Assessment shows you what students have learned and shows students where they stand. Well-designed assessments, clear criteria, and consistent grading help students focus on learning, and they give you reliable evidence of whether students have met your course outcomes. As generative AI becomes part of students’ lives, thoughtful and varied assessment design matters more than ever.
Three Purposes of Assessment
Assessment does more than produce a grade. Strong courses use assessment in three ways (Earl, 2003):
| Purpose | What it does | Examples |
|---|---|---|
| Assessment of learning | Measures what students have learned at the end of a unit or course | Final exams, capstone projects, and final presentations |
| Assessment for learning | Gives you and your students information to guide the next steps in teaching and learning | Practice quizzes, drafts with feedback, muddiest point, and exit tickets |
| Assessment as learning | Helps students monitor and reflect on their own learning | Self-assessment with a rubric, learning journals, and exam wrappers |
Research shows that assessment for learning, often called formative assessment, can raise student achievement (Black & Wiliam, 1998). Most courses rely heavily on assessment of learning. Building in more assessment for and as learning helps students improve before the grades that count the most.
Plan Assessments That Measure Learning
- Start with your outcomes. Every major assessment should measure at least one course learning outcome. See Course Design and Alignment.
- Match the assessment to the level of thinking. A quiz can check recall and understanding. A case study, project, or presentation can show analysis, evaluation, and creation. See Writing Learning Outcomes with Bloom’s Taxonomy.
- Use a variety of formats. Different formats show different kinds of learning, give students more than one way to succeed, and keep your course from relying on a single type of evidence.
- Make expectations clear. Explain the purpose, the steps, and the criteria for each assignment before students begin. See Evidence-Based Teaching Practices.
- Decide where AI fits. For each assignment, decide whether students may use generative AI, and how. Base your decision on what the assignment needs to show about students’ own learning. See Teaching in the Age of AI.
Vary Your Assessments
Move beyond a single format, such as all multiple-choice quizzes or all essays. A mix of assessments gives you a fuller picture of what students can do.
| Format | What students do | What it shows | Canvas tool |
|---|---|---|---|
| Case analysis | Apply course concepts to a real or realistic situation | Application, analysis, and judgment | Assignment or discussion |
| Video or audio explanation | Record themselves explaining a concept or walking through a solution | Understanding and reasoning in their own words | Canvas Studio |
| Oral exam or conference | Answer questions and explain their thinking in a short conversation | Depth of understanding and reasoning | Zoom or in person, with a rubric |
| Concept map | Show how key ideas connect | Organization of knowledge | Assignment |
| Two-stage exam | Take an exam alone, then again in a small group | Individual knowledge and collaborative reasoning | Quiz |
| Portfolio | Collect and reflect on work across the semester | Growth over time | Assignment or ePortfolio |
| Project with checkpoints | Complete a larger project in graded stages | Planning, process, and final product | Assignment with multiple due dates |
| Peer review | Give feedback on classmates’ work using criteria | Understanding of quality and the criteria | Peer review assignment |
| Exam wrapper | Reflect on how they prepared and what they will change | Self-awareness as a learner | Survey or quiz |
| Real-world product | Create a brief, plan, proposal, or design for an authentic audience | Transfer to real situations | Assignment |
For more on real-world tasks, see Designing Real-World Learning.
Design Assessments That Show Students’ Own Thinking
Many of the formats above are harder to hand off to generative AI because they ask students to show their process and reasoning. Strengthen any assessment by adding:
- Process work: proposals, outlines, drafts, and revisions submitted along the way
- In-class work: writing, problem solving, or discussion completed during class
- Oral explanation: a short conference or presentation where students explain their reasoning
- Specific context: tasks tied to your class discussions, local examples, or students’ own experiences
- Reflection: asking students to explain their choices, sources, and how they used any tools
For more ways to design assignments with AI in mind, see Teaching in the Age of AI.
Rubrics and Grading Criteria
Every graded assignment should have a rubric or clear grading criteria that students can see before they begin. A rubric is a set of criteria for student work that describes levels of performance quality (Brookhart, 2013). Rubrics make grading more consistent and faster, and they show students what quality work looks like.
Types of Rubrics
| Type | How it works | Best for |
|---|---|---|
| Analytic | Scores each criterion separately, with a description for each performance level | Major assignments where students need detailed feedback |
| Holistic | Gives one overall score based on a description of the work as a whole | Quick scoring of shorter assignments |
| Single-point | Describes only the proficient level for each criterion, with space to note where work falls short or goes beyond | Drafts, self-assessment, and focused feedback (Fluckiger, 2010) |
Build a Rubric
- List the criteria. Name the three to six most important qualities of the work, based on your learning outcomes.
- Choose the performance levels. Three or four levels, such as Exemplary, Proficient, Developing, and Beginning, are usually enough.
- Describe each level. Write what the work looks like at each level, using specific, observable language.
- Address AI use. If students may use AI, consider a criterion for how well they acknowledge it, explain their process, or evaluate what the tool produced.
- Test it. Score a few samples of student work to check that the descriptions are clear and the levels are distinct.
- Share it early. Give students the rubric with the assignment, and consider having them use it to assess a sample or their own draft.
AI tools can help you draft a first version of a rubric. Review and revise it carefully so it matches your outcomes, your assignment, and your expectations.
Rubrics in Canvas
Add a rubric to any Canvas assignment or graded discussion, then score student work directly in SpeedGrader. Students can see the rubric before they submit and see your scores and comments afterward.
You don’t have to start from scratch. The AAC&U VALUE rubrics are already available in Canvas. When you add a rubric to an assignment, select Find a Rubric to choose one. VALUE rubrics were developed by faculty across the country to assess skills such as critical thinking, written communication, quantitative literacy, ethical reasoning, and teamwork. You can use them as they are or adapt them for your assignment. Learn more about the VALUE rubrics from the American Association of Colleges and Universities.
Exams and Quizzes
Quizzes and exams work well for checking knowledge and understanding, especially as frequent, low-stakes practice. Pair them with other formats to assess higher-level thinking.
Write Strong Multiple-Choice Questions
Well-written questions measure what students know, not how well they take tests. Research-based guidelines include (Haladyna et al., 2002):
- Put the main idea in the question stem, so students could answer it before reading the options.
- Write the stem as a positive question. Avoid words like “not” and “except” when possible, or put them in bold.
- Make every wrong answer plausible, ideally based on common student mistakes.
- Keep answer choices similar in length and grammar.
- Avoid “all of the above,” and use “none of the above” sparingly.
- Use short scenarios or data to test application and analysis, not just recall.
AI tools can help you draft practice questions or plausible wrong answers. Check every question for accuracy, clarity, and alignment with your outcomes before you use it.
Try a Two-Stage Exam
In a two-stage exam, students first take the exam on their own, then take some or all of it again in small groups. In one controlled study, students retained more of what they discussed in their groups, and students at all achievement levels benefited (Gilley & Clarkston, 2014). Because the group stage happens in class, it also shows you how students reason through problems together.
Exams in Canvas
- Use question banks to give students different versions of a quiz.
- Add feedback to quiz questions so students learn from practice quizzes.
- Give students approved extra time or additional attempts through Canvas quiz settings. Accommodations are approved through Services for Students with Disabilities.
- Use Respondus LockDown Browser for any online quiz or exam where AI and other resources aren’t allowed. In-class testing and questions that ask students to explain their reasoning also help.
Grading Practices
- Grade against your criteria. Use a rubric or answer key so every student is held to the same standard.
- Grade without names when you can. Canvas lets you hide student names in SpeedGrader to reduce unintended bias.
- Calibrate with co-graders. If more than one person grades, score a few samples together first.
- Return work while it still matters. Feedback is most useful when students can apply it to the next assignment.
- Keep grades current in Canvas. Students should always know where they stand.
- Review any AI-assisted feedback. If you use AI to help draft comments, read and edit every one before students see it. You are responsible for every comment and grade.
Consider Specifications Grading
In specifications grading, students complete assignments that are graded as meeting or not meeting clear, detailed specifications, often with a chance to revise. Final grades are based on which bundles of work students complete at the required standard (Nilson, 2015). This approach keeps standards high, gives students clear targets, and can save grading time.
When You Have Concerns About AI Use
If you suspect a student used AI in a way your assignment did not allow:
- Compare the work with the student’s earlier work and the assignment’s process steps.
- Talk with the student about how they completed the work. Ask them to explain their reasoning or sources.
- Don’t rely on a detection score alone. See Turnitin for what AI scores can and can’t tell you.
- Follow The Citadel’s academic integrity process.
For more on preventing and responding to academic dishonesty, see the Academic Integrity Certificate.
Want help designing an assessment or rubric? Schedule a course design and teaching consultation.
Sources
Black, P., & Wiliam, D. (1998). Inside the black box: Raising standards through classroom assessment. Phi Delta Kappan, 80(2), 139–148.
Brookhart, S. M. (2013). How to create and use rubrics for formative assessment and grading. ASCD.
Earl, L. M. (2003). Assessment as learning: Using classroom assessment to maximize student learning. Corwin.
Fluckiger, J. (2010). Single point rubric: A tool for responsible student self-assessment. The Delta Kappa Gamma Bulletin, 76(4), 18–25.
Gilley, B. H., & Clarkston, B. (2014). Collaborative testing: Evidence of learning in a controlled in-class study of undergraduate students. Journal of College Science Teaching, 43(3), 83–91.
Haladyna, T. M., Downing, S. M., & Rodriguez, M. C. (2002). A review of multiple-choice item-writing guidelines for classroom assessment. Applied Measurement in Education, 15(3), 309–333. https://doi.org/10.1207/s15324818ame1503_5
Nilson, L. B. (2015). Specifications grading: Restoring rigor, motivating students, and saving faculty time. Stylus.
Have a question? Contact CEITL by phone, email, or in person in Bond Hall, Room 363.