The Learnosity Grading API is an embeddable grading solution for scoring assessments inside your own system. It provides a streamlined interface for teacher review and grading of all Learnosity Items from any activity, including essay, audio, and video, as well as autoscored question types like multiple choice.

View the Grading API demo.

Grading API is designed to seamlessly integrate into your system, enabling educators to score learner responses and leave feedback to support further learning. 

Figure 1: A screenshot of the grading inline UI featuring Item review and a score and comment box.

How it Works

Inline Grading UI

The out-of-the-box grading UI allows graders see the Question stem and the learner response in review mode on the left. On the right is the grading inline component, featuring a score and a feedback section. 

Scoring

Graders can evaluate a learner response and assign a score. This score is then saved with the grading SessionID. If no maximum score was specified for the Item at authoring time, a max score must be added, by clicking the pen icon, before saving.

Figure 2: A screenshot showing an Item with no max score.

Figure 3: A screenshot showing how to update the max score. 

A grader can update the max score using the pen icon. It's important to note this change only applies to the current session, not to the Item nor to other learners who took the same Activity.

Feedback 

In the text box, graders can provide plain text comments to learners, which are then stored with the session. Currently, only plain text is available. For richer feedback see the demo which uses Data API and Learnosity Items. 

Grading Autoscored Question Types

Grading API allows graders to score both non-autoscorable and autoscorable Question types. When a grader changes an autoscore, the new manual score will become the score for the session. However, the autoscore will still be saved and available in a separate score slot. 

Storing Scores and Feedback

The API allows multiple scores and feedback entries per response, storing up to 10 scores and 10 feedback Items per response. These are indexed by the grader ID (user_id in the Grading API initialization config) and are referred to as "slots."

Scores can come from automated scoring or be assigned manually via the Grading API. One score is designated as the reference score - this is the score used for calculating Item and overall scores, and appears on reports. The reference score is always the most recent manually assigned score.

Multiple Graders and Score Updates

Grading API is not currently designed for workflows where multiple graders independently assess the same response. The Grading API is not recommended for handling multiple graders at this stage.

Updating a score

How a score is updated depends on the grader ID. If there is an existing score, then:

• If the updated score has the same grader ID, the previous score is replaced.
• If the updated score has a different grader ID, the score is stored in addition to the existing score. The newest manual score becomes the reference score.
• If 10 scores are already stored, the oldest manually added score is removed.

Since the reference score is always the latest manual score, the grader ID associated with the reference score may change over time.

Feedback Behavior

Feedback is stored in a similar way to scores, with the main difference being that the oldest feedback is replaced with the newest once the limit of 10 feedback entries is reached.

Grading UI Display

The UI shows only the score and feedback for the current grader ID. If multiple grader IDs have been used, the displayed score may not be the current reference score. 

Autoscoring and Rescoring

If a Question is autoscorable, one of the 10 score slots is reserved for the automatically generated score. Rescoring (e.g., on session resubmission) updates this specific slot—it does not create a new score entry and does not change the reference score. 

Therefore, if there is an existing manual score and the session is resubmitted, the manual score assigned before resubmission will remain the reference score. 

The autoscore will never be removed when new scores are added; only manually assigned scores are candidates for removal.

Learner Review

Read-only mode provides an easy way for learners to review their responses, scores, and feedback. 

Figure 4: A screenshot showing learner review mode, with the Item and response on the left and the read only score and feedback on the right.