum for passing the class? How can assignments be distributed in a way that all students, regardless of their selected grade track work towards the same course learning goals? How many re-takes are allowed? And so on. The reward is a highly transparent system that allows for a high degree of student agency while freeing the instructor from micromanaging grade assignments throughout the term.

Ungrading

The ungrading movement has emerged in recent years in response to research indicating that traditional grades may have a detrimental effect on student motivation, engagement, curiosity, willingness to take intellectual risks, mental wellbeing, and community relationships. Proponents of ungrading point out that ranking students by means of grades is a relatively recent invention, one that flourished under the industrial revolution’s factory regime of scientific management, and that has no place in a modern system of education grounded in the science of human learning. While many ungraders lobby passionately for the abolishing of grades altogether, others work within systems where grades are still assigned, but the emphasis is shifted from ranking to feedback. Instructors interested in dipping their toes into the world of ungrading, might consider trying out the following strategies:

• Grade-free zones: consider including ungraded learning activities and assignments to serve as a sandbox for students to practice and learn without the pressures of traditional assessment.
• Qualitative assessment: where possible, replace grades with qualitative descriptive feedback recording the student’s progress and remaining gaps. Ideally, students would engage with qualitative feedback in the form of reflective responses in which they reflect on the feedback and explain how they will apply/have applied it in subsequent practice.
• Pass/no pass options: shifting from letter or percentage grades to binary pass/no pass systems emphasizes attainment of assignment goals over ranking. This approach is the cornerstone of specifications grading (see above) but can be incorporated into any grading schema.
• Focus on process: foregrounding qualitative feedback is part of a broader approach that emphasizes the process of learning, rather than attending solely to the final product. Portfolio assignments, student-led individual learning plans, and reflection logs are all ways to foreground process and foster self-sponsored learning.
• Self-assessment:the guiding principle of self-assessment is that students are themselves the best experts on their own learning. While novices might require guidance when it comes to understanding their level of mastery, students know best how much effort they have expended on their work and how much progress they have made relative to their own prior level. Soliciting reflections and self-assessments from students (from brief ratings of confidence in the correctness of test answers, to detailed process letters and end-of-term assessment conferences) is an essential element of many ungrading systems.

As with any departure from traditional practice, instructors introducing elements of ungrading need to budget time for communicating their assessment system to students and explaining the rationale behind what is likely to be an unfamiliar (and therefore potentially uncomfortable) approach. Successful ungrading requires substantial student buy-in, as the transition from “jumping through hoops to earn a grade” to “learning for learning’s sake” cannot take place without the conscious and willing participation of the learner. The ungrading approach therefore requires enhanced attention to motivating students and to clearly articulating the value of learning tasks, so that students undertake them without the “carrot and stick” incentive system provided by traditional grades.

Next: Additional resources and references