Bringing Consciousness to AI Use: Part II

Active learning through writing in an era when students are using generative AI to produce writing

“Active learning through writing” is a WAC principle that emphasizes the role of writing as not merely a mode of communication or assessment but also as a valuable tool for enhancing the learning process. Before delving into the details of this pedagogical principle, we first clarify the concept of active learning itself.

Active learning is an instructional approach that actively engages students in the learning process, typically through activities such as problem-solving, group discussions, or hands-on experiments. The main idea is to shift from a passive lecture mode where students are mere recipients of information to one where they play a participatory role in constructing knowledge.

WAC advocates use writing as a tool to actively engage students in their learning process. Writing requires students to put words and symbols down on a piece of paper, either digitally or physically. This process, even in the most informal form such as freewriting, asks for students to actively internalize information, reorganize and synthesize various pieces of information, and articulate their understanding. This act enhances comprehension. Additionally, putting words on paper (physically or digitally) also offers a chance to revisit, revise and reflect on them afterwards. Pausing, thinking, and reflecting on their understanding and making cross connections facilitate active learning and critical thinking.

There are various ways of promoting active learning through writing, such as informal writing activities (e.g., journaling, free-writing), formal writing assignments (e.g., essays, papers, abstracts), and collaborative writing (e.g., peer reviews, group writing projects).Formal writing assignments benefit from these smaller, informal writings, which act as scaffolding for these major assignments in their breakdown of work into smaller, more manageable pieces.

With generative AI, instructors may be worried that students are using these tools to generate writing for the writing assignments. Students who heavily rely on generative AI to produce their assignment answers miss out on the cognitive benefits that come with the process of writing. If students use generative AI to do the bulk of their writing, they bypass the critical thinking, active learning and reflection that writing can foster. With the current technology, most students can easily access generative AI tools such as ChatGPT, and it is sometimes hard to accurately detect whether students are using these tools or not. Therefore, instructors may be reluctant to assign writing assignments to students.

Given the current circumstances, I argue that the WAC principle of using writing as a tool to actively engage students in their learning process is still valid and helpful to promote active learning and critical thinking, but to implement the principle well in our classrooms, we need to cultivate and learn AI literacy for ourselves and for our students and rethink our approaches to writing assignments. We need to understand the advantages and limitations of using these tools, have open conversations with students, and make students understand the goals and methods of the writing assignments so that they understand why they are doing the assignment (and less likely to have shortcuts by using the generative AI tools).

Assignment design ideas that encourage genuine engagement, making it difficult or unnecessary to lean on generative AI:

1. Have students keep a journal of their thoughts, questions, and reflections on a topic. Encourage them to write about their emotions, confusions, and epiphanies. Create writing assignments that relate to their personal experiences and feelings.

2. Conduct in-class peer reviews with specific guiding questions. The feedback process requires students’ active engagement.

3. In-class essays: In a monitored environment, students are less likely to write with the help of the AI tools.

4. Interactive group discussions and writings such as collaboratively working on a Google Doc, or assigning a text and asking students to co-create resources by platforms such as hypothes.is: students can annotate a text collaboratively as a low-stake assignment. By doing so, they learn from each other and contribute to annotations together. The dynamic nature of such platforms would make it harder for AI to be seamlessly integrated.

5. Scaffolding formal assignments into smaller pieces (outlines, multiple drafts and revisions) and asking students to reflect on the process and comment on what decisions they make through multiple drafts.

6. Multimodal works: Students may present their work in a non-text way, such as delivering their essays through videos/oral presentations. The process of delivering the text in other modals makes it hard to use AI to complete the task, and it requires students’ active engagement.

Assignment design ideas that creatively integrate generative AI to facilitate active learning through writing (adapted from AI Assignment Flip – 10 Examples):

1. Students use a generative AI tool to draft a paper based on specific guidelines. Then analyze and detail the strengths and weaknesses, focusing on aspects like flow, structure, references, and overall argument. Conclude by reflecting on the evaluation/review process.

2. Present students with a problem and instruct them to use a generative AI for coming up with 2-3 potential solutions. They should then prioritize these solutions, explaining their reasons for each ranking. Furthermore, students should present their own solution, elucidating their choice after considering all the AI-generated options.

3. Students first write a paper and then use AI to review it, exploring feedback in areas such as flow, structure, grammar, clarity, argument, and the overall tone. Based on the AI feedback, students refine their work. Students need to submit their initial draft, AI’s recommendations, the revised paper, and a brief reflection on what advice they take or not and why.

4. In the AI Debate assignment, students choose a topic open to interpretation and use AI to formulate pros and cons. They then examine the soundness and depth of the AI-generated debates, determining which side appears more convincing. Then students challenge the AI’s arguments, pinpointing any logical fallacies, inconsistencies or uncited statements.

“Active learning through writing” emphasizes the instrumental role of writing in facilitating deeper understanding and engagement with the teaching materials. In the era of generative AI, writing can still be a transformative tool in the educational process, promoting critical thinking, reflection, and active participation in one’s learning, but it requires some shift in perspectives and redesigns.

October 17, 2022October 17, 2022

Revisiting “Peer Review” in the Post-Pandemic Era

In WAC, we consider writing as a process that promotes critical thinking. One of its implications is to emphasize revision. The revision process is built into the scaffoldings of big projects during the semester. In revising, we make new connections and discover new ways to understand and express thoughts. Perhaps one support for this argument is to observe the differences between the second edition (2011) of John C. Bean’s Engaging Ideas, one of the key texts for WAC, and its third edition (2021) (in which Bean co-authored with Dan Melzer). Of course, much can be discussed regarding the differences between the two, as we have new developments of writing pedagogy in the past ten years (Preface ix-xi), face new challenges (e.g., covid-19), and adapt ourselves to new ways of learning and teaching. In the preface of the third edition, Bean and Melzer list major changes in the new edition (xi-xii). In this essay, I want to focus on the “peer review,” one of the key strategies for revision. I’ll trace and discuss the changes in Bean’s book and reflect on my own experience in teaching. I’ll also compare the revision process with the one from the design industry and show what I think can be improved in our teaching.

In the second edition of Bean’s book, guidance for peer review is in Chapter 15 “Coaching the Writing Process and Handling the Paper Load,” a chapter focusing on how to minimize teachers’ workload while at the same time encouraging students to write. Peer review is considered one of the key strategies to save time (295-302). Although it is still included as a strategy for saving time in the same chapter in Bean’s third edition, “peer review” is singled out and reframed into a new chapter “Helping Students Use Self-Assessment and Peer Review to Promote Revision and Reflection.” In this newly added chapter, not only does it include all the useful advice in the second edition, but also lays out recent research findings for the benefits of peer review (trying to advocate its value; e.g., research shows that “peer review from at least three peer readers results in feedback that is strikingly similar to instructor feedback and leads to as much or more global revision” (252)), admits the problems with peer review (244-245), and adds a new digital platform (Eli review), in addition to the original two (namely, Calibrated Peer Review and Peerceptiv). I like the shifting from a teacher-centered view to a student-centered one; in other words, the third edition emphasizes more how peer review benefits students.

Including the three digital platforms raises an interesting question: how should we use media to facilitate peer review in a post-pandemic era? Before the pandemic, I used to ask students to print out hard copies and bring them to class (this causes some problems: some students tend to work on their drafts at the last minute and cannot find a place to print their work in an early morning class). In class, they exchange drafts, write comments, and give them back to each other. Transitioning to online teaching, I used Zoom breakout rooms and Google Doc to do the work, with guided instructions. Coming back to the classroom again this semester, I find that students are more likely to bring their laptops to class and feel more comfortable with digital tools, so I continue using online platforms such as Google Doc and Google Form. To me, a digital copy can be accessed more easily than a hard copy, considering sometimes the print copy gets lost. Sometimes handwriting is not easy to recognize compared to the regularized typography. Of course, there are disadvantages. Different platforms can be overwhelming as students need to learn how to manage them. Students also lose the freedom of writing on a piece of paper, and of arranging and rearranging ideas in a non-linear way. Additionally, they are limited by a small screen, instead of a wider view of multiple pages. Handwriting is also more intimate than the standard font. We should also consider students’ needs and accessibility: some students prefer physical copies, while others like the digital version better.

We may continue discussing the pros and cons of learning through different media, but I want to return to the topic of how to do peer review. I believe that we should guide students what we instructors practice when we comment on their papers and lay out some guided instructions and templates for their reference. Without guidance, I find students tend to focus on editing sentences such as syntax, grammar, etc. In the prompt for peer review, I usually ask students to write marginal comments of only higher-order concerns, and an end-comment or a one-paragraph comment (312-313), which consists of comments on the strong and weak points of the draft, as well as the suggestions for revision. In this way, reviewers can focus more on the higher-order elements. I also consider it important to ask students to write specific and descriptive comments and support their comments with evidence and analysis, in line with Bean’s proposal for “descriptive questions” instead of “judgment questions” (247).

Although I previously asked students to talk/communicate with each other about the comments they make, to clarify any confusions and make a to-do list for revision, I found it hard for students to translate the feedback into revisions. I also find Bean and Melzer spend most of their time addressing how to do peer review and give good feedback, but not so much how to integrate comments into revision. Indeed, there is a short section on “Revision Plans” (239), which includes a sheet to ask students to reflect on if the reviewer can correctly identify the important elements (such as thesis, analysis, etc.) and think about how to revise. The set of questions are helpful to guide students to reflect on some important writing elements, but sometimes this top-down approach may ignore certain aspects that are not listed in the guidance but exist in the feedback. Hence, I would propose a bottom-up approach.

Peer review is not only limited to academic practices. It is also a critical process in designing an app or a website. During the summer, I took a series of courses on Google UX Design Professional Certificate, in which they discuss peer review in detail. Some suggestions are also valuable for my teaching.

I think perhaps one missing piece in Bean’s book is how to read feedback. The first tip would be to “stay open minded.” Reading feedback sometimes can trigger mixed and even defensive feelings. It is important to see the feedback as something to expand one’s own limited view, and to “think about where that feedback is coming from,” how it is different from one’s own expectations. Secondly, keep in mind one’s goals. For a writing assignment, what are the goals and rubrics? Could applying the feedback improve the writing to better meet the goals? If not, think about why the feedback is not effective and document it for later consideration. Thirdly, keep one’s own agency. There will be moments when one disagrees with the feedback, and that’s okay. Everyone has their own opinions about the work. When we disagree, we should not completely ignore the feedback. Instead, we need to think about where the disagreement comes from, and whether and how the different thought is supported by evidence and reasonable analysis. We then need to judge if the feedback would fit the writing goals.

In addition to how to read feedback, I suggest arranging the feedback in an “affinity diagram.” The diagram asks one to pull the comments from different sources to one single place (with sticky notes; online platforms such as Miro, Jamboard, Notely, Mural, Padlet can do something similar.) Sometimes marginal comments are spread out through the entire paper, and can be confusing if one receives feedback from more than one reviewer. Writers may not notice some common patterns without stepping back a little bit. The affinity diagram helps with synthesizing the feedback into groups. One may rearrange/reorganize the comments into big categories such as thesis, transitions, structure, support, evidence, citation, etc. The categories/related ideas/themes can also be customized to fit one’s needs. Think: if there are any connections or relations between feedback from different reviewers? Do any common patterns/themes stand out particularly? For example, if one notices that a few comments from different reviewers are about the transitions in different paragraphs of the essay, one may need to revisit some common strategies for making transitions and go back to examine the transitions again. Gathering all the comments together in one place and organizing them according to themes offer a global view on the feedback from bottom up and allows one to identify patterns that may not be easily recognized when the feedback is from different sources.

After organizing and thinking about feedback in the affinity diagram, the next step is to come up with “insights for each theme”/pattern. The “insight” is to write out a specific “next step” for how to improve the writing. The specific insights should be based on the feedback, evidence and analysis, and the goals and criteria of the writing assignment. They also should lead to direct actions to revise. When writing insights, one may follow a template like this: It was observed that xx out of xx reviewers think that ______________ (description of the feedback). This means that ______________ (analysis of the feedback). An insight based on the theme/pattern is: ______________ (what to do next). One should also prioritize key elements in revising. Revising thesis should be prior to revising transitions, as transitions may need to change according to the revised thesis. Grammar and sentence editing should be the last step because high-order concerns will lead to sentence-level changes. Some insights “should be considered a Priority Zero,” meaning “they must be fixed” for your essay, such as having an argument, not a descriptive statement for writing an argumentative essay.

In this article, I focus on peer review, one of the teaching strategies for revision by tracking the changes in two editions of Bean’s book, and by reflecting on my own teaching experience. I’m hoping the above tips from another industry that practices peer review—reading feedback, creating affinity diagrams, drawing insights from feedback—can offer some practical steps for students to translate the feedback into effective revision and complement Bean’s wonderful suggestions for peer review.

Works Cited

Bean, John C. Engaging Ideas. 2nd Edition, Jossey Bass, 2011.

Bean, John C., and Dan Melzer. Engaging Ideas. 3rd Edition, Jossey Bass, 2021.

Google UX Design Professional Certificate. Coursera.

October 22, 2021December 2, 2021

Writing as Revisiting Connections

Sometimes when I put down on a page my riveting ideas in my mind, these ideas suddenly become bland and dull. One reason behind this, I think, is that the linearity of writing works differently from our way of thinking. When I have an idea, or an image in our mind, it does not stand alone; it reverberates on a web with many potentially relevant images or ideas, close or distant. But the moment I put it down on a page, all the fascinating connections are cut off.

Sometimes I don’t have much to say when I start out writing. But putting words on a page triggers some connections I have never thought about before, and I end up having some good stuff.

One thing common in the two scenarios, I think, is to make connections, revisit connections, and find a way to present the pertinent connections in a linear way. Having an idea in mind is like ink dropping in liquid, curling, spreading, and unfolding in time, in all directions and dimensions. But writing is unlike this; it forces one to put letters from left to right, from top to bottom. This one-dimensional writing is inconsistent with thinking.

To bring these connections onto a page well usually doesn’t happen in the first round of writing. In the first round of writing, or freewriting, or “sketches,” which I’d like to call, I find myself too rushed and too overwhelmed to observe all the connections of every sentence and every claim. It isn’t until the third or fourth round that I cool down a bit and could finally examine each possibility and connection I could redraw and rearrange. So, one major task of my writing, I think, is to revisit these connections, make rearrangements, and let time unfold the connections.

Sometimes I share sample essays from an MLA handbook to show what a good essay looks like. But I always have the concern that these sample essays would fail to teach students how to get there from here. In a sample essay, each word has its fixed, appropriate position, as if this were the only version that the writer wrote. What it fails to tell is that it is not written in one shot. Admittedly, there are exceptions in history; Kate Chopin wrote rapidly and without much revision, or Rilke finished the Duino Elegies and 55 Sonnets to Orpheus in a few weeks in February 1922. But we must acknowledge that these may not be practical for us. For most of us, we still need time to revisit our writing, and let time distill the writing into a finer piece. These sample essays would be uninstructive especially for undergraduate students to understand the struggles in writing—the numerous revisions/drafts that lead to this final product, the time, the frustrations, the A-Ha moments—as if someone just simply typed in the letters in a sequence that is on the page.

The illusion may be unconsciously translated into our minds that one may produce a piece of good writing in one attempt. When I say “our” minds, I’m particularly thinking of undergraduate students who work overnight and finish their first draft before the dawn of the deadline and submit it as their final work. But I’m also thinking, how, in many courses, inappropriate or lack of scaffolding of a long paper also contributes to this mindset.

I would argue that making changes to the scaffolding doesn’t need much work. This semester, I’m teaching a writing-intensive course, in which we need to write three papers throughout the semester. I divide each paper into three stages. The first stage is called “preliminary draft” or “sketches,” in which students need to submit 1-2 pages of their writing. Instead of asking specifically for a thesis or an outline, I let them choose the most comfortable way to write: “It could be in various forms, such as a detailed outline, two or three body paragraphs, some close reading analysis, an introduction plus one or two body paragraphs, etc. It gives you some freedom in deciding how you want to approach the topic of your paper. The aim here is to get the paper going.” Students work differently towards their paper; some prefer having an outline, some love having a thesis first, and some just put down everything that comes to them. For this draft, I make minimal marking. Then a week later, I ask for a draft for peer review (2-3 pages). After peer reviewing, students will have another week to revise it into a final paper. Two to three weeks before the “sketches,” students know the prompt and could develop some thoughts through their journal entries and low-stake writings as we read texts. When they submit their paper, I also ask for a cover letter reflecting their writing process: “What are the significant changes that you made to the Essay 1 in your revision? Specify the changes you made and reflect on the changes made from preliminary draft to peer review draft, and then your final paper.”

I’m happy to find out that giving students one week for each draft and three weeks in total to complete a piece works well. When asked what they learned through the whole writing process, one wrote, “I learned there was more value in multiple drafts than I had thought, for most assignments I have a tendency to only do one draft and I could definitely see some improvements between drafts.”; another responded, “I have learned that drafts are very important and it is important to keep making changes to your writing and good writing takes time. When writing an essay it cannot be done in a day but in a course of a few days because it is important to think about what is being written and what ideas I have on the subject.” It is encouraging to know that some early planning and scaffolding, leaving enough time for students to revise and to improve through multiple drafts would benefit them. It is even more encouraging that some students realize how writers bring new connections onto a page through time and multiple attempts.

Author: Weiheng Sun

Bringing Consciousness to AI Use: Part II

Active learning through writing in an era when students are using generative AI to produce writing

Revisiting “Peer Review” in the Post-Pandemic Era

Writing as Revisiting Connections

The OpenLab at City Tech:A place to learn, work, and share

Support

Accessibility

Copyright

Creative Commons

The OpenLab at City Tech:A place to learn, work, and share

Support

Accessibility

Copyright

Creative Commons