Implement the following plan:

Plan: Study Guide Lab — Feb 26, 2026

Context

Midterms are approaching for students' other courses. This activity gives students practical study time while teaching them effective AI-assisted study techniques. Students bring materials from another course, pair with someone studying a different subject (so the partner serves as a naive skeptic who can't be fooled by confident-sounding AI), and work through three rounds of AI-assisted studying — each introducing a different technique grounded in learning science.

Key design decisions from brainstorming:

• Students study for their other courses, not this one
• Partners should be in different courses (naive skeptic model)
• Students bring their own materials (notes, slides, textbooks)
• Use external AI tools (ChatGPT, Claude, etc.) — not the course chat
• Each round: pair kickoff (~3-5 min) → solo study (~10-12 min) → share-out (~10 min)
• Each round suggests a technique but students can choose a different approach
• Each round includes a learning science nugget explaining why the technique works
• Brief text reflection after each share-out
• 75 minutes total (3 rounds × 25 min)

File to Create

content/spring2026/meetings/2026-02-26-study-guide-lab.mdx

Use the same MDX format as 2026-02-24-how-llms-work.mdx — YAML frontmatter, InstructorOnly facilitation notes, ActivityStages with 6 stages.

Frontmatter

date: 2026-02-26
semester: spring2026
title: "Study Guide Lab"
summary: "Use AI to build study materials for your other courses while learning evidence-based study techniques"

Facilitation Notes (InstructorOnly)

Include:

• Philosophy: Students get real study value for their midterms while learning when and how AI helps (and doesn't help) with studying. The three rounds introduce techniques backed by learning science research. Partners are ideally in different courses — the "naive skeptic" can catch moments where the studier just accepts AI output without understanding it.
• Logistics: Ask students to bring course materials (notes, slides, textbooks) for another class. Announce this ahead of time if possible. Pair students in different courses. No formal roles — pairing is for kickoff discussion and share-out, studying is individual.
• Pacing table:

Intro Text (before ActivityStages)

Brief paragraph: Midterms are coming — today you'll spend class time studying for your other courses, using AI as a study partner. Each round introduces a different study technique backed by learning science research. You'll work with a partner who's studying for a different course, which is intentional — they'll help you stay honest about whether you actually understand the material or whether AI is just making you feel like you do.

Instruction to have materials ready (notes, slides, textbook) for another course they need to study for.

Stage Structure (6 stages, no GroupActivity component)

Pairing is informal (sit with a partner), no enforced roles needed.

Stage 1: "Round 1: Active Recall" (15m)

Learning science nugget: Research consistently shows that trying to recall information from memory is one of the most effective ways to strengthen learning — far more effective than re-reading notes. This is called the testing effect: the act of struggling to remember something makes the memory stronger when you do find the answer.

Instructions:

1. Pair kickoff (~3 min): Tell your partner what course and topic you're studying for. Discuss: what's your plan for using active recall with AI?
2. Solo study (~12 min): Work individually with AI using active recall.

Suggested approach: Write down everything you know about a topic from memory — no notes, no AI. Then share your brain dump with AI and ask it to identify what you're missing or got wrong. Pay attention to whether the gaps AI finds are things you genuinely didn't know vs. things you just forgot to write down.

But you could also: ask AI to quiz you verbally with open-ended questions and answer from memory; use AI as a flashcard generator and test yourself; or try any other approach that starts with recall from memory before using AI.

Stage 2: "Round 1: Share Out" (10m)

FacilitationNotes: Ask 3-4 students what happened. Good prompts: "What did AI find that surprised you?" "Did AI identify any gaps that were actually wrong — things it said you missed but you actually know?" "Did anyone catch AI confabulating about your course material?"

TextSubmission (id: study-round1-reflection): What study technique did you use? What did you learn about how well you know this material?

Stage 3: "Round 2: Practice Testing" (15m)

Learning science nugget: Self-testing is the gold standard of study techniques. Students who test themselves consistently outperform those who spend the same time re-reading material — even when the re-readers feel more prepared. That false sense of readiness is called the illusion of competence: material feels familiar when you read it, but familiarity isn't the same as being able to produce an answer on an exam.

Instructions:

1. Pair kickoff (~3 min): How did round 1 go? What topic are you focusing on now? How are you planning to use AI for practice testing?
2. Solo study (~12 min): Work individually with AI using practice testing.

Suggested approach: Share your course materials with AI. Ask it to generate practice exam questions at the level your professor would ask. Answer them without looking at your notes or AI. Then check yourself. Focus on the questions you got wrong — those are your study priorities.

But you could also: ask AI to generate questions in different formats (multiple choice, short answer, explain-a-concept, true/false with justification); have AI role-play as a strict professor giving an oral exam; ask AI to generate questions that target common misconceptions; or try any other approach centered on testing yourself.

Stage 4: "Round 2: Share Out" (10m)

FacilitationNotes: Prompts: "Were AI-generated questions good? Too easy, too hard, or about right?" "How did your confidence compare to your actual performance?" "Did anyone find that AI was bad at writing certain kinds of questions?" This is a good moment to reinforce the illusion of competence — many students will report feeling like they knew the material but struggling on specific questions.

TextSubmission (id: study-round2-reflection): What study technique did you use? What did you learn about your own confidence vs. actual understanding?

Stage 5: "Round 3: Build a Study Aid" (15m)

Learning science nugget: Creating organized study materials — concept maps, cheat sheets, summary notes — forces you to make decisions about what matters most and how ideas connect. This deep processing strengthens understanding. But there's a trap: if AI creates the study aid for you, you skip the cognitive work that makes it valuable. A beautiful, clear AI-generated summary can make you feel like you understand the material when you've really just read something that's easy to follow.

Instructions:

1. Pair kickoff (~3 min): What are your weakest areas from rounds 1 and 2? What kind of study aid would be most useful? How do you plan to keep yourself doing the thinking rather than outsourcing it to AI?
2. Solo study (~12 min): Work individually with AI to build a study aid for your weakest area.

Suggested approach: Pick your weakest area (revealed by the first two rounds). Start by outlining the key concepts yourself, then use AI to help you fill in details, check accuracy, and improve the organization. The goal: a study aid where you did the thinking and AI helped with the polish.

But you could also: ask AI to explain a concept three different ways and pick the one that clicks; have AI generate analogies for difficult concepts; create a "misconception sheet" (common wrong answers with corrections); build practice problems with worked solutions; or try any other approach that produces something you'll use to study later.

Stage 6: "Round 3: Share Out" (10m)

FacilitationNotes: Prompts: "Show your partner your study aid. Can they quiz you on it — and can you answer without looking?" "How much of the study aid came from your understanding vs. AI's generation?" "Which of the three rounds helped you learn the most? Why?" Use this final share-out to connect all three rounds back to the key theme: AI is a powerful study tool, but only if you use it in ways that make your brain do the work.

TextSubmission (id: study-round3-reflection): What study technique did you use? Across all three rounds, which approach helped you learn the most and why?

Files Summary

Verification

1. Start dev server (or confirm it's running): curl -sk -o /dev/null -w "%{http_code}" https://[local-hostname]/
2. Navigate to https://[local-hostname]/spring2026/meetings/2026-02-26-study-guide-lab and verify the page renders
3. Check that all 6 stages appear in the activity stages component
4. Verify InstructorOnly content is visible when logged in as instructor
5. Verify TextSubmission components render in share-out stages

If you need specific details from before exiting plan mode (like exact code snippets, error messages, or content you generated), read the full transcript at: ~/.claude/projects/-Users-challen-www-usingandunderstanding-ai/a195b5ab-fe9f-459c-8ace-4cae7ac7eaa3.jsonl

Let's also create an e2e test for this activity, like the others, opt-in only using the same naming convention.

When I run the tests they fail: 130 | // Verify stage content is visible 131 | for (const page of [s1, s2]) { > 132 | await expect(page.getByTestId("activity-stage-2-content")).toBeVisible({ timeout: 5000 }); | ^ 133 | } 134 | 135 | // === STAGE 3: Round 2 Share Out (has TextSubmission) === at ~/www/usingandunderstanding.ai/testing/e2e/tests/meeting-2026-02-26.spec.ts:132:66

attachment #1: screenshot (image/png) ──────────────────────────────────────────────────────────
testing/e2e/test-results/meeting-2026-02-26-Meeting-64e8b-issions-in-share-out-stages-meeting-2026-02-26/test-failed-2.png
────────────────────────────────────────────────────────────────────────────────────────────────

attachment #2: screenshot (image/png) ──────────────────────────────────────────────────────────
testing/e2e/test-results/meeting-2026-02-26-Meeting-64e8b-issions-in-share-out-stages-meeting-2026-02-26/test-failed-3.png
────────────────────────────────────────────────────────────────────────────────────────────────

attachment #3: screenshot (image/png) ──────────────────────────────────────────────────────────
testing/e2e/test-results/meeting-2026-02-26-Meeting-64e8b-issions-in-share-out-stages-meeting-2026-02-26/test-failed-1.png
────────────────────────────────────────────────────────────────────────────────────────────────

1 failed [meeting-2026-02-26] › testing/e2e/tests/meeting-2026-02-26.spec.ts:25:3 › Meeting 2026-02-26: Study Guide Lab › full meeting flow: three study rounds with text submissions in share-out stages 1 error was not a part of any test, see above for details

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.

OK! Now, describe today's activity to me.

Wait, I think we agreed that the partner wouldn't act as naive skeptic. It's more a person to discuss their ideas with before working with AI to implement them.

Wait, sorry. I want enforced pairing, just without the roles.

Please run the tests since they are failing.

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.

OK! Next, I'd like to create a Feedback component that we can use to gather per-activity feedback going forward. We should have done this earlier but I forgot: oops. This should probably also count toward the activity time in a configurable way. This is always an individual component.