Rubrics & marking — a crash course

Rubrics come in two broad shapes. An analytic rubric scores several criteria separately; a holistic rubric gives one overall judgement against level descriptors. Each fits a different job.

• Use an analytic rubric when students need detailed feedback on distinct dimensions, like argument, evidence and structure.
• Use a holistic rubric for fast, overall judgements where a single integrated quality matters more than the parts.
• Match the rubric to the task and the purpose — don't default to the most detailed grid for everything.

A rubric only makes expectations visible if students can read it. Vague or jargon-heavy descriptors leave them guessing — and leave you re-explaining at marking time.

• Describe what quality looks like, not just labels like ‘excellent’ or ‘poor’ — make each standard observable.
• Share the rubric before students start, so it guides the work rather than just explaining the grade.
• Use exemplars and annotated samples so students can see what each standard means in practice.

A rubric improves consistency, but it doesn't guarantee it on its own. When several people mark, or one person marks a large pile, judgements drift. Calibration and moderation keep them aligned.

• Calibrate before marking — mark a few shared scripts together and talk through where the rubric is read differently.
• Moderate after marking — sample and cross-check borderline and high-stakes cases to confirm standards held.
• Note recurring ambiguities and tighten the rubric wording for next time.

The rubric you built to set expectations is also what makes marking faster and feedback clearer. Let it carry the routine load so your comments can do the rest.

• Tie each comment to a criterion so feedback points to where the work sits and what to improve.
• Let the rubric handle repeated judgements, freeing your written comments for the specific and the personal.
• Return the completed rubric with the work so students can see how the standards applied to them.