 |
| A few weeks ago, I experimented with a method for integrating AI into peer review. My goal was to see if AI can be used by students as a tool for evaluating their own writing. When I do peer review, I give students a guide with a series of steps and questions designed to help them dissect their drafts and analyze them via my grading criteria. I used the same strategy for this activity, but I integrated AI by having students do the steps with both a partner and an LLM. I put students into pairs and guided them through the steps in this slide show.
As with any peer review activity, the benefit of using an LLM is not in the content of feedback but in the conversation the process encourages. Note that most of the prompts in the slideshow do not ask readers to evaluate; they ask readers to identify and summarize various aspects of the essay (e. g. thesis, paragraph topics, topic sentences, transitions, etc.). The goal is to help students pay closer attention to what they actually say and do in their writing (not what they *think* they say or do in their writing), which is the first step in learning how to evaluate writing. AI can be helpful in doing that. The challenge is to help students understand that the feedback isn’t the point; it is in how the process helps them see their writing differently. And helping students understand this point is the same challenge that has plagued peer review “since the beginning of time” (to quote the start of way too many college essays). |