Overview

A rubric is a key component of Feedback Aide. It defines the scoring criteria that a human or AI evaluator can use to assign a score and provide feedback to a learner.

The Rubric Editor enables you to create new rubrics for both essay-style and short-response assessments. This article explains when to use each rubric type, how to create them, and how to configure their scoring options.

Rubric Types

The Rubric Editor supports two major categories of rubrics: Essay Rubrics and Short Response Rubrics.

Essay Rubrics are used for evaluating longer responses such as essays, where performance is assessed either across multiple traits or as a whole.

• Analytic: Evaluates multiple criteria separately and assigns scores for each. This type supports both discrete scoring and multi-discrete (range) scoring.
• Holistic: Evaluates the response as a whole and assigns a single overall score. Like the analytic rubric, it supports both discrete and multi-discrete (range) scoring.

Short Response Rubrics are designed for short-answer questions where students must demonstrate specific correct responses.

• General: Provides correct/incorrect binary scoring with support for alternate responses.
• Key Points: Evaluates whether key elements are included in the response.
• Categorize: Assesses how well students classify criteria into correct categories.
• X from Y: Awards points when students provide a specified number of correct answers from a larger set.

Creating an Analytic Rubric

Analytic rubrics evaluate student responses across multiple traits or criteria, with separate scores assigned for each. This type of rubric provides clear, detailed feedback to learners and helps identify specific areas of strength and improvement.

When to use an Analytic Rubric

• Use an analytic rubric when a task requires learners to demonstrate different skills or traits. For example, an essay may be evaluated for organization, strength of arguments, grammar, and integration of evidence, with each aspect scored independently.

• Choose this rubric type when there are specific, measurable performance requirements. For instance, if learners must integrate three sources to earn full points, or if correct use of a particular grammatical concept is required, these expectations should be explicitly stated in the rubric.

• It is ideal when detailed feedback is needed to guide learners. By breaking performance into separate criteria, instructors can show exactly which areas meet expectations and which need improvement, making it especially effective for formative assessments.

• Analytic rubrics are also well-suited for complex or multi-step tasks, such as research projects or portfolios, where each component (e.g., research quality, analysis, presentation) must be evaluated transparently and fairly.

How to create an Analytic Rubric

To create an Analytic rubric, go to the Create rubric view. From the rubric type dropdown. In the Point range box, you can choose between None and Multiple. When the None option is selected, each rubric level is assigned a fixed score. When the Multiple option is selected, you can define a range of possible points for each level. By default, the None option is applied.

The Rubric levels section defines the scoring levels. Three default levels are provided—Poor with a score of 1, Satisfactory with a score of 2, and Outstanding with a score of 3. You can rename these levels as needed. If the Point range is set to None, each level must be given a fixed score. Additional levels can be added by clicking the plus icon (+), up to a maximum of six levels. Levels can be reordered using the move-row option on the left side, and they can be deleted using the trash icon.

The Criteria section allows you to define the traits that will be assessed against the rubric levels. Each criterion represents a scoring dimension, and the associated points or ranges are automatically added to calculate the total point. To add a new criterion, click the + Criteria button at the bottom of the section. You can add up to ten criteria in total. For each criterion, you must provide a name and description of the trait, along with level-specific descriptions that explain performance at each rubric level. Criteria can be reordered or deleted using the ellipsis (⋯) icon.

In addition, the rubric includes an Additional grading guidance field. This optional text field allows you to include extra context or instructions for Feedback Aide that apply to the rubric. For more information, refer to the Additional Grading Guidance section within the Feedback Aide Rubrics article.

To save the rubric, click Save button at the bottom of the Rubric Editor.

If the Point range is set to Multiple, you must define a score range for every level within each criterion.

Once the rubric is defined, you can review it in two different views. By selecting the Preview tab at the upper-right corner of the Rubric Editor, you can see how the rubric will appear when used. By selecting the Source tab, you can view the rubric’s JSON data.

Creating a Holistic Rubric

Holistic rubrics provide a single overall score that reflects the quality of a response as a whole. Instead of assigning points to multiple criteria independently, you define one set of scoring levels within each criterion. This format is best suited when the goal is to emphasize the overall performance of a response rather than breaking it down into detailed parts.

When to use a Holistic Rubric

• Use a holistic rubric when the primary goal is to assess the end product or overall impression, rather than analyzing individual components. For example, scoring a final essay or creative project where the combined effect of ideas, organization, and style matters most.

• It is ideal when speed and efficiency in grading are important, as it allows instructors to evaluate quickly without assigning separate scores to each criterion.

• Choose this rubric type for summative assessments where detailed feedback is not the focus, such as standardized tests or large-scale evaluations.

• Holistic rubrics are also effective when all criteria are closely interconnected, making it difficult to separate them meaningfully. For instance, assessing the overall persuasiveness of an argument or the overall creativity and impact of a presentation.

How to create a Holistic Rubric

To create a holistic rubric, go to the Create rubric view. From the rubric type dropdown, select Holistic. In the Point range menu, you can choose between None and Multiple.

• If None is selected, each level is assigned a fixed point value.
• If Multiple is selected, you can define a point range for each level, which allows scoring granularity within that level.

By default, the None option is applied.

All rubric levels are defined directly inside each criterion section. Each criterion can have between one and six levels. A level requires:

• A name (for example, Poor, Satisfactory, Outstanding),
• A score or score range (depending on the Point range option selected), and
• A description that defines performance expectations for that level.

To add another level, click the + Level button within the criterion. Levels can be reordered using the drag handle on the left or deleted using the trash icon.

To add more criteria, click the + Criteria button at the bottom of the rubric. You can create up to ten criteria. The total points across all criteria are automatically calculated and shown at the bottom of the Rubric Editor.

In addition, the rubric includes an Additional grading guidance field. This optional text field allows you to include extra context or instructions for Feedback Aide that apply to the rubric. For more information, refer to the Additional Grading Guidance section within the Feedback Aide Rubrics article.

To save the rubric, click Save button at the bottom of the Rubric Editor.

Once the rubric is defined, you can review it in two different views. By selecting the Preview tab at the upper-right corner of the Rubric Editor, you can see how the rubric will appear when used. By selecting the Source tab, you can view the rubric’s JSON data.

Creating a General Rubric

The General short response rubric type is best suited for questions that require a short answer or explanation. In this format, the student’s response is evaluated as either correct or incorrect. The scoring is binary:

• 1 indicates the response is correct.
• 0 indicates the response is incorrect.

The rubric includes a single Correct response, along with two optional supporting lists: Alternate responses and Incorrect responses. These lists help guide the AI by showing which answers are close to correct (but may be vague or incomplete) and which are incorrect. This ensures consistent scoring and makes it clear how generously or strictly responses should be marked.

When to use a General Rubric

• The question requires a short, focused response, such as a definition, identification of a concept, or a simple explanation.

• You want to check for basic comprehension or understanding of key ideas rather than evaluate multiple skills or criteria.

• There is a single core correct answer, but alternative phrasings are acceptable, and a supporting list of alternate responses can help guide consistent scoring.

How to create a General Rubric

To create a general rubric, go to the Create rubric view. From the rubric type dropdown, select General.

• In the Correct response box, enter the expected correct answer.
• In the Alternate responses section, add variations of the correct answer that are acceptable. You can add up to 10 alternate responses. To add one, click the plus icon (+). To remove one, click the trash icon next to the alternate response box.
• In the Incorrect responses section, enter answers that are wrong or misleading. You can add up to 10 incorrect responses. To add one, click the plus icon (+). To remove one, click the trash icon next to the incorrect response box.
• In the Additional grading guidance section, the optional text field, allows you to include extra context or instructions for Feedback Aide. For more information, refer to the Additional Grading Guidance section within the Feedback Aide Rubrics article.

Once you have added the responses, the Total points value is automatically set to 1. To save the rubric, click Save at the bottom of the Rubric Editor.

Once you define the rubric, you can view it in two formats. Select the Preview tab to see how it will appear during scoring or select the Source tab to view the underlying JSON data.

Creating a Categorize Rubric

The Categorize rubric type is used to evaluate a student’s ability to classify or group information into categories. Each rubric includes a set of categories, with each category containing one or more criteria. Every criterion in a category is worth 1 point, and the total point is calculated as the sum of all criteria across all categories.

When to use a Categorize Rubric

• When students are required to group a list of items into specific, predefined categories, such as sorting animals by habitat, literary devices by type, or historical events by time period.

• To assess a learner’s ability to recognize relationships, patterns, or connections between different concepts or objects.

• When evaluating understanding of classification systems, such as scientific taxonomies, grammatical rules, or data groupings.

• For tasks where students need to match concepts to definitions, like pairing vocabulary words with their meanings or linking causes with effects.

• To measure how well students can organize ideas or evidence, such as arranging arguments by theme or categorizing data by type.

• When students must differentiate between similar or closely related concepts, demonstrating the depth of their understanding.

How to create a Categorize Rubric

To create a categorize rubric select Categorize from the rubric dropdown. 

In the Category name box, enter the name of the category (for example, Photosynthesis or Cellular Respiration).

• In the Points boxes under the category, enter the criteria that belong to that category. Each criterion counts as 1 point.
• To add another criterion (point) to the category, click the plus icon (+).
• To delete a criterion, click the trash icon next to the criterion.

To add a new category, click the + Category button. Then enter the category name and add the corresponding points. You can create up to 10 categories in a single rubric.

You can reorder or delete a category by clicking the ellipsis (⋯) icon at the right side of the category row. The Total points value is automatically updated as you add criteria across categories.

In addition, the rubric includes an Additional grading guidance section. This optional text field allows you to provide overall instructions or context for Feedback Aide, such as clarifying how to handle ambiguous responses or outlining specific expectations for categorization. For more information, refer to the Additional Grading Guidance section within the Feedback Aide Rubrics article.

To save the rubric, click Save at the bottom of the Rubric Editor.

Once you define the rubric, you can view it in two formats. Select the Preview tab to see how it will appear during scoring or select the Source tab to view the underlying JSON data.

Creating a Key points Rubric

The Key points rubric is used to evaluate the correctness of a student’s response based on how many essential points they correctly identified or calculated. Each rubric includes a list of key points that the student is expected to include in their answer, with each point assigned a specific score. The total score is calculated as the sum of the values assigned to all the key points.

When to use a Key points Rubric

• When students need to identify multiple specific components or steps in their response, such as stages of a process or elements of a concept.

• For multi-step calculation problems in subjects like math or science, where each step must be correct to reach the final solution.

• When a question asks for a specific number of reasons, causes, or effects, such as explaining three causes of a historical event or listing two effects of a scientific phenomenon.

• For evaluating detailed written responses, where students must include several key ideas to demonstrate full understanding.

• When you want to pinpoint partial understanding, as this rubric allows students to earn credit for the correct parts of their answer even if the response is incomplete.

How to create a Key points Rubric

To create a Key points rubric, go to the Create rubric view. From the rubric type dropdown, select Key points.

• In the Key points box, enter the key element the student should include in their response (for example, Contribution per unit = $6.40).
• Enter the score for the key point in the box next to it. Each new key point must have a score assigned. By default, a new key point is set to 0.
• To add another key point, click the plus icon (+). You can add up to 10 key points in a rubric.
• To delete a key point, click the trash icon next to it.

The Total points value at the bottom is automatically updated as you add or adjust the scores for key points.

In addition, the rubric includes an Additional grading guidance section. This optional text field can be used to clarify how partial credit should be assigned, define acceptable alternative responses, or note specific expectations for reasoning or justification. For more information, refer to the Additional Grading Guidance section within the Feedback Aide Rubrics article.

To save the rubric, click Save at the bottom of the Rubric Editor.

Once you define the rubric, you can view it in two formats. Select the Preview tab to see how it will appear during scoring or select the Source tab to view the underlying JSON data.

Creating an X from Y Rubric

The X from Y rubric type is used when a student must provide one or more (X) correct responses from a larger set of (Y) possible correct responses. This is useful when a question allows for multiple valid answers, but the student is only required to provide a subset.

Each response is scored in a binary manner:

• 1 if the student has provided a correct response with sufficient detail.
• 0 if the student has not provided the correct response.

The total score is calculated by summing the number of correct responses identified by the student, up to the maximum specified in the Points needed box.

When to use an X from Y Rubric

• For questions such as “Name three causes of…” or “List two examples of…”, where several possible correct answers exist, but only a certain number are required.

• When you want to assess a student’s breadth of knowledge without requiring them to list every possible correct response.

• In assessments that test recall and understanding of key concepts, such as identifying symptoms of a condition, types of figurative language, or examples of renewable resources.

• For tasks where variety in answers is expected, and students may demonstrate knowledge in different valid ways.

• When you need simple, binary scoring, with each provided response marked as correct (1) or incorrect (0), making grading quick and consistent.

How to create an X from Y Rubric

To create an X from Y rubric, go to the Create rubric view. From the rubric type dropdown, select X from Y.

• In the Points needed box, enter the maximum number of correct responses a student can provide to receive full credit. This value is displayed as the total points for the rubric.
• Enter a correct response in the Correct response box.
• Enter an alternate response for the correct response in the Alternate responses section.
• To add more alternate responses for a correct response, click the plus icon (+) below the alternate response box. You can add up to 10 alternate responses for each correct response.
• Enter an incorrect response for a correct response in the Incorrect responses section.
• To add more incorrect responses, click the plus icon (+) below the incorrect response box. You can add up to 10 incorrect responses for each correct response.
• To delete an alternate or incorrect response, click the trash icon associated with it.
• To add additional correct responses, click the + Correct response button at the bottom of the rubric. You can add up to 10 correct responses.
• If you have multiple correct responses, you can reorder or delete them using the options available in the ellipsis (⋯) icon next to each correct response.
• The Additional grading guidance section allows you to provide overall instructions or context for Feedback Aide. For more information, refer to the Additional Grading Guidance section within the Feedback Aide Rubrics article.

The Total points value at the bottom is automatically updated based on the number specified in the Points needed box.

To save the rubric, click Save at the bottom of the Rubric Editor.

Once you define the rubric, you can view it in two formats. Select the Preview tab to see how it will appear during scoring or select the Source tab to view the underlying JSON data.

Testing your rubric with Rubric Practice

The Rubric Practice feature lets you test your rubric before deploying it to learners. By running sample responses through Feedback Aide and comparing the results to your own scoring, you can catch gaps or ambiguities in your rubric's criteria early, so learners receive accurate, consistent feedback.

This guide covers the Analytic rubric type. Documentation for other rubric types will be available separately.

This guide walks you through how to access Rubric Practice, how to use it, and how to interpret the results.

1. Before you begin

Rubric Practice must be enabled for your Feedback Aide consumer. If you do not see the Test tab, contact your system administrator or Learnosity Customer Success Manager.

When it is active, a Test tab appears inside the Rubric Editor alongside the Edit, Preview, and View Source tabs.

2. Opening the Test tab

1. Open the Rubric Editor for the question you want to test.
2. Click the Test tab in the top-right area of the editor.

The “Stimulus” is the question the learner will answer. To run tests in Rubric Practice, the question stimulus is required. The stimulus is passed in through your application's integration with Rubric Editor.

3. Adding sample responses

Each response panel allows you to enter or paste a sample learner response to see how Feedback Aide will grade it against the rubric. You can test up to five responses at a time.

Test a range of quality levels, for example, one strong response, one average response, and one weak response. This gives you the most useful picture of how your rubric performs across the full score range.

1. Enter your sample response in the text editor. You can also paste from a document.
2. The editor supports basic rich-text formatting: bold, italic, underline, bullet lists, numbered lists, and indentation.
3. To add another response, scroll to the end of the panel and click the button for adding a new response. New responses are added below the existing ones.
4. To remove a response you no longer need, click the Delete button inside that response panel.

Character limit: For best results, keep responses under approximately 5,000 words (30,000 characters).

4. Selecting your expected results

Before running Feedback Aide's evaluation, you can select the scores you expect for each response. This is optional, but it lets you compare your judgment to Feedback Aide's and identify any discrepancies.

1. Below the response text editor, find the rubric with its criteria listed.
2. For each criterion, click the level that best matches the response. The selected level is highlighted and marked with a blue person icon.
3. Repeat for each criterion.

You do not need to score all criteria before running a test, but doing so will allow for better comparison between the expected and Feedback Aide score.

5. Running a test

There are two ways to run a test:

• Run test - evaluates a single response. Click the Run test button inside the response panel you want to evaluate.
• Run all tests - evaluates all responses at once. The Run test buttons on individual panels are disabled while "Run all tests" is in progress.

A loading indicator appears while the evaluation is running. Depending on the length of your response, this typically takes a few seconds.

6. Understanding the results

Once a test completes, several components are updated inside that response panel.

Response header badges

At the top of the response panel you may see one or two badges:

• Credits badge - shows how many credits were consumed for this evaluation. Credit deduction is the same as if grading a real learner response. 
• All criteria scores match badge - appears in green when Feedback Aide's scores agree with all of your expected scores.
• Mismatches found badge - appears in red when Feedback Aide's scores differ from your expected scores on one or more criteria.

The comparison badge only appears if you have entered expected scores for at least one criterion before running the test.

The score comparison

After a test runs, the rubric beneath the response shows two sets of scores side by side:

• A blue person icon on a level indicates your expected score for that criterion.
• An orange icon on a level indicates the score assigned by Feedback Aide.

The criterion label uses color to signal the comparison at a glance:

• Green with a checkmark - your expected score and Feedback Aide's score are the same for this criterion.
• Red with a warning icon - your expected score and Feedback Aide's score differ for this criterion.

For short response questions, your expected score and Feedback Aide’s score are shown side by side for comparison. Below is an example of an x-from-y question.

Feedback to learner

Below the rubric, Feedback Aide shows the criterion-by-criterion feedback it would deliver to a learner. Each criterion displays the level it selected and a written comment explaining the score and suggesting how the learner could improve. This is the same feedback your learners would see in the live question, so reviewing it here helps you confirm the tone and content align with your expectations.

7. What to do when you find mismatches

A mismatch does not always mean something is wrong. Use mismatches as a prompt to ask the following questions:

• Are the level descriptions clear enough? If Feedback Aide is consistently scoring one level above or below your expectation, the boundary between two adjacent levels may be ambiguous. Consider tightening the description to make the distinction more explicit.
• Is additional grading guidance needed? The rubric editor has an Additional grading guidance field at the bottom of the Edit tab. Use this to give Feedback Aide specific instructions, for example, "Focus on argumentation, not grammar" or "Ignore UK/US spelling differences." Then re-run the test to see whether the guidance resolves the mismatch.
• Was your own scoring correct? Sometimes reviewing the rubric levels alongside the response reveals that Feedback Aide's score was the more accurate one. In those cases, no rubric change is needed. The test has simply validated that the rubric is working as intended.

After making changes, switch back to the Edit tab, update the rubric, then return to the Test tab and run the test again. Repeat this cycle until you are satisfied with the results.

8. Confirming your rubric

Rubric Practice is a testing tool only and does not modify your rubric automatically. 

When you are satisfied with your rubric and testing results, click Confirm rubric at the bottom of the Rubric Editor. This returns the final rubric to your application. Note, this button may not be present depending on your integration, if so, no further action is required.

Any sample responses and expected scores entered in the Test tab are not included in the final rubric JSON. They exist only for the current session to help you validate your rubric before deploying it.

Feedback Aide credit usage with rubric practice

Each time Feedback Aide evaluates a response, it consumes credits. The number of credits used is shown in the credits badge on the response panel after each test. Your administrator sets the number of credits available per session. When the credit limit is reached, the run test buttons become disabled and a warning message is displayed.

If you need more credits for testing, contact your platform administrator.