12 · Final quiz: the science-fair capstone

1. What is the scientific method, in plain words?

A way to make AI do your whole project for youA careful, step-by-step way to ask a question and find a trustworthy answerA rule that science projects must use a robotA trick to always get the result you hoped for

2. Which of these is a good, testable hypothesis for an AI project?

AI is really cool and helpful.The model will be right at least 8 out of 10 times on new photos.I hope my project wins.Robots are the future.

3. What does Google Teachable Machine let you do?

Write your essay for youTrain an image, sound, or pose model in your browser by giving it examples — no codingFind someone's home addressGrade your science fair automatically

4. Why must you test your model on examples it has NEVER seen before?

To make the score look as high as possibleBecause testing on the training examples just shows it memorized them, not that it learned the patternBecause new examples are easierThere is no reason — any test is fine

5. Your hypothesis was '8 out of 10,' but the model scored 7 out of 10. What's the right thing to do?

Quietly remove the photos it got wrong so the score hits 8/10Report the real 7/10 honestly — a guess being wrong is a valid resultPretend you never made a hypothesisSay the AI cheated

6. Which project is the SAFEST and simplest choice for data ethics?

Collecting photos of your classmates' facesRecording strangers' voices at the storeSorting photos of objects (like ripe vs. unripe fruit) or your own posesUsing people's names and addresses

7. What does 'consent' mean in your project plan?

Asking first and getting a clear yes before involving anyone or anything that isn't yoursLetting the AI decide everythingKeeping your project a secretCopying another student's idea

8. Your model got 'ripe' right far more than 'unripe.' What likely caused this bias?

The model dislikes unripe fruitYou probably gave it many more 'ripe' examples than 'unripe' onesBias is impossible in AIThe fruit was too colorful

9. When you do background research and an AI helper gives you a 'fact,' what should you do?

Paste it straight into your project and cite the AICheck it in a trusted source and cite that trusted source — never invent a citationAlways believe it because it sounded sureSkip citations entirely

10. Why put your sources and limitations right on your science-fair poster?

To fill up empty spaceBecause showing your sources and limits makes your project MORE trustworthy — judges can check your workTo hide your real resultsBecause the rules say to use big words