What If There Were an AI Tool that Could Do My Grading for Me?
How I Use My Essay Feedback for Formative Assessment in Writing Courses
Key Takeaways (from TLDR Bot)
AI Feedback Tool for Grading: The author uses an AI tool called "My Essay Feedback" to provide detailed and personalized feedback to students, making grading easier and more efficient.
Student Engagement: Students can interact with the AI at any time, receiving immediate feedback, which reduces wait times for grading and enhances learning.
Customization & Control: Instructors can tailor the AI's feedback prompts and even involve students in creating rubrics, ensuring that the feedback aligns with classroom needs.
Integrated & Secure Platform: The tool works within Learning Management Systems (LMS) like Blackboard, ensuring privacy by not sharing student data with external models.
Positive Student Response: Students appreciate the AI's constructive feedback, noting its helpfulness in refining their writing and research skills.
When I started this Substack, I wanted to share a cool formative feedback tool I had the opportunity to pilot in my Fall 2023 and Spring 2024 English 102 and literary analysis courses. My Essay Feedback is the creation of Eric Kean, a viola and statistics professor at Western Washington State University who is “keenly” aware of how writing professors want to create a responsive feedback tool for their students. Anna Mills, one of my composition heroes and author of the excellent first-year writing OER How Arguments Work on Libre Text, first led me to this tool, and I have loved the opportunity to work with Eric and pilot/develop features.
Okay, I was kind of joking about outsourcing my grading to AI, but I also kind of wasn’t. I teach a 5:5 load that is heavy with composition courses (mostly English 102). I like to provide my students with lots of feedback, but it just has not proved to be sustainable. This tool allows me to design a prompt for weekly writing assessments where I once gave brief and “stock” comments. The output provides much more detailed feedback, the kind of information a student would get if they came to office hours, only now any student can access this, at 2:00 in the morning.
The assignments automatically populate in my gradebook once the student has met the upload, interaction, reflection, and revision criteria. Students love it because they don’t have to wait for me to grade. I love it because it makes my life so much easier. Basically, I just check the feedback and let students know whether I agree or disagree with the AI. About 95% of the time, the feedback is great! About 5% of the time, the tool goes off the rails, and we all laugh about that when we debrief in class. It’s a great real-world example of the need to be skeptical and critical consumers of AI output.
What I like about My Essay Feedback:
It integrates into my LMS (Blackboard). Students can purchase codes and use the website (as you do with Eli Review, for example). But the tool can also be integrated so that students don’t have to leave the LMS environment.
The instructor (sometimes with student help) creates the assignments and the prompts! I piloted another tool, Feedback Fruits, for my institution last year, that promoted its AI assistance. It gave me a very narrow set of AI prompts and did not allow me to create my own. This is not how I want to use generative AI in my classroom. I tailor my prompts and persona to the types of feedback I would give if students attend my office hours. I also have my students create a rubric for their essays using prompting.
Instructors can also leave feedback in various modalities (text, audio, or video).
Students can complete peer reviews on each other’s work using the platform.
Students can submit their assignments in a variety of ways. They can upload files of various types or copy and paste their responses directly into a textbox. The user interface is very intuitive for both students and me.
Students do not have to make accounts with ChatGPT or other tools (though this tool is powered by ChatGPT).
The platform protects students’ information and privacy. It does not share their work with training models.
I can view all the interactions they have with the AI including their prompts. This helps me to work with them on their prompting skills.
If I want to, I can choose to allow the tool to rewrite sections of the assignment. Sometimes I turn this feature off. I can also set guardrails on the chat if I want to.
The grades populate in my LMS gradebook automatically once students have met the criteria I set for their participation and interaction with the LLM.
What Students Are Saying:
I am working on an IRB proposal where I will be able to analyze the quality of the feedback more closely, but in the meantime, I wanted to share a few English 102 student feedback reflections with their permission. I require students to reflect on the quality of the feedback they receive and tell me what they plan to do with it. Here’s what a few of them have said:
The AI gave me feedback on how my first two sentences are a bit choppy and needs some more smoothing out and it makes perfect sense! It pointed something out to me that I hadn't noticed myself.
I think the feedback system provided adequate feedback to my research question. It suggested really valuable suggestions and gave me more in-depth insight that I could use moving forward in drafting my essay. I feel more confident in my topic now.
The feedback seemed honest and direct. I know what I must do, make my transitions smoother and switch my essay to appear a little bit more with APA format.
An Example of a Prompt: Narrow and Focused Research Question
So how does this all work?
Here are the student instructions for this assignment, a weekly writing task where students narrow and focus their research questions for their exploratory research essays. It’s the same assignment I used to assign and grade with brief stock comments.
This week, you will post your narrowed and focused research question and a short written explanation of why you have chosen this topic and how it relates either to you personally or to our local community. As part of this assignment, you will interact with the My Essay Feedback chatbot to refine your research question. I will also be providing feedback to you on your research question.
Instructions:
Post your narrowed and focused research question for Essay One (Exploratory Research) here. Then write 2-3 paragraphs explaining why you chose this topic, how it relates either to you personally or to our local community, and what you expect to find in your research.
Prompt: This type of feedback using the following criteria:
Review the student's research question. Is it narrow, focused, and appropriate for a 6-8 page exploratory research paper? In the student's explanation of the topic, do you see either personal connections or relevance to a local community? Is there any background knowledge that is important for the student to know? Propose an outline for a research paper that would answer this research question.
I hope this gives you some ideas about ways you can integrated generative AI tools without changing too much about your current workflow or assignments. Happy prompting!