AI and Course Design

 

Guidelines for using AI in your course

Instructors and students often ask: "What is the AI policy at the U of A?" 

AI is a flexible technology that is applied in different ways, depending on the context. A policy that would be sensible in one situation could be restrictive or damaging in another. 

Therefore, the U of A has developed a framework and guidelines for the use of AI. Anyone using AI in their teaching, learning or research is responsible for evaluating the risks and benefits for their situation. They must also act in alignment with the university's core strategic commitments to innovation, equity, sustainability, and reconciliation.

The Framework for the Responsible Use of AI at the University of Alberta describes six principles for the responsible use of AI in learning, teaching, research, administration, and other university-related work. It seeks to build trust among those who will be impacted by our use of AI. The AI Data Safety Guidelines for the University of Alberta provide guidance on privacy and data protection for the use of AI tools.

---

AI and backwards design

The fact that a technology makes something is possible does not guarantee that it is useful. GenAI is remarkably flexible and can emulate or mediate some aspects of teaching and learning, but it is up to instructors to evaluate whether the possible is indeed useful in the context of the course or program of study.

"Backwards design" can help clarify whether and where GenAI could bring value to students in your course. In most cases, a blanket yes-or-no policy about AI is insufficient. GenAI is too flexible, and teaching and learning is too dynamic an environment for an all-or-nothing AI policy.

Since Wiggins & McTighe described it in the mid-1990s, "backwards design" has become a go-to method for building effective courses in postsecondary education. The core idea is simple: instead of starting with a long list of content and seeing how far you can get over the term, you design the journey by first identifying the learning outcomes that you want your students to reach.

• Learning outcomes: Your course learning outcomes cover the knowledge, skills and attitudes students need to develop. Backwards design begins by asking what a student needs to know, what they need to be able to do, and what values or ethics they should develop by the end of your course.
• Design assessments: Once you know the destination, the first step "backwards" is to decide how you will measure your students' progress. These generally fall into two assessment categories.
  • Formative assessments: low-stakes activities where students can practice and receive helpful feedback.
  • Summative assessments: formal ways you evaluate learning, such as exams or final papers
• Plan teaching and learning activities: Plan specific activities to help your students succeed on those assessments.vary depending on your field, and they typically include a mix of independent work, such as readings and reflection, in-class activities such as lectures and participatory activities, and hands-on experiential learning.

---

Where does AI belong in my course?

AI is sometimes viewed as an enemy of teaching and learning because it can mimic the competencies you are trying to help students develop. This worry is legitimate, but it has led some instructors to simply forbid their students from using artificial intelligence entirely.

Before you implement a "no AI" policy, consider these potential limitations:

• Standard tools: AI is increasingly integrated into the standard tools students use for coursework, such as word processors, research platforms and search engines.

• Professional preparation: In many fields, artificial intelligence is already embedded in the specific workflows that students need to master to become competent in their disciplines.

• Inclusive learning: Some neurodivergent students and those facing language barriers rely on support from AI tools. A strict ban may disproportionately affect these learners.

• Unintended consequences: While it may be simple to forbid the use of artificial intelligence, it is difficult to detect or prove whether a student has used it. Prohibiting it might simply drive the practice underground, leaving students to use these tools without your guidance.

Backwards design is a more productive place to start when deciding about AI in your course. Begin with your learning outcomes and work backwards through assessments and teaching and learning activities, considering ways that AI might support or undermine a student's learning, at various points along the way.

---

AI as a course design partner

Course design is complex, creative work. If you ever feel stuck or want to try something new, AI can act as a versatile creative partner to explore and evaluate ideas alongside you.

Current versions of GenAI models (like Google Gemini) are surprisingly well-informed about the knowledge, skills, and ethical considerations in many fields. What this means for you, as an instructor, is that you can engage with AI as though it were an experienced colleague.

A couple of cautions, before you engage AI as a course-design partner:

• Approach every AI output with skepticism. It is up to you to validate and apply your critical judgment.
• Model transparency. If you have used AI to brainstorm or plan a learning activity, share this with your students. In other words, model the transparency you want back from your students.

Sample course design prompts

Some sample course-design prompts are given below. These can be cut-and-paste straight into Gemini. See the section called Interacting and Building with AI if you're unsure how to do that.

Notice in each sample prompt that Gemini is prompted with a specific role (instructional design partner), goal (help me...), sequenced process, and output format. You can adapt this prompting approach to other aspects of course design where you would benefit from a design partner.