How does grading scale extraction work?

1️⃣ What does “grading scale” mean in Ed.ai?

In Ed.ai, a grading scale can be:

• a points-based grading scale (e.g., 2 points for setting up the equation, 1 point for the answer),

• or a rubric (e.g., “Reason abstractly and quantitatively”, “Construct viable arguments”).

👉🏼 Ed.ai handles both formats and uses them as the basis for its analysis.

2️⃣ Is a grading scale required?

No, but it’s strongly recommended.

When you create your assessment, you can add a grading scale (points-based or competency-based). It:

• gives the AI a structure for grading,

• gives you a clear framework you can edit at every step,

• aligns the scoring with your own expectations.

📌 If you don’t add one, you can always create one later, or edit the one Ed.ai builds automatically.

3️⃣ Can Ed.ai work with a partial grading scale, or none at all?

✅ Yes. If you add only part of a grading scale (or none), Ed.ai automatically builds or completes the detailed grading scale from your question sheet and your answer key.

This grading scale:

• is built on our math skill frameworks, which follow Common Core, AP and state standards,

• breaks each question down in a consistent way,

• is fully editable by you before grading.

⚠️ Without a points-based grading scale or a rubric, the assessment is graded and analyzed, but it isn’t scored, with points or by competency (message “No rubrics detected”). To get a score, add a points-based grading scale or a rubric to the assessment documents.

📌 Example of a teacher’s grading scale:

📌 Example of the grading scale Ed.ai structures from that answer key:

💡 To go further, see Sur quels contenus pédagogiques Ed.ai entraîne-t-il ses modèles d’IA ?.