Quality rubric
Use this rubric to judge whether the AI output is classroom-ready or needs another iteration.
| Dimension | Classroom-ready output should include |
|---|---|
| Observable criteria | Descriptors mention visible evidence, not vague traits like excellent or weak. |
| Fair weighting | Scores align to taught objectives and avoid double-counting the same skill. |
| Feedback value | Each level includes actionable next steps students can use. |
| Bias check | Language avoids assumptions about background, disability, language status, or resources. |