Updated January 2025
What is the updated AI feedback model?
Writable’s updated AI feedback model, or Version 2 (v2), combines advanced AI technology with sample writing submissions aligned to specific skills and standards.
This integration results in more accurate scoring and meaningful feedback. By analyzing curated sample submissions, the AI gains a clearer understanding of writing expectations, ensuring alignment with the appropriate rubric scores and customizing feedback to those criteria.
Think of Version 2 (v2) like a new, improved GPS for writing. It combines:
Powerful AI technology (the “GPS directions”)
Sample writing submissions (the “map points”)
The AI “studies” real student examples from Writable and it understands how to better align each piece of writing to the right standards and skills. This way, the feedback is more precise and aligned with what teachers expect.
On This Page:
More about Sample Submissions
Sample submissions will act as benchmarks for the scoring, helping the AI:
Distinguish between strong, average, and weak responses based on grade-level expectations and rubrics.
Calibrate its scoring to align closely with rubric criteria, ensuring more accurate and consistent writing scores.
Generate targeted, contextual feedback that reflects rubric and genre nuances while remaining actionable and student-friendly.
Note: Sample writing submissions are a unique mix of Writable's curated student writing from across genres and grade levels combined with the prompt and rubric for the specific assignment. Teachers will not upload their own sample submissions at this time.
The AI analyzes different papers (the sample submissions) and figures out what makes the writing strong, average, or needs improvement.
Since it sees clear exemplars of writing, it better understands how to score student papers based on the assignment specifics.
As a result, feedback is targeted (directly related to the rubric and more) and actionable (students know exactly what to fix or improve).
How does it impact grading?
Writable’s enhanced AI feedback model empowers educators with more reliable and precise tools for evaluating student writing. Teachers save time by aligning feedback with specific skills, standards, and rubric criteria while maintaining consistency and fairness in grading. Students, in turn, benefit from actionable feedback that helps them understand their strengths and areas for improvement, fostering growth and confidence in their writing skills.
Teachers can trust the AI to consistently measure student writing against the right criteria, saving time and minimizing guesswork.
Students receive more than a score - they get specific tips on how to improve their writing. This builds confidence and helps them become better writers faster.
Can I still use the v1 scoring model?
Yes, the Version 1 (v1) scoring model is still available. To select it, go to the "Edit Assignment" page and click "More." You can find more step-by-step guidance here.
While we recommend transitioning to Version 2 (v2) for new assignments due to its enhanced capabilities, we understand that maintaining consistency with v1 may be important for tracking growth in common assessments across the school year.
What does that mean?
If you have ongoing projects or assessments where you want to keep the old method (v1) for consistency, you still can.
However, if you want the improved features (v2), just switch it on for new assignments.
Glossary
Here are some common terms used in this article and in other resources regarding Writable's AI feedback tools.
Draft Scores - An AI-suggested score on the rubric items and the overall writing. These can be edited by teachers.
Grade Assist - The general term referring to AI-powered grading supports, including Draft Scores and Suggested Comments.
RevisionAid - A student-facing AI support that provides actionable feedback directly to students. This can be enabled when assigning or by editing the assignment options.
Sample Submissions - Benchmarks that teach the AI what “strong,” “average,” or “needs improvement” writing looks like.
Suggested Comments - AI-generated comments based on student writing and the assignment prompt and rubric. Teachers can edit these before sending to students.
Version 1 (v1) - The previous AI scoring model used in Writable assignments.
Version 2 (v2) - A new, more accurate AI scoring tool that uses sample submissions alongside genre, rubric, and prompt specifics.
Need help with the new AI scoring model?
If you have any questions about our new AI scoring model please contact our team at support@writable.com.