The start of term is a good time to make your assessment expectations explicit. Before students begin their first assignment, take a fresh look at the rubric they will use to guide their work.
A useful university assignment rubric connects learning outcomes to evidence in student work, distinguishes levels of performance, and helps students understand how to improve. These principles apply whether you grade independently, work with TAs, or use AI-assisted grading.
Choose one upcoming assignment and work through these five checks.
1. Connect each criterion to evidence
For each criterion, ask: what would a student actually write, explain, or demonstrate? If you assess critical thinking, specify the reasoning you want to see, such as evaluating an alternative explanation or justifying a decision with evidence.
2. Make adjacent performance levels distinct
Read two neighboring levels without their labels. Could you tell which represents stronger work? Explain what changes in the quality of the reasoning, application, or evidence. Words such as “excellent” and “good” need supporting descriptions.
3. Check for overlapping penalties
Review criteria such as analysis, argument, and use of evidence. If one weakness could lower all three ratings, clarify what each criterion measures and when separate deductions are justified.
4. Leave room for different valid approaches
Check whether students can meet the learning outcome through a different defensible answer, method, or example. Describe the qualities of successful work while allowing room for choices appropriate to the assignment.
5. Try the rubric before students rely on it
Apply it to a short sample response. If you work with TAs, compare ratings and the evidence behind them. If you use AI, review its suggestions against your own judgment. Clarify ambiguous wording before sharing the final rubric with students.
A rubric example you can adapt
Imagine a business case assignment asking students to recommend a course of action. A criterion labeled “Critical thinking,” with ratings of “Excellent,” “Good,” and “Needs improvement,” leaves much of your judgment unexplained.
One way to make it clearer is to assess the justification of the recommendation:
Adapt the language and levels to your course. This is one illustrative criterion, not a complete rubric.
Help students act on feedback
Once students have a draft, ask them to use the rubric to identify one improvement before requesting feedback. After receiving feedback, have them add a brief revision note:
This gives you something concrete to discuss: how students interpreted feedback and made decisions. It also makes your evaluative thinking visible while keeping students responsible for their revisions.
Putting this into practice with TimelyGrader
For courses using TimelyGrader’s Formative Feedback feature, students can receive feedback on drafts against the assignment rubric without seeing grades or suggested scores. They can also ask Hugo, TimelyGrader’s AI assistant, for clarification before revising.
See the student feedback walkthrough
Try the five checks on one assignment this week, and share them with a colleague preparing their first assessment.
Further reading: Cornell’s guide to using rubrics.
