Generative AI and Writing as a Process: Part II

Revising a WAC Informal Assignment Using Generative AI

In the previous blog posts, we explored how WAC principles—in particular, active learning through writing and writing as a process—can facilitate students’ learning and deepen their engagement in class. In addition, WAC fellows also introduced competing perspectives on using generative AI for writing assignments, acknowledging its transformative aspect as an educational tool and, at the same time, suggesting its inherent limits and the need for redesign. In this section, we will explore specific exhibits suggesting possibilities of how professors can use generative AI as an educational tool to develop a WAC informal assignment. Here is one of the most common scaffolding assignments of Writing Intensive courses: 

———————————————————————————————-

Exhibit 1. A module of a WAC informal assignment

Suppose you are an emerging scholar applying for a conference. The call for proposals states that you should submit an abstract of 250 words.

In your abstract, you will walk the readers (in this case, the conference committee) through your 1) premise (essential information your readers should know about your object of study), 2) topic (what interests you from the object of study), 3) tentative thesis (your own argument which will be supported by your analysis and observations of the object of study), and 4) blueprint (a roadmap of your essay).  

In order to brainstorm and develop your abstract, follow the following steps:

  1. Freewrite what you want to write in your conference paper.
    • For ten minutes, write down what comes to your mind regarding your object of study. Do not worry about mistakes, grammar, or spelling errors. Do not reread your writing or edit it. If you are a multilingual speaker, feel free to write words in a language you find more comfortable if you cannot think of a word in English. The point of this exercise is to let your thinking flow so you can use this time to brainstorm your thoughts.
  2. Based on your freewritten draft, write your tentative thesis (one core sentence that can summarize the main argument of your paper) and abstract. 

Make sure your abstract includes 1) premise, 2) topic, 3) tentative thesis, and 4) blueprint. 

———————————————————————————————-

The case of Exhibit 1 is helpful when students are not familiar with writing an abstract. It specifically outlines what a professor expects from students’ abstracts: 1) premise, 2) topic, 3) tentative thesis, and 4) blueprint. Yet, it is designed as an assignment with top-down instructions, which can confuse students if they are not engaged in a thorough conversation regarding the definitions and significances of these elements. Therefore, professors often prepare an in-class session to explain what these elements are and why they are needed in this piece of writing. They might also give detailed handouts to students.

In this process, what makes both professors and students frustrated is that some students do not grasp the definitions and significances of these elements even after conversations in and out of class which, from the professors’ point of view, seemed thorough. However, this situation can happen since these elements are often entirely new concepts to students, even though they are familiar to professors who have already written multiple abstracts during their academic journey. Some students might understand the definitions but have not yet grasped the significances (i.e., why they should include these elements and write them in a particular order). In contrast, other students might understand both definitions and significances but struggle to apply the concepts in their own writings. 

Here, WAC fellows suggest a module of a WAC informal assignment that uses generative AI as an educational tool. This assignment is designed to supplement assignments like Exhibit 1. For instance, you can assign this assignment only to students needing help writing an abstract. Another option is to assign this module as an in-class activity that you will reflect on with students as a group. In either case, we recommend doing this module after finishing Exhibit 1 or its equivalent in that this is a module designed for students’ reflection on their writings. 

———————————————————————————————-

Exhibit 2. A module of a WAC informal assignment that uses generative AI

Instructions for students

  1. Ask Chat GPT 1) to generate an abstract based on your freewritten draft and 2) to summarize your freewritten draft in one sentence (which will be your tentative thesis). Make sure to provide specific details about your target audience. (e.g., “I will submit this abstract to a journal that focuses on modern American drama.”) 
  2. Compare and contrast two abstracts (an AI-generated abstract and your own).
  3. Write a one-paragraph reflection, considering the following questions:
    • Did two abstracts look similar? Did the AI-generated abstract include the main point of your tentative thesis? Did it miss any details you think are important and necessary in your abstract draft? 
    • If so, how can you make your abstract clearer—so your revised abstract can include your main theme and thesis as well as a blueprint of your paper? 
    • Did you take any advice from the AI-generated abstract? If so, what was it? What did your own abstract miss to include?
    • Are there any changes you want to make further based on this comparison?
  4. Revise your abstract according to your one-paragraph reflection.

Submit 1) an AI-generated abstract, 2) your revised abstract, and 3) your one-paragraph reflection.

———————————————————————————————-

Using generative AI as a tool, students can revise their own abstracts, comparing and contrasting them with an AI-generated abstract, a product of AI’s machine learning algorithm based on the numerous abstracts it collected. This act of self-revision is helpful in that students often find it hard to revise their own writings, and many of them end up editing their work instead of revising it. In this context, this exercise provides them with an accessible AI reader—who will not judge them or whom they will not feel bad asking to read their work—while students themselves can be critical readers of this model writing produced by AI. 

Possible concern 1: What if AI’s model writing is a bad example?

Some model writings of AI can be really bad. They might have overly decorative rhetoric and/or a structure not addressing scholarly conventions. Yet, this WAC module is designed to help students with WAC principles (active learning through writing and writing as a process). Writing a one-paragraph reflection, students will have a chance to independently think about whether the AI-generated abstract is well-written. In other words, it gives them a chance to think critically about 1) whether the AI-generated work is fully credible and 2) how they can make use of this technology as a writer. In doing so, they can identify points of confusion and develop their ideas regarding their abstract. Through this metacognitive process, students will have a chance to think about their target audience and their own thinking process both as a reader and a writer. 

Possible concern 2: More writing means more grading!

Some professors might share a concern that this new module burdens them with two extra assignments to be read and graded (an AI-generated abstract and students’ one-paragraph reflection). Considering professors’ heavy workload, here are three possible suggestions based on WAC principles:

  1. Minimal marking: Instead of changing sentence level errors or giving line edit comments, give a few comments highlighting strengths and encouraging students to identify their patterns of errors. 
  2. Exploratory writing: One-paragraph reflection is designed to be an exploratory writing assignment that guides students to discover, develop, and clarify ideas. In other words, reading this short paragraph will allow professors to follow students’ trajectories of thoughts. Therefore, this exploratory writing assignment can lessen professors’ time used to find (or sometimes guess) what students might have wanted to write in their assignments and, in doing so, assist professors in giving students more pinpointed feedback.
  3. Low-stake in-class activity: Instead of grading them, professors can consider making this an in-class freewriting activity, which will also improve class preparation and discussion. 

In this section, we explored a sample exhibit of the WAC informal writing assignment (Exhibit 1) and another exhibit that uses generative AI as an educational tool to supplement the WAC informal writing assignment (Exhibit 2). Acknowledging some inherent limits of generative AI, this section suggested some possible uses of generative AI based on WAC principles, focusing on how professors can encourage students to learn through reflective and metacognitive writing.