Training data and bias

Trace unfair AI outputs back to biased or incomplete training data, and see how an overall accuracy score can hide the problem.

⏱ 20 min❓ 8 questions

What you'll be able to do

✔I can name and recognise common sources of bias in training data
✔I can calculate per-group accuracy and explain why an overall score can hide unfairness
✔I can suggest practical ways to detect and reduce data bias

How this works

1Read the short explanations first — they teach you what you need before each question.
2Answer each question, then tap Check. A right answer moves you on automatically; if it's wrong, you'll see a friendly explanation of why.
3Finish all 3 steps to see your score — sign in and subscribe to unlock the full lesson and save it.

What's inside

8 questions and 1 quick teaching card, including:

🔢Number answer — Work out the answer and type the number.

👪Grown-up tip: sit alongside younger learners for the first few lessons. Read the questions together and chat about the explanations — that's where the real learning sticks.

Curriculum: AC9TDI10P08