Helin and I had our presentation today, and time flew by faster than we expected. Here’s my personal reflection on Inquiry 1. Preparing and delivering our presentation on “How AI improves student learning and academic performance” pushed me to move beyond abstract enthusiasm into evidence-anchored practice. Going in, I expected a familiar story—AI helps because it’s “smart”—but the reading and our rehearsal cycles clarified that the gains come from three very specific mechanisms: personalization (adaptive paths that meet students where they are), immediate feedback (formative, targeted), and teacher analytics (dashboards that surface patterns and outliers). Designing the mini activity where classmates used an LLM to critique their own thinking made this concrete: the strongest moments usually weren’t generic explanations, but the model diagnosing that person’s misconception and proposing a next step. Presenting under time pressure also taught me to prioritize claims tied to data (e.g., K–12 math meta-analyses and physics case studies) and to name limitations up front along with guardrails (process logs, prompt scaffolds, and assessment designs that reward reasoning). For Inquiry 2, I want to carry forward two threads: (1) a small, classroom-level study in math/physics that compares an AI-supported formative workflow (diagnostic prompt → targeted practice → reflection) against a conventional one, using pre/post performance and a simple engagement/progress tracker; and (2) a practical “AI use agreement” for students that aligns with academic integrity while still encouraging metacognitive help (e.g., “explain my error,” “generate isomorphic practice,” “ask me probing questions”). This first phase has opened unexpected directions around equity: AI appears especially helpful for multilingual learners and students who benefit from private, low-stakes rehearsal before speaking. My curiosity now points to designing prompts that teach habits, not just give answers, and to building lightweight teacher dashboards that convert model feedback into actionable small-group plans. In short, Inquiry 1 provided me an oppurtunity to expand from “AI as content generator only” to “AI as formative infrastructure too”.
Below is the link to our google slides:
https://docs.google.com/presentation/d/1Dhpj_U8SHdDNISd-rr8wr9umChSctiC5T_-dau6eDhU/edit?usp=sharing
