🧮 Rubric based grading
When creating Essay Questions, you have the option to enable rubric-based grading. This allows your AI Assistant to assess responses more accurately and provide structured, meaningful feedback based on clearly defined expectations.
Rubrics help set the standard for evaluation and ensure consistency across learner submissions.
🎥 Watch: Setting up Rubric-Based Grading
See how to create and apply rubrics for Essay Questions in action.
https://youtu.be/gdCyz8tyAYg
There are 2 ways to set up your scoring rubric:
-
Scoring Rubric with Descriptions
Define each skill or competency you want to assess, along with a description that helps the AI understand what to look for in learner responses.
Example:
- Clarity of Argument – Is the learner’s main idea clearly stated and logically supported?
- Use of Evidence – Does the learner support their points with relevant examples?
The AI will provide feedback based on these descriptions, all within the maximum score assigned
-
Full Breakdown of Scoring Rubric
Go deeper by assigning a maximum score for each criterion. This gives you both qualitative feedback and a numeric score for each part of the essay.
Example:
- Thesis Statement (5 points) – Clear, concise, and relevant
- Organization (5 points) – Logical flow and structure
- Grammar & Language (5 points) – Proper usage, tone, and vocabulary
This setup is ideal for formal assessments in which requires more control and precision in the grading
<aside>
⚠️
Tips for creating a good scoring rubric
- Define Clear Objectives: Before creating your rubric, clearly define what you want to assess. What are the learning objectives or skills you want to measure? Ensure these objectives are specific, measurable, achievable, relevant, and time-bound (SMART).
- Identify Criteria: Break down the objectives into specific criteria or dimensions that contribute to achieving those objectives. For example, if you're assessing a writing assignment, criteria might include organization, clarity, grammar, word count and creativity.
- Use Clear Language: Write clear descriptors for each level of performance that are easy to understand. Avoid jargon or ambiguous terms. Use concrete examples whenever possible to illustrate each level.
- Balance Criteria: Ensure that the criteria you choose are balanced and cover all aspects of the task or skill you're assessing. Avoid overly focusing on one aspect at the expense of others.
- Consider Weighting: Decide if certain criteria are more important than others and assign appropriate weights to them. This helps ensure that the rubric accurately reflects the relative importance of different aspects of performance.
- Determine Levels of Performance: Establish different levels of performance for each criterion. Typically, these range from "Excellent" to "Poor" or "Exceeds Expectations" to "Does Not Meet Expectations." Clearly define what each level entails in terms of performance.
- Test Your Rubric: Before using the rubric for assessment, test it with sample work to ensure that it effectively differentiates between levels of performance and aligns with your expectations.
Remember: the AI depends on you to provide specific instructions. You can imagine you are giving instructions to a colleague on how to assist you with grading papers and feedback. That would be the level of detail you need to provide the AI in the scoring rubric. If there are references your students need to quote, you need to be very specific what they are as the AI does not look for answers outside of the rubric provided.
</aside>
🎥 Watch: Learners’ view of rubrics
Take a peek at the how your learners’ receive their feedback when they submit their submission
https://youtu.be/Qotkvb5FjzA