"

AI and the CRAFT Discussion Strategy

Theresa Huff

Learning Objectives

By the end of this chapter, you will be able to

  • Identify Appropriate AI Usage Guidelines
  • Identify the Steps of the CRAFT Strategy

The call for digital literacy (Breakstone et al., 2018) and removing the digital divide (Ragnedda & Muschert, 2013) in education has been a driving force in the last twenty years or so. More recently, another call has added to this demand: AI literacy (Ng et al., 2021). As instructional designers and leaders, the speed at which AI is developing can sometimes make it feel like an impossible task to a) daily wrap our heads around it, (b) train faculty in how to wield it (or even adapt to it), and c) design instruction on it for students. But, you, leader-and/or-instructional designer-on-your-way-to-getting-your-expert-degree-in-this-field, will be uniquely called upon to do all of the above.

Disclaimer: Even as I draft this chapter, new AI tools and new capabilities by AI are being launched. So, by the end of this course, this information may be obsolete (sigh). But, in this moment, I’ll offer what best practices and knowledge I have.

Using AI

I’ll be asking you to use some free AI tools in this course. You’ll need to choose an AI tool, but you don’t have to stick to the same one week-to-week. As our syllabus states, it’s up to you to read the fine print for any tool you use. So, be sure to read how the AI tool is using your data. I recommend the free versions of ChatGPT, Claude, Gemini, or Copilot. For image creation, play around with a free version recommended for educators. I’ll be recommending other free AI tools for specific uses in the course as we go.

Regardless of how much experience you have using AI, it’s always good to review the basics (how does AI work and what should we be aware of?) and some tips for effectively prompting AI (how do I get AI to do what I want it to do?). As your level of AI knowledge calls for, view the two videos below from Ethan and Lilach Mollick of Wharton’s Generative AI Labs.

The AI Basics

How to Effectively Prompt an AI

Note: Some AI experts say that as AI progresses, the focus on crafting a prompt is becoming less and less important or necessary. However, for this moment in time and for this course, I’ve offered guidance through the above video on crafting prompts, and we will be transparently sharing the prompts we used with one another in the the CRAFT assignments.

Some Final AI Guidance

Before you begin using AI in the course, review these important pieces of information. Select the arrow next to each title to read more.

The CRAFT strategy

If you’ve taken any courses with Dr. Curry at Idaho State University, you will likely be familiar with the MANIC assignment (Curry & Cook, 2014). Used in Discussion posts, this framework helps facilitate thoughtful, specific, posts and meaningful interactions with classmates. Students follow the MANIC acronym to critically review a journal article, a lecture, or other information and point out key talking points. The MANIC acronym stands for:

  • Most important – What was the most important thing you gleaned from this information?
  • Agree – What is something you agree with from this information?
  • Not agree – What is something you disagree with from this information?
  • Interesting – What was something you found interesting or insightful?
  • Confusing – What was one thing that was confusing to your or that you feel needed clarification?

In this asynchronous course, we have only a few ways to communicate with one another: Office Hours and Discussion posts. Therefore, having meaningful Discussions is vital. Additionally, you will be regularly interacting with and critically evaluating AI in the rest of this course. We will be sharing our prompts, our AI outputs, our critical analysis, and our reflections with one another in the interest of:

  • transparency for integrity
  • learning at least ten times as much as we would on our own
  • gathering a fuller understanding about the abilities, limitations, pros, and cons of this evolving technology

To aid your process and sharing, I give you my MANIC spin-off, the CRAFT framework. Full disclosure, I co-created this framework with ChatGPT 4.0 just for you.

Here’s how this will work. Like the MANIC strategy, each week you’ll be prompted to use an AI tool to complete an assignment. This might be using an AI tool to act in a particular role like a Subject Matter Expert (SME) or your debate coach, to generate something for your textbook chapter, or to apply a learning theory to create instruction for Part II of that week’s ORB and post to a Discussion. You’ll use the CRAFT framework to complete these and organize your Discussion post.

C – Clarify

Before interacting with the AI, clarify what it is you are planning to do. A clear purpose ensures your prompts are targeted and effective. Also, it is easy to get distracted once you start interacting with an AI tool, so stating your goal can help if you get lost in fascination and wonder “now what was I looking for?”. For this step, you’ll simply decide and state your goal AND state which AI tool you are using.

Clarify – Example

  • ASSIGNMENT: AI as Subject Matter Expert (SME) directions- Use one of the free AI tools to collaborate as if they were a Subject Matter Expert the area of instruction for which you are designing instruction. The SME is also a behaviorist.
  • Your stated goal:
    • Clarify: Using ChatGPT, I need to create instruction for a fourth grade teacher who is a behaviorist on the topic of the life cycle of a butterfly. 

R – Request

Now that you’ve got a plan, make your request (or prompt) of the AI. While you can just engage the AI as you would a human in conversation and get pretty decent results (“Create instruction on the life cycle of a butterfly.”). Wharton College Professor and author of Co-intelligence, Ethan Mollick, says that ensuring you give the AI context can greatly improve its output.

Tip: be specific about the audience, tone, and format to get tailored outputs.

Request (prompt) – Example

  • Request (prompt): “From a behaviorist perspective, identify measurable learning objectives for a 4th-grade lesson on the butterfly life cycle. Ensure the objectives focus on observable behaviors (e.g., naming stages or sequencing steps), use language appropriate for 4th graders (audience), and are formatted in a clear, student-friendly tone (tone) suitable for inclusion in a lesson plan or presentation (format).”

Note: Read more about prompts at Ethan’s blog, One Useful Thing,

A – Assess

This step ensures you continue to exert your critical thinking skills rather than relinquish it to AI.

Once you’ve submitted the prompt request to the AI, the AI will give you an output. This is where the work really begins for you, because AI, as we have heard in the AI Basics video above (and maybe experienced for yourself), AI is an expert liar.  Review the output against what you know regarding the theory you are asked to apply or the instructional design principles you know. Ask yourself:

  1. Is this accurate?
  2. Is it concise or will I need to cut a bunch of this?
  3. Is it appropriate for my audience?
  4. What’s missing? (and sometimes “who is missing” in the case of AI bias)
  5. What’s unclear?

Example:

AI output:

Here’s a list of measurable learning objectives, keeping the audience, tone, and format appropriate for a 4th-grade lesson:

Measurable Learning Objectives

  1. Naming Stages:
    • “Students will be able to name the four stages of a butterfly’s life cycle (egg, caterpillar, chrysalis, butterfly) with 100% accuracy after completing a matching activity.”
  2. Sequencing Steps:
    • “Students will correctly sequence the stages of the butterfly life cycle on a worksheet, scoring at least 4 out of 4 on the activity.”
  3. Identifying Characteristics:
    • “Students will identify one key characteristic of each stage of the butterfly life cycle in a group discussion, contributing at least one observation.”
  1. Appreciating Nature:
    • “Students will develop an appreciation for the beauty of nature and the importance of butterflies in the environment.”

Assess: In reviewing the above output, the bolded action verbs reveal an inaccuracy. “Develop an appreciation” is not a measurable learning objective. I may want it more concise, and therefore want to remove the four steps (Naming Stages, Sequencing Steps, etc.).

F – Frame

This step ensures you continue to maintain your creativity, voice, and human intuition.

Regardless of what you prompt the AI to generate for you, after analyzing it, you must put your own spin on it. Why? Because it’s too easy to offload that process to a machine, and if we do, your brain, your imagination, your voice, your gut all are silenced. The world will miss out on your unique viewpoint, vision, and voice, and that’s the unique value you bring to the table as an instructional designer. It’s also not ethical to pass off work as yours if only the AI did the work. So, in this step, you’ll act on your analysis by making the changes needed and putting it in your voice.

Example

Reframe the output:

Measurable Learning Objectives

    • Students will be able to identify the four stages of a butterfly’s life cycle (egg, caterpillar, chrysalis, butterfly).
    • Students will correctly sequence the stages of the butterfly life cycle on a worksheet.
    • Students will identify one key characteristic of each stage of the butterfly life cycle.
    • Students will state the importance of butterflies in the environment.

Reframe: Based on the analysis, I adapted the one unmeasurable learning objective to make it measurable, and I made the objectives more concise by removing the four steps. Based on my intuition and experience, I knew fourth grade teachers might need some flexibility in what activities they actually end up using in a classroom to assess these objectives. So, I removed the specific details from the objectives, too.

Note: If the AI output has major issues, the Reframe step may call for further prompting before you make your changes and add your voice, but always put your spin on the output.

T – Think

This step (potentially) is what can make you a leader and expert.

Sometimes as instructional designers our work is very task-driven (“design a course in Canvas”, “run a task analysis”, “create an instructional video”, etc.), and there is nothing wrong with that. However, becoming an expert or leader in this field requires nearly as much time thinking about and reflecting on what we are doing, the tools we are doing it with, and why we are doing it, as actually doing the tasks. Specifically with AI, which brings so many new changes and concerns, we need to build in the time to reflect on our own use of it. For the Think step, stop the “doing” for a few minutes and consider by answering the following questions:

  1. How did using AI in your instructional design process affect your confidence in your abilities as an instructional designer?
  2. How did AI enhance or limit your creativity when designing instruction?
  3. How did AI’s involvement support or challenge the human-centered aspect of your design, such as tailoring to different learners’ needs
  4. What else did you learn during this process?

Reflect – Example

Think: Using AI in my instructional design process made me reflect on my own abilities. At times, I felt more confident when AI provided helpful suggestions, but I also questioned whether I was relying on it too much. AI gave me ideas I hadn’t considered, but some felt too generic, so I had to adjust them. It was useful for structuring content, but I wasn’t sure if it fully accounted for different learners’ needs. This process made me think more critically about AI’s role in instructional design.

Note:  Writing these reflections and reading the reflections of others, alongside the experience of using AI, will give you vocabulary and use cases to discuss with others at conferences, at work, or with your classmates. Using daily reflections on all you are learning can greatly help guide you to dissertation topics and their underlying theories and frameworks. I highly recommend it as a life practice!

A Completed Assignment

Each assignment that calls for using the CRAFT model in this course is a little bit different, but based on the assignment in the examples aboveuse the arrow below to see what a completed CRAFT assignment would look like:

 

Completed CRAFT Assignment

Completed CRAFT

Clarify

Using ChatGPT, I need to create instruction for a fourth grade teacher who is a behaviorist on the topic of the life cycle of a butterfly.

Request

(prompt): “From a behaviorist perspective, identify measurable learning objectives for a 4th-grade lesson on the butterfly life cycle. Ensure the objectives focus on observable behaviors (e.g., naming stages or sequencing steps), use language appropriate for 4th graders (audience), and are formatted in a clear, student-friendly tone (tone) suitable for inclusion in a lesson plan or presentation (format).”

AI Output:

Here’s a list of measurable learning objectives, keeping the audience, tone, and format appropriate for a 4th-grade lesson:

Measurable Learning Objectives

      1. Naming Stages:
        • “Students will be able to name the four stages of a butterfly’s life cycle (egg, caterpillar, chrysalis, butterfly) with 100% accuracy after completing a matching activity.”
      2. Sequencing Steps:
        • “Students will correctly sequence the stages of the butterfly life cycle on a worksheet, scoring at least 4 out of 4 on the activity.”
      3. Identifying Characteristics:
        • “Students will identify one key characteristic of each stage of the butterfly life cycle in a group discussion, contributing at least one observation.”
      1. Appreciating Nature:
        • “Students will develop an appreciation for the beauty of nature and the importance of butterflies in the environment.”

Assess

In reviewing the above output, the bolded action verbs reveal an inaccuracy. “Develop an appreciation” is not a measurable learning objective. I may want it more concise, and therefore want to remove the four steps (Naming Stages, Sequencing Steps, etc.).

Adapted AI Output:

Measurable Learning Objectives

        • Students will be able to identify the four stages of a butterfly’s life cycle (egg, caterpillar, chrysalis, butterfly).
        • Students will correctly sequence the stages of the butterfly life cycle on a worksheet.
        • Students will identify one key characteristic of each stage of the butterfly life cycle.
        • Students will state the importance of butterflies in the environment.

Frame

Based on the analysis, I adapted the one unmeasurable learning objective to make it measurable, and I made the objectives more concise by removing the four steps. Based on my intuition and experience, I knew fourth grade teachers might need some flexibility in what activities they actually end up using in a classroom to assess these objectives. So, I removed the specific details from the objectives, too.

 

Think

Using AI in my instructional design process made me reflect on my own abilities. At times, I felt more confident when AI provided helpful suggestions, but I also questioned whether I was relying on it too much. AI gave me ideas I hadn’t considered, but some felt too generic, so I had to adjust them. It was useful for structuring content, but I wasn’t sure if it fully accounted for different learners’ needs. This process made me think more critically about AI’s role in instructional design.

 

AI and CRAFT Discussion Strategy Check (Not Graded)

That was a lot of information. Test yourself before jumping into this week’s assignments.

 

References

Breakstone, J., McGrew, S., Smith, M., Ortega, T., & Wineburg, S. (2018). Why we need a new approach to teaching digital literacy. Phi Delta Kappan, 99(6), 27-32.

Curry, J. H., & Cook, J. (2014). Facilitating online discussions at a manic pace: A new strategy for an old problem. Quarterly Review of Distance Education, 15(3), 1.

Mollick, E., & Mollick, E. (2024). Co-Intelligence. Random House UK.

Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.

Ragnedda, M., & Muschert, G. W. (2013). The digital divide. Florence, KY: Routledge.

License and Attribution

“AI and the CRAFT Discussion Strategy” by Theresa Huff is licensed under CC BY 4.0.

 

License

Icon for the Creative Commons Attribution 4.0 International License

Design in Progress: A Collaborative Text on Learning Theories Copyright © by Theresa Huff is licensed under a Creative Commons Attribution 4.0 International License, except where otherwise noted.

Share This Book