Topics · Graphs, Data & Chance
Collecting data
Ask a good question, choose who to ask, and collect data you can trust.
Asking 5 friends and asking the whole school can give different answers.
More ways into this idea
- Play with itSampling LabDraw sample after sample and see when the guess goes wrong
Step into a storyThe Bird CountYou can't check every tree. You can still count the birds.- Make it a gameSample WiselyDesign who gets asked — before any answers arrive
- Make it a gameSample SleuthInvestigate the headline. Can the survey carry the claim?
- Test yourselfA fresh mixed set, whenever you likeExplanations after every question. Nothing is recorded.
Common sticking points
- A common sticking point is judging a survey only by how many people it asked. A survey of 200 people from one corner still cannot hear anyone else — how they were chosen matters more than how many.
- Another is treating whoever happens to be nearby as standing in for everyone. The useful question is not 'were they students?' but 'who had no chance of being picked?'
- A third is reading any two different survey results as proof someone made a mistake. Two fair surveys of the same group will differ a little — that wobble is normal, and it is a different thing from a survey that left people out.
Try this together
Ask each other the same question twice: once asking only the people in one room, then asking a name pulled from a list of everyone at home. Did the answer change? Then write the counts as 'out of' — 3 out of 4 — and check they add up to everybody you asked.
1–5 minutes, no materials to buy, nothing to hand in.
Ontario curriculum details
- D1.1Explain why good samples matter
- D1.2Collect data with a suitable sample
Plain-language summaries. See the full curriculum for every expectation and the official source.