Quick post: How can the FreeStyle tool in CoPilot be used to give feedback on student writing samples against written descriptors?
This is a quick experiment in using GPT to give feedback on writing, using MYP assessment descriptors*. I used CoPilot’s AI FreeStyle space, but it’ll work in any GPT writing tool. I wouldn’t advocate for using this as final assessment, and is not intended to replace strong teacher-student relationships or interactions. It might be useful to you in your workflow of feedback and improvement for students, or in student-teacher conferences/coaching towards better writing.
The key to making GPT tools work for you is being clear and specific in the prompts you give them. They (currently) won’t work for mathematical notation or images/graphs, but could be helpful for written essays. For two useful structures on prompting for teachers, check out Dan Fitzpatrick’s PREP model, or Tom Barrett’s CREATE framework.
In these experiments with MYP Assessment:
- I told CoPilot what role to play and the purpose of its task.
- I identifed and presented the sample work.**
- I identified the assessment descriptors.
- I gave it clear instructions for what to look for and how to present the feedback.
CoPilot for Student Feedback by Stephen Taylor Go Fullscreen here
Some Productivity Tips:
- Invest time to make time. Experiment with prompts until you get the type of feedback you’re looking for.
- Keep your own documents of prompts and formats, so you can edit, iterate and paste in to be reused.
- If it’s not useful, don’t use it. You might already have great workflows for feedback, so no need to reinvent the wheel.
Some Go-To Tools:
- CoPilot’s AI Freestyle space
- Perplexity, for GPT + Bing Search, chat and sources
- WAB AI Innovation Guide
- (If You) USEME-AI
*For quick reference, here are the MYP Descriptors.
**(actually, it was CoPilot-generated examples, but you know what I mean)
Robo-tiger teacher giving feedback to tiger student, generated in MidJourney by Stephen.
Thank-you for your comments.