Designing and teaching online courses for a diverse group of students can be a complex and time-consuming task. From developing clear instructions to creating assessments that accurately measure students’ understanding of the material, faculty members face the challenge that their course materials are engaging and fair for everyone. Recently, educators have begun turning to AI to help refine these processes—testing assignments, restructuring rubrics and improving communication with students. Here’s how one faculty member, Margaret Sass, uses AI to transform her online course design process.
AI-powered personas: Testing assignments before launch
When preparing an online course, one of the biggest challenges is ensuring the assignments are accessible to all traditional and nontraditional students. With a diverse mix of students in every class, it can be difficult to create instructions that all students understand. By creating personas based on her students, such as a 35-year-old full-time working woman and a 21-year-old male student working part-time, Margaret has been able to “test” her assignments before the course begins.
With the assistance of AI, mainly ChatGPT, Margaret crafts these personas with specific details about their backgrounds, demographics, barriers, and potential perspectives. This approach allows her to see how different students might respond to her instructions, helping her fine-tune the assignments and identify what’s missing. Instead of waiting for student feedback during the course, she can address issues in advance.
Restructuring rubrics that supports all students level of learning
After testing assignments through AI-generated personas, the next step is refining the rubrics. Bias can creep into assessment design, particularly when assignments inadvertently favor certain demographics. For example, Margaret realized that designing rubrics based on responses from one group (e.g., 35-year-old women) could unintentionally disadvantage younger students or those from different backgrounds.
By using AI, she was able to redesign her rubrics to be more inclusive. The rubrics are now simplified, making the assessment criteria clearer for all students, regardless of their background or experience level. The result reflects the diversity of the classroom. (See below for an example of a conversation with ChatGPT analyzing an assignment rubric using personas.)
Discovering biases—A self-reflection
While using AI to test assignments and develop better rubrics, Margaret also discovered biases inherent in both the AI and herself. These AI-generated personas brought to light assumptions she hadn’t realized she was making. For instance, certain questions or instructions may have unintentionally favored one persona over another based on her own limited knowledge, prompting her to reflect on these biases and adjust her approach.
This realization encouraged her to re-examine her course content. With AI’s help, she became more aware of potential gaps in her assignments, ensuring the learning environment was as inclusive as possible.
What happens after testing the assignment?
Once the assignment has been tested using AI personas, the next step is refining it based on the feedback from the AI’s simulated responses. Whether it’s simplifying the instructions or making the rubric more transparent, this iterative process helps create more student-centered and inclusive assignments before they reach actual students.
Tips for getting started
If you are looking to get started, here are steps to use to do so:
- Create diverse personas: Simulate students from various backgrounds and demographics to preview how different groups might engage with the assignment.
- Test and refine instructions: Use AI responses from these personas to identify confusing or biased language in assignment prompts.
- Restructure rubrics: Leverage AI insights to simplify and make grading criteria fair for all students, regardless of their backgrounds.
- Identify and address biases: Use AI to reflect on and eliminate unconscious biases in your assignments or rubrics.
Conclusion: A new approach to course design
By incorporating AI into the design process, you can create more effective, inclusive, and fair online courses. Testing assignments with AI-generated personas help anticipate how diverse students might engage with the material while restructuring rubrics ensures fair grading. While AI is not a replacement for human insight or connection, it offers a powerful tool for refining and improving the online learning experience.
Resources
The following resources can help you incorporate Journals in your online course:
- Example of Using ChatGPT and Personas to Test an Assignment Rubric
- Innovation Playground: AI Personas (Nov 12, 2024, 12:00 – 12:50 p.m.): In this Zoom session, get a chance to try out using ChatGPT to create personas and test your assignments and rubrics.
- AI Webinars and Seminars: Search our eCampus faculty development events page in CampusGroups to learn more about using AI for teaching and learning.
- AI Articles on TALK (Teaching and Learning Knowledge Base)
For additional help with using AI personas for your online course design, consider contacting eCampus Center to request a consultation.
Article credit
Thanks to Professor Margaret Sass and Anthony Saba, the eCampus faculty development coordinator who worked with Sass to write this article. ChatGPT was used to help revise and edit the original article draft.