Skip to content
Get in touch
← Browse labs

Lab / activity

Try, Check, Explain

In development
About this activity

Can I use this?
This preview contains developing learning material. Its instructions, practice, and supporting materials still need a complete usability review.

Is help wanted?
No specific development request is open.

Has it been reviewed with learners?
Learner review is not yet documented.

Jump to related resources ↓

What you will make

A first attempt and a checked retry that show what you can explain yourself, what one piece of AI help changed, and what you still need to learn. These may be spoken, drawn, demonstrated, or written. Keeping a record is optional.

Before you start

Required: access to a chatbot for the live AI activity. Operate it yourself, direct a partner, or guide a facilitator-operated exchange. No learner account or recording is required.

Choose a low-risk learning goal and a readable reference you can open. The required Resource Reference packet offers two starting selections and guidance for choosing your own. Use public or learner-owned material that you may share. If you need a sharing checkpoint, open Module Safety: Share Less, Notice More.

For self-study, you may inspect one Resource Gallery example, then work on your own question or an unworked branch of the example. In a Zoom or in-person session, the demonstration opens the activity; learners then practice individually, in pairs, or in groups. Watching alone prepares you to begin but does not complete the practice. If live AI is unavailable, prepare the goal and first attempt and return for the live step.

Lab materials

Activity

Follow the complete activity here or open the optional Notebook Lab Notebook. You may leave note fields blank and explain your thinking aloud.

  1. Choose a goal and reference

    For self-study, you may inspect one Resource Gallery example, then choose your own question or an unworked branch. In a Zoom or in-person session, use the demonstration as an opening, then practice individually, in pairs, or in a group. Watching alone is preparation; completion requires your own attempt, chosen live AI help, source check, and retry. You may work aloud without an account or saved record.

    Complete this sentence: β€œAfter this practice, I want to be able to explain or do…” Choose one bounded idea or skill. For example, explain why a mountain comparison needs a starting point, or revise a historical claim so it stays within its evidence.

    Open the required Resource Reference packet. Select a linked source or choose a suitable source for your own inquiry using its guidance. Locate a passage, diagram, or worked example that can help you check your attempt. Name who produced it and what it can establish. Keep the reference available, but do not copy its answer into your first attempt.

  2. Make an unaided first attempt

    Close or cover the relevant explanation and try the task before asking AI. Say, sketch, demonstrate, or write what you currently think and why. Mark one uncertainty instead of disguising it. β€œI do not know yet; I think the difficulty is…” is a useful starting attempt.

    Keep enough of that attempt in mind or in optional notes to compare later. You are finding a place to practice, not earning a score.

  3. Ask for one piece of help

    Required: access to a chatbot for the live AI activity. Operate it, direct a partner, or choose the request in a facilitator-operated exchange. You do not need a learner account or a recording.

    Use public or learner-owned material you may share. Open Module Safety: Share Less, Notice More if you need help deciding what to remove or keep out before sending a passage.

    Give the chatbot your goal, first attempt, and a short permitted source passage or reference description. Ask for one hint, example, or practice question rather than a completed explanation. For example: β€œI want to learn [goal]. My first attempt is [attempt]. Using this source passage, give me one hint about [difficulty]. Do not finish the answer for me. Wait for my next attempt.”

    Inspect what actually comes back. If it gives away the answer, set that response aside and request a new practice question. If it refuses, invents support, or cannot work with the passage, identify the limitation and try a narrower request. A generated hint still needs checking.

  4. Check the help against the reference

    Reopen your selected source from the Resource Reference packet, or the source you selected for your own goal. Point to the passage, diagram, or worked example that supports or contradicts the hint. Separate the source's statement from the chatbot's interpretation and from your own inference.

    Decide whether to use, correct, reject, or leave the help unresolved. A second chatbot answer is not an independent source check. If support is unclear, say so and narrow the claim. Module Reality Check offers a further checking path.

  5. Try again and transfer

    Put the AI answer aside and make a new attempt in your own words or actions. Explain what changed and point back to the reference that justifies it. If you still cannot explain it, name the missing step rather than copying the answer.

    Now change one condition: a different measurement, another claim, a new example, or a new question. Try that transfer yourself before requesting more help. Check it against an appropriate reference; the original source may not cover the new case. The Resource Reference packet gives starting transfer questions with visible limits.

  6. Explain and choose the next step

    Explain your first attempt, the help you inspected, the evidence you used, and what your retry shows. Include what remains uncertain. A partner may ask β€œWhat makes you think that?” without supplying your explanation.

    Choose whether to practice another case, find a better reference, ask a person, return to the lesson, or stop. If the retry remains uncertain, that is an honest result; identify the next learning need. Completion is participation in this checked practice, not proof of mastery.

Check your result

  • The learning goal names something you wanted to understand or do.
  • The first attempt came before AI help.
  • You can locate the support or the gap for the help you received.
  • Your retry and transfer attempt show your own thinking, including uncertainty.

You are done when

  • You made an unaided attempt at a bounded learning goal.
  • You chose and participated in one live chatbot request for bounded learning help.
  • You checked that help against a reference and made a reasoned decision about it.
  • You tried again, attempted a changed case, and explained what changed or remains difficult.
  • You chose a next learning step or a reason to stop; no saved record is required.

Try another version

  • Swap roles with a partner and practice a different question.
  • Compare a hint with a worked example, making an attempt before each new piece of help.
  • Follow a new question from your Curiosity exploration using a reference you select yourself.

Help improve this lab

Tell us what helped, what was unclear, or what you would change. A quick observation is welcome.

Share a review β†’