ChatGPT Prompts for Course Design

Instructions

Looking for an AI co-author for your course?

Start anywhere.
Use with ChatGPT free or Plus.
One ChatGPT conversation for one course.
Don’t want to mix up courses? Review managing your memory in ChatGPT.

The Course Design Model

Stage 1: Develop a Course Outline and Learning Objectives

We will define your course’s learning outcomes and organize them by module or lesson. This step ensures a strong foundation, making sure your objectives align with Bloom’s Revised Taxonomy.

Stage 2: Develop Assessments for Each Lesson

For each lesson or module, create assessments aligned with the learning objectives. These assessments will measure student success and ensure that learning goals are met.

Stage 3: Develop Activities and Scaffold Student Learning

Develop instructional activities and identify engaging, accessible content that will help scaffold students’ learning toward achieving the desired outcomes.

Stage 4: Continuous Feedback and Improvement

Throughout the course, integrate continuous feedback loops to refine content and instructional strategies based on student performance and engagement.

Notes

Here are a library of prompts, inspired by the work of Ethan Mollick and the Prompt Library, for ChatGPT to help with course design.

Faculty “often work with subject-matter and content knowledge, which can be so expansive and complex” (Fang & Broussard, 2024) that finding a starting point or developing aspects of a course can be very challenging. These prompts give you the freedom and flexibility you need to design your course. This is about AI augmenting you to achieve more (Birss, 2023).

Following David Birss’s (2023) approach of Preparation, Interaction, and Continuation. I have done the preparation for you and have structured these prompts to allow you to interact and continue development with AI.

Disclaimer

This is for experimental purposes.
This uses AI, check everything.

References

Birss, D. (2023). How to Research and Write Using Generative AI Tools [Video file]. Linkedin. https://www.linkedin.com/learning/how-to-research-and-write-using-generative-ai-tools/how-to-work-with-ai

Fang, B., & Broussard, K. (2024, August 7). Augmented Course Design: Using AI to Boost Efficiency and Expand Capacity [Review of Augmented Course Design: Using AI to Boost Efficiency and Expand Capacity]. Educause. https://er.educause.edu/articles/2024/8/augmented-course-design-using-ai-to-boost-efficiency-and-expand-capacity

4 thoughts on “ChatGPT Prompts for Course Design”

kkm11 says:

October 4, 2024 at 9:55 am

Adding Development Notes to each prompts to track progress, issues, lessons learned, and future ideas.

Next: engineer a prompt for developing question banks, research other’s work since nothing here is new
kkm11 says:

September 19, 2024 at 1:53 pm
I just spoke with Amy (alh245) and got her initial impressions:
- She asked about the stages and why I had them broken up the way I did. I didn’t answer very well. I haven’t put any more thought into them since I first wrote them down. Amy made me realize that I should review and refine the stages to more closely match a backwards design approach and make sure that accessibility is included from the start of developing assessments and activities. It cannot be a separate step. While I always had that in mind, I don’t think that is well stated in the way I have things currently.
- I need to to a better job of emphasizing how to manage conversations in ChatGPT to ensure that information is retained and information is not conflated. Do I need to emphasize that the output will likely be different each time, even if you input the same prompts? I don’t think so. I think people understand the basic nature of GenAI at this point.
- Amy shared some interesting notes from a training event presented by Garima Gupta on prompt engineering. The speaker mentioned “Shots, Chains, Emotional Prompts, Chain of Destiny” and other concepts I’ll have to look into more. I feel pretty solid with my work so far, but am always willing to learn and improve my work.
- I’m going to keep going based on everyone’s feedback. My next steps are to review the stages, build out the rest of the stages simultaneously, knowing they will be unpolished, and then continue to refine all the prompts making sure they accomplish what I anticipate is needed from my users. I wanted to remind myself that I want to develop a specific prompt that mirrors work that our multimedia specialist, Aimee (asb193), has done on building a custom GPT to assist with image to text descriptions for accessibility. She wanted to improve upon ASU’s tool. For the same reasons why I chose not to use a GPT for my work here, we’ve discussed the benefit of possibly creating a complex prompt that could mirror what she’s done with her GPT.
kkm11 says:

September 17, 2024 at 5:28 pm
I shared this with Aaron (acy5072) and got his initial impressions:
- In our office, we often start working with faculty who already have some content. They might be revising an existing course. They might not be that far along but still have a define course description with course goals and a scope document for the development. I was thinking that while the current model doesn’t account for this, this isn’t a problem. While this started off sounding similar to what Logan shared earlier, Aaron was making a different point. After I have built out the stages, it would be easier to build in accommodations for these particular circumstances. I might build out alternate prompts or just build these into the existing ones. Since our office has Plus accounts, it would be relatively easy to upload content as zipped archives. I don’t know what limits we’d run into. I don’t think we could upload a single course archive, consisting of a GB of data, into ChatGPT but perhaps we could cherry pick and work lesson by lesson or in other similar chunks.
- Aaron also mentioned how he was reminded of an older Reddit post where one user shared their powerful experience of using ChatGPT as a therapist. Aside from the obvious dangers with doing this, Aaron shared this because it was an interesting use of AI with a very practical use case. There was something about that post that reminded Aaron of what I am trying to achieve. While I am not trying to engineer a therapist, I am trying to create a professional co-author for faculty to use to augment their expertise and create a tool that makes the process of designing or revising a course easier and more efficient.
kkm11 says:

September 17, 2024 at 2:13 pm
Just met with Logan (ljh5825) who had some great feedback when I asked him for his first impressions:
- Could be developed as a GPT but there considerations around sharing and Non-Plus (free) account limitations that needs to be addressed. I thought about that and this is one of the reasons why I went the route of developing a “complex” prompt. In addition, I like the openness of anyone having access to the prompt directly. That said, if I could learn more about the process of training a GPT, that might be something I look into to better refine this project.
- The time commitment might be an issue. If the first stage took 1 hour or 30 minutes to complete, would faculty still be willing to use this tool? That’s a great point and my hope is that the process is engaging enough and the results would be useful enough that the time would be worth it. In my testing, I invested something in the neighborhood of 15-20 minutes to complete stage 1 and was still impressed with the results. Would spending 30+ minutes yield better results, I don’t know. Is 15 minutes too much? Is it not enough? I don’t know. I’ll have to ask my faculty partners what they think. One benefit of using ChatGPT is that the user can stop the dialog at any point and return to finish the conversation. If they only have 3 minutes now, that is perfectly fine. The conversation can pick right up where it was left off.
- The process of revising a course is different than a new course development project. How might this prompt be adjusted to a revision project? Is it necessary? Would the process benefit from splitting the prompt into separate tasks? I don’t know but changes can be easily made. Even the existing prompt could be adjusted to ask whether the faculty was interested in a new course development or a course revision. I don’t think I’ll make any changes yet. I think it will be easier to engineer all the stages to address a new course development first and then adjust for other contexts, like a course revision, later.
- Logan closed with encouraging me to consider sharing this with others like Educause, TLT, etc. as a submission or presentation. I will definitely do so as I continue this work. I have already planned on sharing with partners around the University like HHD, WCLD, and others.