Learn how Institutions use LLM to improve Teaching & Learning

GergelyTar
Instructure
Instructure
3
1001

Canvas.png

✏️This is a guest post written by Brittany Ward, Product Manager of Doowii Inc - partner on delivering Ask Your Data of Intelligent Insights.

How are Institutions using AI enabled natural language to Improve Student Outcomes?

As part of Intelligent Insights, Ask Your Data simplifies data exploration and analysis within Canvas. This tool enables institutions to use their data more effectively to support student success and institutional growth. No need for 3rd party tools or scripting, Ask Your Data provides a user-friendly interface to query, analyze, and visualize key performance metrics, facilitating data-driven decisions. Ask a question about your data and we’ll use AI to grab the result for you, summarize it then create an easy to understand visualization. 

Ask Your Data Insights

To date, we’ve seen thousands of questions asked! We’re highlighting some top themes and questions, but we would love to hear more from you in the comments section. What questions or themes have you found helpful?

Top Themes and Questions

Student Performance and Engagement
Ask Your Data helps administrators dig into student performance such as missing assignments and low scores, giving educators the opportunity to act before students fall too behind. Understanding the courses where students are struggling most, can help allocate student support resources where they are most needed.

  1. How many students have a current score lower than 70% in each course?

  2. Which students have more than 2 missing assignments?

  3. Show me all students with a score of 50% or less.

  4. Give me a list of current active term names. In each term, give me a count of students with a current score in the following ranges: <60%, 60% to 69.95%, 70% to 79.95%, 80% to 89.95%, and >89.95%.

System and Course Management
Course setup and management can be challenging and time consuming. Ask Your Data helps ensure courses are published and ready for student access as well as identifying trends or problem areas needing attention. Identifying courses needing attention can help ensure that students get the most support possible in their learning journey.

  1. Which teachers have the most ungraded assignments?

  2. Provide a count of all unpublished courses.

  3. How many assignments have the word "x" in the title?

  4. Which courses have more than 25 ungraded assignments?

User Activity and Engagement
Ask Your Data makes it easier to track individual user engagement and last login, which is helpful for instructors, student success teams, financial aid administrators, and many others. Both scores and engagement, while some students may be perfectly capable of succeeding without a lot of engagement, the combination of both is a pretty telling sign that something may be amiss.

  1. When was the last time user x was active in Canvas?
  2. Show students with a total grade lower than 65% or students whose total activity time was less than 20% of the average activity time for their class.

Keywords

Low Scores

Missing Assignments

Ungraded Assignments

Unpublished Courses

Average Activity Time

Conclusion

Let us know how you have found Ask Your Data helpful in the comments, or in our Community forum.

3 Comments
vancej
Community Participant

Sorry for  asking but what is  LLM?

dbrace
Community Coach
Community Coach

Hi @vancej,

"LLM" is the abbreviation for "large language model", also commonly referred to as GenAI (Generative Artificial Intelligence).

-Doug

mbmacdonald
Community Participant

"When was the last time user x was active in Canvas?"

I want to say that AYD seems like it is going to make life easier for a lot of Canvas admins - less time spent trying to understand data definitions etc. However, one thing I'm not sure it will help us understand is Mobile usage data. Is mobile usage data always tracked? Are timestamps accurate? Do users ever get logged out of the mobile app? If not, does that make all their access data suspect? I would really appreciate a one-stop shop for documentation/FAQs about mobile data. It would help me craft questions and understand the results I see.