To set up your rubric in JeddAI, enter or paste your own marking criteria, give each criterion a performance level and a clear descriptor, and allocate marks, so the tool drafts marking against your standards rather than a generic model. You review and adjust every result afterwards.
What does a marking rubric need to contain?
A rubric needs three things to mark well: a set of clear criteria, a performance level for each criterion, and a descriptor that says what work at that level looks like. Together these tell the tool what to judge and how to tell a strong response from a developing one, which is exactly what you weigh up when you mark by hand.
Add a mark or weighting to each criterion as well, so JeddAI knows how much each part contributes to the overall result. You do not need an elaborate document. A concise rubric with a handful of well-defined criteria usually produces more consistent marking than a long one with overlapping or vague categories.
- Criteria: the specific things you are assessing, such as thesis, use of evidence, or structure.
- Performance levels or bands: the tiers a response can fall into, from limited to sophisticated.
- Descriptors: a short statement of what work looks like at each level.
- Mark allocations: how many marks or what weighting each criterion carries.
How do you add your existing rubric to JeddAI?
Add your rubric by entering or pasting your own criteria rather than adopting a generic template, because JeddAI marks against what you provide. If your faculty already has an agreed rubric aligned to curriculum outcomes, use that wording directly, so the drafted marking reflects the standards your students are actually assessed against.
Keep the language close to how you would explain the criteria to a colleague or a student. Where a criterion refers to a specific outcome, such as an ACARA achievement standard or a syllabus requirement, name it plainly. The clearer the source of each criterion, the more closely the first draft will track how you mark, and the less you will edit afterwards.
If you mark several classes, or a faculty shares a common task, agree the rubric once and enter that shared version, so every class is judged against the same criteria. This is where a well-defined rubric pays off beyond time saved: it makes marking across a cohort more consistent, and it gives you a clear reference when a student or parent asks how a mark was reached.
How do you write descriptors JeddAI can mark consistently?
Write descriptors that name observable features of the work, because concrete language is what lets JeddAI apply a criterion the same way across a whole class. Integrates textual evidence to support a sustained argument gives the tool something to look for; good analysis does not. Keep one idea per criterion, with clearly different descriptors at each level, so the marking does not drift.
Avoid overlap between criteria and between levels. If two bands could both describe the same paragraph, the distinction is too soft for consistent judgement, whether a person or the tool is applying it. Sharpening those boundaries is the single most useful thing you can do before your first run, and it improves your rubric for hand-marking too.
It helps to test a draft rubric against a couple of responses you have already marked before you run a full class. If the descriptors capture why one script sits above another, they are specific enough. If you find yourself relying on a gut sense the wording does not explain, that is a signal to add the missing detail.
How do you set mark allocations and performance bands?
Set mark allocations by giving each criterion a weighting that matches its importance in the task, then define the bands a response can reach. If your assessment uses a marks-out-of scale, an A-to-E judgement, or NESA-style performance bands, mirror that structure so JeddAI’s draft lines up with how you report results.
Anchor the top and bottom bands first, since those are the easiest to describe, then fill in the middle. Aligning bands to published standards, such as Australian Curriculum achievement standards or your senior syllabus, keeps the marking defensible and consistent with what students meet elsewhere. You can adjust weightings after seeing how the first drafts fall.
Keep the number of bands manageable. Four or five well-separated levels are easier to apply consistently than eight finely graded ones, and they map cleanly onto most reporting scales. If your task carries a total mark, make sure the criterion marks add up to it, so the drafted total is transparent rather than something you have to reconcile by hand.
How do rubrics, success criteria and comment banks fit together?
A rubric, success criteria and comment banks do different jobs, and using all three gives JeddAI the fullest picture. The rubric defines the criteria and standards it marks against; success criteria describe what a strong response looks like in student-friendly terms; and comment banks hold the feedback phrases you reuse so students hear a consistent voice.
You do not need every element on day one, but they reinforce each other. The rubric tells the tool how to judge, success criteria help it explain the judgement, and the comment bank shapes the wording of the feedback. Together they keep both the marks and the comments aligned to your expectations across the whole class.
A practical order is to build the rubric first, since it does the heaviest lifting, then add success criteria and a small comment bank as you notice the phrases you write again and again. There is no need to complete all three before you start; each one you add simply sharpens how closely the drafts match your voice.
How do you refine your rubric after the first marking run?
Refine your rubric by looking at where you edited JeddAI’s drafts most, because those edits show which criteria or descriptors were unclear. If you repeatedly nudged the same mark or reworded the same comment, tighten that part of the rubric, and the next run will need less correction. Consistency comes from the criteria, not the automation.
Once a rubric works, reuse it across similar tasks so feedback stays consistent over time and you are not rebuilding it each term. A tool like JeddAI helps you apply the same criteria evenly across a class and cut the time spent drafting routine feedback, while you review and approve every mark and comment. Built in Australia, it keeps the teacher in control. You can Get started with JeddAI and test your rubric on a single assignment.
- Note the criteria you edited most, and clarify their descriptors.
- Split any criterion that tried to assess two things at once.
- Reuse a working rubric on similar tasks instead of starting over.
- Keep reviewing each draft so the marks reflect your judgement.
| Rubric element | Vague wording | Specific wording that marks well |
|---|---|---|
| Criterion | Analysis | Analysis of how language shapes meaning in the text |
| Descriptor | Good use of evidence | Integrates well-chosen quotations to support each point |
| Performance level | Excellent response | Sustained, insightful argument with precise textual support |
| Mark allocation | Marked holistically | 8 marks: 4 for argument, 4 for use of evidence |
Frequently asked questions
Can I use a rubric I already have in a document?
Yes. JeddAI marks against your own criteria, so you can enter or paste an existing rubric rather than building one from scratch. Keeping your original wording helps the drafted marking match how you already assess.
How detailed should my rubric be?
Detailed enough that each level is clearly distinct, but no longer than needed. A handful of specific, non-overlapping criteria usually marks more consistently than a long rubric with vague categories.
Can I use NESA, VCAA or QCAA marking bands?
Yes. Mirror whatever band structure your assessment uses, such as performance bands or an A-to-E scale, so JeddAI's drafts line up with how you report results.
Does the rubric change how JeddAI marks work?
The rubric is what it marks against. JeddAI drafts marks and feedback using the criteria and descriptors you provide, and you review and adjust every result before it reaches a student.
Do I have to build a new rubric for every task?
No. Once a rubric works well you can reuse it on similar assignments, which keeps feedback consistent and saves setup time each term.
Get started with Jeddle
Jeddle gives teachers and students instant, syllabus-aligned feedback powered by JeddAI.
Looking for study material? Browse Jeddle's Australian-English subject resources, or explore more articles on Using Jeddle.
