Inclusive & equitable assessment — a crash course

A lot of what an assessment measures is incidental — confusing wording, an inaccessible format, an unstated assumption. Whenever a barrier is irrelevant to what you’re actually assessing, it’s just noise, and it falls hardest on students who already face the most friction. Designing it out helps everyone.

• Write instructions in plain language, and state explicitly what a good answer needs to do.
• Provide accessible formats by default — readable structure, real headings, good contrast, and content that works with assistive technology.
• Use the Universal Design for Learning idea of multiple means of expression: where the format isn’t the point, let students demonstrate the same learning in more than one way.
• Check that the task tests the intended outcome, not unrelated reading speed, cultural in-knowledge, or tech access.

Flexibility is often confused with leniency. It isn’t. You can offer choice in how students evidence their learning while holding the learning outcomes and standards completely constant — the criteria stay the same, only the route changes.

• Anchor every option to the same outcomes and the same marking criteria, so no choice is an easier path to the grade.
• Offer choice where the mode is incidental (for example, a presentation or a written piece) rather than where it is the skill being assessed.
• Build in flexibility by design — deadlines, formats, or question choice — so fewer students need to ask for it as an exception.
• Keep options manageable: a few well-designed routes beat an unmaintainable menu.

Some students will still need individual adjustments, and providing them is both a professional and a legal expectation. But every barrier you remove up front is one fewer adjustment anyone has to request — good design quietly shrinks the queue and reduces who has to disclose to be treated fairly.

• Know your institution’s adjustment process and apply it promptly, without making students justify themselves repeatedly.
• Treat agreed adjustments as a floor, not a favour — they protect a student’s right to a fair assessment.
• Where an adjustment keeps recurring across a cohort, fix the task instead so it stops being needed.
• Protect privacy: adjustments and disclosures are confidential and shouldn’t single anyone out.

Even a well-designed task can be marked unevenly. Marking is a human judgement, and human judgement drifts — across a long pile, between markers, and under the pull of unconscious bias. A bit of structure keeps grades tied to the work, not to who produced it.

• Mark against a clear, shared rubric so judgements stay anchored to the criteria.
• Mark anonymously wherever the assessment format allows it.
• Use moderation or second marking to catch drift and keep standards consistent across markers.
• Watch for bias triggers — name, language background, handwriting, early-pile fatigue — and re-check borderline decisions.