Working School & college skill
Design an assessment rubric
Design an assessment rubric with outcome-aligned criteria, observable performance levels, fair weighting, accessibility, calibration, and feedback use.
Assess the intended learning, not familiarity with hidden expectations.
When to use
- Use for essays, projects, performances, labs, portfolios, presentations, or practice.
- Do not score disability, language background, style preference, or resources unless they are legitimate stated outcomes.
Preconditions
- Gather learning outcomes, task instructions, learner context, accommodations, stakes, samples, and policy.
Procedure
- Map each criterion to one stated outcome and remove criteria unrelated to learning.
- Choose analytic or holistic structure and a manageable number of consequential criteria.
- Describe observable evidence at each level without vague adjectives or deficit labels.
- Set weights that match teaching emphasis and test the arithmetic.
- Separate correctness, reasoning, communication, process, and conventions where appropriate.
- Review cultural, linguistic, disability, technology, and resource bias; support equivalent evidence routes.
- Calibrate several assessors on varied samples and resolve material disagreement.
- Give learners the rubric early and revise from use evidence.
Failure plan
- If assessors cannot agree, clarify evidence and recalibrate before consequential grading.
- If a criterion penalizes an access need, redesign the task or evidence route.
Worked example
A project rubric scores question, evidence, reasoning, and reflection separately, permits an accessible presentation format, and is calibrated on three anonymous samples.
--- name: design-an-assessment-rubric category: school description: Design an assessment rubric with outcome-aligned criteria, observable performance levels, fair weighting, accessibility, calibration, and feedback use. Use when work must be evaluated consistently and transparently. --- # design-an-assessment-rubric Assess the intended learning, not familiarity with hidden expectations. ## When to use - Use for essays, projects, performances, labs, portfolios, presentations, or practice. - Do not score disability, language background, style preference, or resources unless they are legitimate stated outcomes. ## Preconditions - Gather learning outcomes, task instructions, learner context, accommodations, stakes, samples, and policy. ## Procedure 1. Map each criterion to one stated outcome and remove criteria unrelated to learning. 2. Choose analytic or holistic structure and a manageable number of consequential criteria. 3. Describe observable evidence at each level without vague adjectives or deficit labels. 4. Set weights that match teaching emphasis and test the arithmetic. 5. Separate correctness, reasoning, communication, process, and conventions where appropriate. 6. Review cultural, linguistic, disability, technology, and resource bias; support equivalent evidence routes. 7. Calibrate several assessors on varied samples and resolve material disagreement. 8. Give learners the rubric early and revise from use evidence. ## Failure plan - If assessors cannot agree, clarify evidence and recalibrate before consequential grading. - If a criterion penalizes an access need, redesign the task or evidence route. ## Worked example A project rubric scores question, evidence, reasoning, and reflection separately, permits an accessible presentation format, and is calibrated on three anonymous samples. ## Done - An assessment-rubric document records outcomes, criteria, level descriptors, weights, accessibility, and feedback use - Outcome-map, arithmetic, bias, sample-calibration, learner-clarity, and accommodation checks verify fairness