AI Literacy Instruction in First-Year Writing Courses
| Category | Writing assignments Developing AI literacy |
| Instructors | Tracy Hamler Carrick, Senior Lecturer |
| John S. Knight Institute for Writing in the Disciplines | |
| College | College of Arts and Sciences |
| Name of Course | WRIT 1370/80: First-Year Writing Seminars |
| Discipline | Writing |
| Course-level | First-Year Undergraduates |
| Course size | 12 students |
| Implemented | Spring 2023/4, Fall 2023 |
Brief Summary
Throughout the semester, a variety of AI-generated texts are brought to class for workshopping and discussion as students explore best practices for integrating generative AI into academic writing processes. Students contemplate when, where, and how the tools can support their development as academic writers.
Learning Outcomes Targeted
Writing Processes
Reading Comprehension
Research Skills
Information Literacy
Context
When first learning about generative AI, instead of seeing it as a problem or something to be afraid of, Tracy Carrick saw it as another tool to help students. For example, she had been using QuillBot and Grammarly with her students for years. Using ChatGPT seemed like a logical extension of this work. She tells students to, “Figure out ways to let tools and resources do some of the heavy lifting for you. When used ethically, responsibly, and strategically, they can help you learn to be a stronger writer.”
Working together, students explored the capabilities of generative AI in the classroom. As a group, they determined how generative AI can be useful to writers with specific writing projects and when transitioning into new writing contexts as they move more deeply into disciplines and professions. Through these interactions, the instructor quickly discovered that students vary widely in their understanding of AI, including:
- What generative AI is.
- How they might already be using different forms of artificial intelligence in their practice.
- What uses of generative AI violate Cornell’s academic integrity policy.
- How to use generative AI effectively, strategically, and ethically in their writing.
“Figure out ways to let tools and resources do some of the heavy lifting for you.”
Using this technology surfaced sometimes hidden inequities and differences in foundational educational experiences across her student population. Some students, for example, are not exposed to academic writing before starting at Cornell; others may have limited experiences; and others still may have diverse learning needs that require wider-ranging pedagogical approaches. In these cases, generative AI has the potential to bridge gaps in achievement, development, and experience by modeling the many types and unique structures of formal academic writing.
Implementation
In the classroom, two types of activities and assignments, “AI-Free Zones” and “Generative AI Integration and Writing Instruction” guided students in understanding the effective and collaborative use of generative AI.
AI-Free Zones
These assignments focused on ensuring students develop learning and writing practices without the use of AI. Some strategies used were:
- Writing with and about proprietary material
- One-on-one writing conferences
- Graded drafts
- Explicit assignment guidelines that penalize AI use
Generative AI Integration and Writing Instruction
These assignments focused on how generative AI can be used as a reading and writing assistant in ways that help meet learning goals.
Students experimented with generative AI in class – workshop-style – at different stages of the writing process; for example, when figuring out the language for a research question/topic, as a research assistant by directing generative AI to compile relevant source materials, as an outlining tool, as an invention tool to spark ideas or highlight critical questions, and as an editing tool to find alternative ways to compose sentences. Students also experimented with generative AI as a reading assistant; for example, by putting prompts into ChatGPT to generate a summary of a reading assignment.
Most of the work students did with generative AI was in the classroom. When students used generative AI tools independently to develop their own research and writing projects, they accounted for their work in conferences with the instructor and/or with written reflections that describe and reflect on the ways they used generative AI.
Challenges
Preparing students to use generative AI tools for assignments – such as explaining what generative AI is and setting up accounts – required more time than expected, as some students were not familiar with generative AI.
Reflection and Future Directions
“I have become increasingly committed to providing basic AI literacy instruction in my first-year writing courses and to encourage other instructors of First-Year Writing Seminars to join me in asking: How can we use our classroom spaces to provide instructional guidance on when, where, and how to use generative AI ethically and responsibly?” Tracy Carrick, instructor.
It is also important, Carrick notes, to not get too caught up in the hype. She and students have found that generative AI cannot replace traditional writing instruction. If anything, it is now even more necessary. To Carrick, generative AI’s greatest benefit in a writing classroom is its capacity to reinforce and reinvigorate writing instruction. It has proven, so far, to be a useful instructional tool for helping first-year students learn how to write with greater agility, confidence, and stylistic force.
“How can we use our classroom spaces to provide instructional guidance on when, where, and how to use generative AI ethically and responsibly?“
Tracy Carrick
Some practical challenges going forward are:
- Continuing to develop lesson plans that thoughtfully embed AI literacy into coursework in ways that support and enhance learning goals.
- Redesigning learning outcomes and grading practices to equitably acknowledge the many ways that students choose (and are able) to use generative AI tools to support their writing and learning processes.
- Providing equal access to premier tools and resources.
These writing classes were small – capped at 12 students. Instructors in larger classes might not find whole-class discussions or workshops manageable. Instead, after students complete a writing assignment, instructors of larger classes might have students pair up, talk about or share their assignments, and respond together to a series of reflective questions that reinforce the metacognitive awareness:
- Which generative AI tools did you use?
- When did you integrate these tools?
- What impact did the material you generated have on your process? Did it help or hinder it?
- How did you feel about using these tools?
- How do you think others (classmates, faculty, parents, friends, future employers, etc) might feel about the ways you used these tools?
- How is your work different as a result of using these tools (not better or worse, necessarily, but different)?
- What have you learned about generative AI and yourself as a researcher or writer after using these tools?
- What are the risks and rewards of using generative AI tools?
- Will you use them again? If so, how, when, and why?
How to Implement This in Your Class
Provide spaces for students to experiment with and have conversations about generative AI. Two assignment examples include:
Assignment Example 1:
- Take a stage in the writing process and, for homework, have students complete a task using their typical methods; for example, coming up with a research question or finding scholarly resources in the library.
- Then, come together as a class to talk about their process. What go-to strategies do they use organically?
- The instructor then introduces a prompt for students to workshop in class to learn new strategies, some “analog” and some using tools such as generative AI. For example, have students use generative AI to find different ways of phrasing a research question or finding search terms to use on the library website.
Assignment Example 2:
- For homework, have students write a summary of a reading using their own prompts and a generative AI tool.
- Instruct students to not stop prompting the tool until they get a summary that is close to what they would write themselves.
- In class, have students share what prompts they used to get their final summary. What is important for this assignment are the prompts used – not the resulting summary. The prompts demonstrate not only the student’s understanding of the reading but also their understanding of how writing in a particular genre works.
Resources and links to additional materials
- Filling in Research Gaps with Generative AI (Tracy Hamler Carrick, KNIGHTLYnews, 5/01/2023)
- Guidance for AI Best Practices in FWS (Knight Institute AI Committee, Chaired by Tracy Hamler Carrick 10/20/2023)
- Can FWS Instructors Ban ChatGPT? (Tracy Hamler Carrick, KNIGHTLYnews, 10/23/2023)
- gAI in my FWS (Tracy Hamler Carrick, Cornell Faculty Forum presentation, 11/29/2023)
- Welcoming AI into the Classroom (Kathy Hovis, A&S News, 12/05/2023)
- Slide Presentation: Creative Responses to Generative AI Embedded AI Literacy Instruction in my FWS