Perspectives on AI

A variety of perspectives were shared during our conversation with TRU faculty members at our “Artificial Intelligence: Questions and Guidance” session on May 17, 2023:

Overall, we advocate for an open, transparent approach to addressing the changes that generative AI is making in our disciplines. As you determine where you fall along the continuum of AI use, we offer some classroom activities and conversation starters to explore with your learners. If you have an AI activity you’d like to share with us, please email learningtech@tru.ca.

Have Conversations with Students about AI

Having conversations about AI in the classroom is beneficial to gather insight and share perspectives. By talking to students about AI, it helps you understand what students think about it and how they may use it. It also helps students understand what you think about it and how you might use it. These kinds of conversations build trust and transparency between you and your students.

Below are some guiding actions and conversation starters you may be interested in using with your students.

Students will have different levels of awareness when it comes to generative AI. Some guiding questions for conversations with your students may include:

  • What do you know about artificial intelligence?
  • What AI app(s) do you use? 
  • What value does this app(s) provide you?
  • Why did you choose to use them?
  • What are your concerns about using AI?
  • How can you use AI to support and show your learning?
  • When is it inappropriate to use AI?
  • How do you think AI could be integrated into the classroom or discipline area?

Sharing your perspective with students will let them know how you feel about AI and what your expectations are for the class. The following prompts can be used to help develop your perspective:

  • Take a moment to consider your general perspective about AI use. What do you think about Artificial Intelligence? Do you use AI tools in your own practice? Do you think AI has a place in the classroom? Why or why not?
  • Explore AI apps that your students may use to help develop your awareness of what tools exist.
  • Consider how AI is impacting your academic field and/or the profession that students are pursuing. Is it suitable to integrate AI into your classroom to support future career realities students will face? 
  • Reflect upon your existing assessments from the perspective of artificial intelligence. Evaluate the current alignment between your course learning outcomes and your assessments, while also contemplating the influence of artificial intelligence applications.
    • Which types of assessments demonstrate the strongest alignment with your course learning outcomes? How does AI impact these assessment methods?
    • Which types of assessments demonstrate the weakest alignment with your course learning outcomes? Is there an opportunity to consider alternative assessment methods as a way to enhance alignment and proactively consider the impact of AI?
    • Can you make minor modifications to your overall course design to proactively consider the impact of artificial intelligence applications?

The “generative” in generative AI means that the tool will make things up if it cannot access an answer to a question. This means that generative AI can perpetrate misinformation and generate false citations. Demonstrating this in class can be useful for students.

  • If you have a presence online prior to 2019, try asking generative AI to create a biography for you. Go through it with your students to explore what is correct and incorrect.
  • Use an assignment question as a prompt and work together as a class to fact-check the output.
  • Ask generative AI to generate a list of citations on the class topic, and then have students try finding the original sources.

If you have a clear rubric for your assignments, it can help students to see that generative AI is missing much of the classroom context required to do well on a particular assessment. Consider creating an AI-generated response of your own assignment and then marking it as a group against your rubric. This can also be helpful for students to understand the rubric and how it is applied.

By allowing students to be part of the conversation regarding the use of AI in your classroom, this will encourage them to share how AI could be used to show their learning in a responsible way. Some ideas to consider are:

  • Collaboratively establish class guidelines with your students regarding the suitable and unsuitable use of artificial intelligence apps in your courses.
  • Jointly define expectations for citing, referencing, or acknowledging the incorporation of artificial intelligence applications in course work and assessments. These expectations may also be included in assignment descriptions, rubrics, and syllabi as a reminder.
  • Co-develop rubric criteria that will be used to assess assignments, focusing on skills and actions that students should be showcasing
  • Create a space to discuss AI throughout the semester. For example, this could be done asynchronously by using a Forum in Moodle.

Strategies for Minimizing the Use of Generative AI in Assessments 

Concerns about students using AI to fully complete their assignments are real. However, the uncertainty over who wrote a student’s paper has always existed – was it written by a parent, a sibling, or was it contracted out? If we spend the majority of our time determining if a computer has completed student work, it reduces our time and capacity to focus on creating meaningful learning experiences for our students. Rather than worrying about how students may use AI to cheat, perhaps a more productive question to consider is: how can we best educate our students and design assessments that are authentic, relevant, and meaningful? 

Here are a few strategies for revising assessments to encourage student creation and demonstration of knowledge which in turn reduces the usefulness of AI tools.

Although AI has the potential to be a helpful tool, it is crucial for students to understand that relying on AI as a substitute for their own work is not acceptable. 

If you are not allowing students to use AI in your class be sure to explain why:

  • It can be considered cheating because they are misrepresenting their knowledge by employing unauthorized devices or aids.
  • It can be considered plagiarism because if students submit work as their own creation, even if it was generated by an AI tool, it means presenting the ideas of others, even if those “others” are not human beings.
  • Students may be submitting fabricated or false information as AI tools sometimes make up information including references and sources.
  • Students may be infringing on copyright. For example, using a text-generating AI tool like ChatGPT to compose a song resembling a particular artists’ song or using an image-generating AI tool like DALL-E to produce an image in the style of a contemporary artist, could potentially violate copyright laws. AI tools rely on existing works and may generate derivatives that infringe upon intellectual property rights.

Consider replacing assessments like short answer questions and essays with different kinds of assessments, such as:

  • Multimedia (video, podcast) presentations: Ask students to create multimedia submissions (audio and/or video). This requires conceptualization and technical skills in order to put together a clear and concise product. Also, it encourages students to demonstrate their knowledge in a manner that is most suitable for them. Kaltura is a TRU supported tool that allows students to create and host audio and video content that can be integrated with Moodle assignments for seamless submission. 
  • Mind maps: Ask students to create a mind map that illustrates how they conceptualize a certain topic and/or how they see relationships between topics. This encourages more holistic and wide-reaching thinking.
  • Debates: Ask students to debate a key question or challenge in your discipline. This encourages students to look at topics from varied perspectives and communicate their ideas clearly. You can decide whether you want students to use a tool such as ChatGPT to help them prepare for their opening statement or to gain ideas of possible counter arguments they can prepare for. 
  • Oral presentations: Ask students to share what they learned with the class. By encouraging class-wide dialogue and discussion after the presentation, this prompts students to know their material well so they are prepared to answer follow up questions. 
  • Portfolios: Ask students to curate their learning and showcase it in a digital portfolio. TRUBOX is TRU’s locally-hosted instance of WordPress where customized portfolio templates can be created for your classroom. Connect with the Learning Tech team to learn about the support provided with website development projects (learningtech@tru.ca).

Scaffolding an assignment and breaking it into smaller parts helps to minimize student stress and pressure. Also, it provides multiple opportunities for self reflection and feedback from peers and/or the instructor. 

Create assessments that allow students to develop products and ideas over time, with checkpoints integrated throughout the assignment. Examples of checkpoints may include:

  • Creating an outline
  • Drafting a proposal
  • First draft of a project
  • Revised drafts of a project
  • Annotated bibliography
  • Documentation of the research process
  • Reflective journals and/or logs pertaining to the assignment
  • Asking process-oriented questions for students to answer

Tools like ChatGPT do not have access to information or events beyond September 2021, nor cannot pull information from specific sources (yet). So, to lessen the appeal of using artificial intelligence, ask students to draw from, incorporate ideas/evidence from, make connections to, or reflect upon sources such as: 

  • Course materials: Course textbooks, readings/videos in Moodle, class discussions (online or in-person), etc.
  • Current events: Recently published studies, current events in your discipline, current events from around the world, etc.
  • Specific sources: Articles, textbooks, videos, etc.
  • Physical artifacts brought into the classroom

Provide students with opportunities to reflect and draw connections between their personal experiences and what they are learning. This makes the assessment relevant which is more motivating for students. The University of Cambridge describes a few models for reflection:

  • ERA Cycle (experience, reflection, action)
  • Driscoll’s What Model (What? So what? Now what?)
  • Kolb’s Experiential Learning Cycle (Concrete experience, reflective observation, abstract conceptualization, active experimentation)
  • Gribb’s Reflective Cycle (Description, feelings, evaluation, analysis, conclusion, action plan)

Ask students to complete certain work during class time. For example, utilize entrance or exit tickets that capture student reflections about material learned. This strategy can also help students prepare for activities that build upon this, such as group work or discussions.

Strategies for Integrating Generative AI into Assessments 

Providing students with opportunities to use AI tools in the classroom can help them learn how to use them in safe, ethical, and constructive manners, as they relate to their field of study. Below you will find a few strategies for integrating generative AI into activities and assessments.

However, a few cautions to consider before engaging with or assigning AI tools:

  • Please note that none of these tools have been through a privacy impact assessment. Do not put student data, including names, email addresses, and student numbers, through any artificial intelligence tool. This includes the use of AI detection tools.
  • Generative AI tools may not work with screen readers and other adaptive technologies. Beware of limiting who can engage in the classroom experience with you.
  • Remember that there are significant ethical issues in using AI tools, including issues of bias, climate harm, and copyright. Deciding when it is appropriate to choose to engage with AI is an important part of developing a responsible relationship with generative AI.

Be as clear as possible when describing how AI can be used in an assessment. This will help students understand how AI can be used ethically and responsibly in their work. For example, ask yourself questions such as:

  • What is considered appropriate use of AI in this assignment?
  • What is considered inappropriate use of AI in this assignment? Why?
  • Do students have to acknowledge their use of AI in this assignment? If so, how? For example, TRU’s Artificial Intelligence: A Guide for Students page shares guidance on how students can acknowledge the use of AI and how they can add citations.
  • Will you allow any AI tool or will you provide one or more approved AI tools? 
  • Will you utilize generative AI as a study aid? If so, will you allow your students to upload content you created to generate practice quizzes and questions?
  • If you think students will ask AI to explain theories and concepts, make sure they know that they need to evaluate the generated content for things such as accuracy, bias, and comprehensiveness. 

Once expectations are determined, share them in assignment descriptions, rubrics, and/or course outlines. Keep in mind, involving students in the co-creation of guidelines for AI use in assessments can be a powerful and meaningful experience. 

Model the safe and ethical use of artificial intelligence by being clear and transparent about how it can be used for learning. Demonstrate effective practices when using artificial intelligence with students. For example:

  • Describe how you determined which AI tool you used, mentioning consideration of ethical concerns such as environmental impact or privacy and data collection
  • Describe how you used the AI tool (to summarize text, create an outline, generate content, etc.)
  • Describe how you evaluated and fact-checked the content AI generated, mentioning consideration of inherent inaccuracies and bias
  • Describe the overall experience with the tool and whether it helped or hindered your learning
  • Describe how your experience with the tool aligned/misaligned with the scholarly standards of your field

When used thoughtfully and carefully, generative AI tools may help students engage in higher order thinking skills. For example, have an AI tool generate a response to a question/prompt and have students (individually, in pairs, or in small groups): 

  • Analyze the strengths and weaknesses of the response (What did it capture well? What did it miss?)
  • Fact check the response and identify incorrect information (What aspects may be inaccurate? What would need to be revised and improved?)
  • Evaluate the response for logic, consistency, accuracy, and bias (Does the response make sense? Is the information accurate? Is the information inclusive?)

To utilize generative AI in the classroom in a low-stakes and casual way, you may consider using it for formative or ungraded assessments during class time. York University shares a number of examples:

  • Use generative AI to create a starting point for class discussion on a specific topic. Evaluate it as a class by discussing what it gets right, what it’s missing, and how it would need to be revised.
  • Have small groups of students explore AI tools to generate content on a specified topic. Then, have them compare the results and/or the process (What grade would they assign the response using a course rubric? What prompts and adjustments were needed to generate the response?)
  • Engage in a class debate and use generative AI to create counter arguments. This helps students evaluate different perspectives and strengthen their own arguments.

Having students revise the work generated by a machine may be less intimidating than reviewing the work of a peer. The following activities are examples of ways AI tools can be used in small group activities: 

  • Revision: Have students edit a piece of text generated by AI – expand upon the paragraphs, combine or separate sentences, add supporting evidence, or rewrite introductions/conclusions.
  • Prompt Competition: Share a major question or challenge in your field (something without a single answer) that generative AI could create a response to. Have students determine the criteria that will be used to assess the AI response (e.g. accuracy of content, multiple perspectives, etc.). Then, have students (individually or in small groups) write their prompts that will be used to generate a response. After the prompts generate their responses, students evaluate the responses based on the criteria they determined.

Did you know?

TRU Library created Artificial Intelligence: A Guide for Students which covers how Artificial Intelligence (AI) may be used responsibly and ethically in student work, how to critically evaluate content created by generative AI (e.g., ChatGPT), and how to acknowledge and cite AI.

References

AI technology and academic integrity. Academic Integrity. (2023, June 14). https://www.yorku.ca/unit/vpacad/academic-integrity/ai-technology-and-academic-integrity/

Artificial Intelligence in higher education. Teaching & Learning – Trent University. (n.d.). https://www.trentu.ca/teaching/teaching-resources/artificial-intelligence-higher-education

How do I consider the impact of AI tools like CHATGPT in my courses?. Center for Teaching & Learning : UMass Amherst. (n.d.). https://www.umass.edu/ctl/how-do-i-consider-impact-ai-tools-chatgpt-my-courses

How to make your assessment more ai-proof. UvA Teaching and Learning Centres (TLC). (2023, May 9). https://tlc.uva.nl/en/article/how-to-make-your-assessment-more-ai-proof/

Prochaska, E. (2023, January 11). Embrace the bot: Designing writing assignments in the face of ai. Faculty Focus | Higher Ed Teaching & Learning. https://www.facultyfocus.com/articles/course-design-ideas/embrace-the-bot-designing-writing-assignments-in-the-face-of-ai/?st=FFdaily%3Bsc%3DFF230123%3Butm_term%3DFF230123&mailingID=4402

Teaching and learning with Artificial Intelligence Apps. Taylor Institute for Teaching and Learning. (2023, January). https://taylorinstitute.ucalgary.ca/teaching-with-AI-apps?fbclid=IwAR3D9cEnKf8cmQVZgxI5zfo7hxGzJA2GDJTxcmWNZQAyySq8RUdVIFUuOb4

Valuing process as equal to, or greater than, product. Valuing Process as Equal To, or Greater Than, Product | Center for Teaching & Learning. (n.d.). https://teaching.berkeley.edu/valuing-process-equal-or-greater-product

Watkins, R. (2022, December 19). Update your course syllabus for chatgpt. Medium. https://medium.com/@rwatkins_7167/updating-your-course-syllabus-for-chatgpt-965f4b57b003