Mike Frazier, Michigan State University
Lauren Hensley, Ohio Wesleyan University
This assignment proposes the integration of generative AI tools, such as ChatGPT, into a college learning and motivation strategies course, with the dual focus of enhancing metacognition and promoting ethical AI use. Students engage with AI-generated artifacts, compare outputs with their own work, and reflect on the implications of AI in their academic and professional lives. The approach can be adapted for various courses, encouraging a critical examination of AI's role in learning and its potential impact on future careers.
Learning Goals:
Original Assignment Context: middle of elective college learning and motivation strategies course
Materials Needed: an accessible AI text generation program (i.e. ChatGPT)
Time Frame: 1 class session
Introduction
Technologies like generative chatbots provide students with opportunities to evaluate outputs generated by AI as part of the metacognitive process of thinking about their learning and funds of knowledge. Using AI tools in instruction also provides important opportunities to have discussions with students related to ethical and responsible use of these tools, which is appropriate and important to facilitate in contexts designed to support students’ success in college and beyond. In this submission, we provide a model of scaffolding metacognition based on a multi-phase project in our college learning and motivation strategies course. In the remainder of this submission, we will describe the traditional version of the course project in our college learning strategies course, followed by a detailed description of the changes made and reasons for these changes, so that others could use this assignment–or its approach–themselves. Finally, we will provide a sample activity from a lesson that asks students to analyze AI-generated output in order to increase metacognition on an academic task.
The Traditional Assignment
The course for which we have adapted a project–Learning and Motivation Strategies for College Success–is a three-credit elective at a large public university. The traditional version of the project involved a series of scaffolded assignments leading up to a final product. Students:
With the recent release of OpenAI’s ChatGPT, we found that, with the help of AI, we were able to complete all the above tasks almost instantly, with convincing original, specific outputs from ChatGPT that were on-par with what we would have previously considered average to above-average student work. This discovery motivated us to revise the course project with the goal of promoting metacognition through students’ engagement with AI. We focused on metacognition in particular because students had been discussing the complexities of self-regulation and metacognition leading up to this lesson, which makes the integration of a generative AI tool at this point a timely and relevant activity to help students think about their thinking. By testing out an AI tool that appears to “think” differently from them, they are able to better identify and navigate the ways in which they do and do not think in order to improve and apply their understanding of theories from educational psychology and strategies behind learning and motivation.
Model for Adapting the Project
In this section, we describe how we integrated a GPT-3.5 powered tool (ChatGPT) into the course project for our course on learning and motivation strategies, and provide the rationale for what we changed. In adapting the project, we used the course on college success as a fitting context for open and honest conversations with students about ethical uses of AI in learning and instruction. The main talking point throughout these conversations is that as human authors, we need to be utilizing any new writing tools ethically and responsibly to elevate the critical and creative thinking that we do if we want to advance in our respective fields. Having these conversations in an educational space with a figure they trust, such as a caring and non-judgmental instructor, is essential to making thoughtful decisions related to academic integrity. Throughout each of the learning tasks associated with the course project, we gradually integrated generative AI tools, with each task asking students to incorporate and reflect on generated AI artifacts. This gradual approach allowed students time and space to consider the implications for themselves as writers and as future practitioners in their fields, while also allowing them to gradually build skills related to AI, which is quickly emerging as a potential essential literacy in our world. We modeled for students appropriate and ethical uses of AI by (1) completing the task independently, (2) exploring an AI tool relevant to the task, (3) engaging in discussion and reflection to compare the human and the AI outputs, and (4) discussing the broader implications of AI on our lives and the importance of ethical AI literacy. This final step emphasized thinking about the structures and conventions of writing in particular contexts, the veracity of the AI output, the appropriateness of the types of detail and depth generated by AI, expanding or condensing ideas, and potential biases in our systems and conventions that AI brings to light. As this was the first time many students were witnessing ChatGPT’s capabilities, they expressed a range of emotions, from fascination with the technology to potential anxiety about its impact on their future careers.
Model Lesson Plan from Course Project
Now we turn more specifically to this example lesson. This lesson is designed for undergraduate students in a course on learning and motivation strategies, but its method for introducing a generative AI tool and utilizing AI in instruction could be applied to any course or lesson that asks students to engage metacognitively with AI-generated artifacts. Prior to this lesson, which takes place in week three of the course, students have had a general overview of the course, learned about the science behind learning and motivation, and practiced specific techniques to improve their learning and motivation strategies applicable to other courses and areas of their lives. Students have also demonstrated an understanding of mastery-based versus performance-based goals. Finally, they have chosen a course project book from a book list that includes curated titles related to learning and memory, productivity and focus, motivation and well-being, and preparing for life beyond college. The objective of this specific lesson as it is nested within the semester-long course project is to help students develop S.M.A.R.T. goals for themselves to attain by the end of the academic semester, identifying two specific strategies from their chosen book to help them achieve those goals.
In class, students discuss in small groups what their general goals are for the semester and construct draft versions of their goals. Second, they define and identify S.M.A.R.T. goals, consider non-examples of goals that could be more specific, measurable, attainable, relevant, and/or time-bound, and revise their draft goals to make them S.M.A.R.T. Third, the instructor presents a five-step process to writing and revising goals: (1) write what you want to accomplish, (2) identify any obstacles, (3) identify resources to overcome those obstacles, (4) list your motivation to accomplish your goal, and (5) review and revise goal statements. Fourth, students are introduced to a generative AI tool such as ChatGPT. The instructor uses an interactive, exploratory, and Socratic approach to showcase the tool and some things that it is capable of related to students’ specific interests and questions. Students then attempt to prompt a generative AI tool like ChatGPT to revise their initial draft goal so that it is S.M.A.R.T. They compare the two S.M.A.R.T. goals (one created by them, one generated by AI) and discuss perceived power and limitations of generative AI tools. Finally, the instructor leads a broader discussion on some critical points related to issues like data privacy, transparency, responsible and ethical use, and critical exploration of AI tools in two contexts: (1) as a learner in academia, and (2) as a future practitioner in their individual fields. This final discussion identifies essential implications to guide students’ thinking as they engage metacognitively about what AI tools mean for their critical learning, thinking, and writing going forward.
This overarching discussion exploring with students the ethical implications of Generative AI tools like ChatGPT can be a daunting process—particularly if, like us, instructors find themselves suddenly inundated with resources, opinions, and immediate threats to the status quo of instruction in higher education. Naturally, as we work as instructors to better understand how Large Language Models work, we have an opportunity to learn alongside students, implementing AI tools through a steady, incremental process. Particularly at the university level, students have the opportunity–and perhaps the necessity—for more philosophical and ethical discussions around these tools. In order to support our students’ and our own understanding throughout these discussions, we turned to the tool itself. In Appendix A, we have included an example prompt that we gave ChatGPT (Version 3.5) that could help beginning users understand what it is and how it could help undergraduate students. Appendix B includes a more advanced list, provided by GPT-4 (currently available only to paid users when we prompted it in May, 2023), which dives deeper into the issues and implications of using generative AI tools in the field of education. All of these questions
From what we have been able to tell up to this point in our discussions with undergraduates, many of them are taking the introduction of generative AI tools into the general population very seriously; some are worried about what the jobs in their field will look like in the future. For example, in this first discussion, a third-year student in chemical engineering wanted to know how AI could impact jobs in that field, so we asked ChatGPT. The student, who had been looking at entry-level internships in the field that week, mentioned that the tasks described by ChatGPT were very similar to the job descriptions for entry-level internship positions in her field. This is one example of a talking point during the discussion that further reiterated the potential for generative AI to elevate the critical and creative thinking we do in various fields by automating certain aspects of our jobs. Generally speaking, it was a bit difficult to understand exactly how students felt during this initial session—they seemed to still be grappling with understanding exactly what ChatGPT was. It was in later discussions over the next several weeks, as we started using ChatGPT for more tasks and they read more about it in the news, that students started expressing their thoughts and concerns more often. Additionally, having regular and honest discussions with students about Generative AI tools has helped us position ourselves as figures of trust, and we have found that students often want to know how they can use these tools ethically to help them improve their lives and work. If we, as instructors, are able to incrementally identify particular ethical use cases of generative AI in education, we can incorporate these use cases in our discussions with students to help them improve their ethical AI literacy.
To conclude, navigating these ethical conversations with students requires a careful and nuanced approach; however, it seems essential to help students grapple with these difficult issues, not only to support their personal and professional growth in their potential fields, but to further support metacognitive practices to improve their learning, thinking, and work.
(Backward Design Template based upon Wiggins & McTighe, 2005)
Lesson Topic: Introducing and Creating S.M.A.R.T. Goals with AI in the Loop
For what student population and type of course is this lesson appropriate?
This lesson is designed for undergraduate students in a course on learning and motivation strategies, but its method for introducing a generative AI tool and utilizing AI in instruction could be applied to any course or lesson that asks students to engage metacognitively with AI-generated artifacts. Prior to this lesson, which takes place in week three of the course, students have had a general overview of the course, learned about the science behind learning and motivation, and practiced specific techniques to improve their learning and motivation strategies applicable to other courses and areas of their lives. They have also demonstrated an understanding of mastery-based versus performance-based goals. Finally, they have chosen a course project book from a book list that includes curated titles related to learning and memory, productivity and focus, motivation and well-being, and preparing for life beyond college.
Materials needed: (1) a projection method for the instructor; (2) individual student devices with internet connections would be highly beneficial, so that students can all access the chosen generative AI tool during the live session. Students will also need to create a free account with OpenAI (phone verification is also required).
|
|
Established Goals
What relevant goals (e.g., content standards, course or program objectives, learning outcomes) will this design address?
Essential Questions
Enduring Understandings
Key Knowledge and Skills Students Will Acquire
Students will know. . .
Students will be able to. . .
|
|
Performance Tasks
On a notecard, students will record four pieces of information as evidence of their learning:
Students will then submit their formal S.M.A.R.T. goal assignment, due the following week, which asks them to articulate the final version of their goal(s), how they arrived at those goals and plan to track them, and how they are related to the strategies discussed in their course project books. Students will also be required to be fully transparent in a reflection about any generative AI tools they used to help them write their goals.
Other Evidence
|
|
Learning Activities (~30-50 min.)
The key teaching/learning activities in numbered, chronological sequence.
Appendix A
We asked ChatGPT to “make a list of good exploratory prompts for ChatGPT that could be used for an undergraduate class to learn more about the tool itself and how it is relevant to them as individual learners with different backgrounds and interests'' (February 10, 2023). An abridged list of the questions generated by ChatGPT can be found below. This is a good starting point for ideas, but again, we would recommend a more natural, exploratory approach that includes specific questions that students in your classroom want to ask.
Appendix B
In May of 2023, we decided that, given the potential that many students are already using tools like ChatGPT, there may be a need for a more updated list of questions that delve more into deeper issues around the use of generative AI tools in educational contexts. We used GPT-4 (May 12 Version of ChatGPT) to help us revise this list of questions and received the following list of questions which are designed to go beyond a superficial introduction to ChatGPT and delve into its applications, ethical implications, and future potential. They also aim to highlight the value of using ChatGPT as an advanced learning tool in the classroom setting, rather than as just a casual AI application:
This set of questions encourages students to engage critically and reflectively with ChatGPT. They are designed not just to answer the what and how, but also to stimulate discussions on the why and what if, thereby fostering a more in-depth and nuanced understanding of the technology in a pedagogically sound manner.