As I’ve previously discussed on my blog, I am one of the nodes in the network of faculty, staff, and student support helping our campus community navigate the implications of AI on teaching and learning. Salem State University’s Center for Teaching Innovation continues to support folks throughout the summer, so as part of a “Summer @ CTI” June initiative I suggested a session that helped instructors who may be feeling overwhelmed by the wide range of tools, resources, and debates related to AI in higher education. We have previously used checklists in other professional development contexts, since they tend to work well as starting points that we can use to anchor live programming and then disseminate asynchronously to folks who can’t make live sessions.
I came up with what we are calling ADAPT, a Generative AI Course Design Checklist.
The bare-bones checklist you see at the top of this post is designed as a pared-down starting point, but it’s accompanied by additional questions for folks to consider as they review upcoming courses and plan relevant materials.
Why ADAPT?
- AI is a big and disruptive topic; it can be hard to know where to start in terms of course design
- ADAPT is informed by SSU AI Advisory Group outreach and conversations re: immediate needs across classrooms here; soliciting feedback and perspectives across campus is an important component of resource development (and some of the concerns informing our checklist may align with your own)
- For example, our office places a strong emphasis on the value of “AI literacy” rather than “AI evangelism”; many of our students, faculty, and staff wish to know more about this topic but also have concerns related to climate impact, copyright, misinformation, bias (among others)
- ADAPT is a framework to get started and a recommended approach; it’s designed to help instructors self-assess their emerging AI literacies and plan next steps related to their more immediate course design priorities
AI Literacy
“AI Literacy” is a frequently deployed phrase in teaching and learning contexts, but it can be helpful to consider more specifically what “AI literacy” might entail.
For example, in reviewing resources on this topic, I came across a 2024 framework developed by Michelle Kassoria, Maya Georgieva, and Allison Papini published by EDUCAUSE:
I found these four areas to be helpful, but in talking with our faculty I decided to rephrase them into more colloquial language (while also sharing their point of origin) and de-emphasize certain elements (for example, many of our faculty are still weighing the benefits of application, integration, and implementation):
- Technical: What are large language models? What are chatbots? What are hallucinations?
- Evaluative: Can we slow down and test some of this stuff out more? Do I have an idea for what I’d like to test, and why?
- Practical: Am I OK with students using some of these tools in specific course contexts? What do they need to know? Who else is using this stuff in their classrooms?
- Ethical: Should I be using this stuff myself? Should I have information to share with students about what not to do?
More specifically, I shared some particular contexts and lines of inquiry that we have found helpful in our own emerging AI literacy work:
- Moving from “AI” in the abstract to considerations of specific AI tools and use-cases in our classrooms and fields
- Identifying “ethical implications” in specific contexts and in relation to core community values (for example, SSU’s HSI-MSI status)
- Developing learning outcomes that address the impact of AI tools and use-cases by students, faculty, and staff
- Reviewing specific implications for higher ed (teaching, assessment, emerging K-12 classroom AI use-cases)
- Accepting that “AI literacy” is ongoing and what it entails will change as new tools, use-cases, and perceptions continue to emerge and change
What might AI literacy look like across campus and beyond an individual course? I shared some suggestions and ideas and referenced some ongoing work we are involved with here at the CTI, including:
- Assessment of AI syllabus language and course policies
- AI’s impact on student support/student success work
- AI and Academic Integrity resources for faculty, staff, and students that prioritize equity, encourage AI literacy, assess impact on research/writing
- Department-specific support and dialogue on AI’s impact
- AI’s impact on student professional development/experiential learning
- Transparency and piloting in potential administrative uses of AI
AI and Assessment Design/Redesign
In introducing ADAPT, we spent a bit of time discussing the particular challenges instructors face regarding assessment design.
We started with a consideration of framings that recommend “AI-proof” or “AI-resistant” assessment activities, with the following points of emphasis:
- “AI-Proof” language may be useful when searching for assignments to model, but most assignments can be completed/aided by generative AI tools accessible to students.
- “AI-Resistant” assignment recommendations acknowledge this reality. That being said, consider whether your assignment is “AI-Resistant” at the expense of teaching and learning.
- For example, consider the negative pedagogical implications of handwritten major assessment activities, surveillance technology, and onerous submission guidelines on student success.
- If an AI tool can be used by a student to outsource and complete an assignment, that doesn’t necessarily mean that you shouldn’t assign that assessment activity! But you may want to be transparent and direct with students about these possibilities, their limitations, and their impact on learning
I shared some general recommendations with instructors about assessment after AI:
- Start by explaining the value of completing this assignment within your desired parameters (with or without AI). Encourage questions.
- Scaffolded assignments that use small assessment activities/in-class work may discourage some AI usage beyond your course policies; the value of working iteratively/incrementally should be made clear to students.
- Some instructors have increased usage of archival materials/special collections and field work/experiential learning productively.
- Some instructors have used multimodal/presentation outputs productively.
- Rubrics that anticipate the quality/tendencies of AI outputs may also discourage some AI usage beyond course policies.
I also noted that instructors who have encouraged or modeled use of AI tools in assessment activities tend to:
- Emphasize use of specific, vetted tools (at SSU, instructors and students can access Copilot/Gemini licenses that have been vetted by our ITS department for use)
- Identify specific points in assignment workflows to introduce/consider AI tools (brainstorming, copyediting, code review, audience/genre activities)
- Model genres/use-cases from specific professional contexts relevant to the course
- Incorporate AI tools in smaller assessment or ungraded assessment activities
- Incorporate self-reflection requirements to help students assess the experience of using AI
For interested faculty, I also suggested a self-assessment approach for assessment activities that may be impacted by AI use-cases:
- Use SSU’s Copilot or Gemini accounts (or your own vetted AI tools) to gauge what an AI output generated by a student might look like.
- Reflect on the quality (or lack of quality) of the output. How would you discuss this use-case with students? What is lost from outsourcing this work to AI?
- What might be gained from incorporating AI into some component of the current assignment or a revised assignment? How would you discuss this use-case with students? Why might you still refrain from allowing this use-case?
- How would a current or revised rubric help you assess this assignment and guide students towards what they should prioritize in this work?
- How might you revise this activity and still maintain the course outcomes associated with it?
Next Steps
ADAPT was introduced in a 90-minute virtual workshop (attended by around 20 or so instructors and staff). Attendees were compensated for their attendance. Before the session, they were asked to have a syllabus (or notes for a new course) in mind, as well as at least one major assessment activity. They were also asked to send an informal follow-up message (around 250 words in length) via email to reflect on the workshop and identify/discuss relevant areas of interest and concern.
The ADAPT checklist was designed to function as an asynchronous resource due to available time for professional development, instructor preferences, and to acknowledge that it is summer and folks may have other priorities. But we do really see it as a starting point that will ideally lead to additional conversations and discussions with faculty support sectors like CTI and with department/cross-campus colleagues.
ADAPT is one of several support initiatives we are providing to our faculty to help navigate the implications of AI. I’ll try to share other work and efforts if they seem useful in contexts beyond our own institution.
As I note in the checklist document, I am happy to talk more. I am sharing more of my AI work here because I have been encouraged to do so by colleagues and because I am hoping to model some AI literacy efforts that do not begin or end with AI adoption as an inevitability.
Note: The views expressed in this post may not be shared by my employer or all of my colleagues and should not be regarded as an institutional or departmental perspective. If you’d like to talk more, feel free to email me at jimmcgrath[dot]us[at]gmail