Teacher reference · public page
Sampling: does more mean better? — Teacher reference
Prepare the lesson, inspect the reasoning and keep the evidence limits visible.
Before you begin
Frequency, percentages and reading a bar chart. No programming required.
An authored 45-minute sequence using an existing synthetic inquiry. The 200-person school and its activity group are invented, not a real school survey.
Synthetic starting counts: the whole school has 200 people, with 100 choosing A; the activity group has 40 people, with 36 choosing A. These fixed counts support a paper comparison and are not a newly drawn sample.
- Check that the sampling activity opens. Arrange an operator, recorder and questioner; every learner writes an individual prediction first.
- Keep the chart axis at 0% for the main comparison. The synthetic school has 100 A responses out of 200; the group has 36 out of 40.
One-class sequence
- 5 min
Predict before asking
Will asking all 40 group members be closer to the school’s 50% than asking 20 randomly chosen pupils? Explain before running.
Teacher reasoning / look forCollect reasons before displaying results. Distinguish “more people” from “different people have a chance to be selected”.
- 12 min
Keep size, change who is asked
Use the selected seed (default 2026) to compare school-20 and club-20. Record the A count, sample size and percentage. Save both pieces of evidence in the original tool.
Teacher reasoning / look forOnly the sampling frame changes. A sample’s percentage uses its own sample size as the denominator, not 200.
- 10 min
Increase size within each frame
Compare club-40 and school-40 with the earlier cases. Which pupils are still excluded by asking only the group?
Teacher reasoning / look forThe full group gives 36/40 = 90%, while the full school is 50%. Surveying that entire group removes within-group sampling variability but does not fix its mismatch with the school.
- 10 min
Draw again and revise
Use the next-seed control and inspect the listed 30 repeats. Revise “more people always means more accurate”. Do not pick only the nicest result.
Teacher reasoning / look forThe 30 repeats begin at the displayed seed. They describe these deterministic draws, not all possible surveys. Larger random samples generally reduce variability; a single draw need not improve.
- 8 min
Plan a lunch survey
A pupil asks only the basketball team about favourite lunches. Propose a better sampling frame, note possible non-response, then answer the exit question.
Teacher reasoning / look forLook for a whole-grade sampling frame and opportunities across classes/interests. An improved plan still cannot promise an exact population percentage.
All comparison settings
Whole school · 20 people
- Sample size
- 20
- Sampling frame
- Whole school, random
- Chart axis starts at (%)
- 0
- Experiment seed
- 2026
Activity group · 20 people
- Sample size
- 20
- Sampling frame
- Activity group only
- Chart axis starts at (%)
- 0
- Experiment seed
- 2026
Activity group · 40 people
- Sample size
- 40
- Sampling frame
- Activity group only
- Chart axis starts at (%)
- 0
- Experiment seed
- 2026
Whole school · 40 people
- Sample size
- 40
- Sampling frame
- Whole school, random
- Chart axis starts at (%)
- 0
- Experiment seed
- 2026
Open the original activity
The interactive tool and this classroom guide support Traditional Chinese and English.
- Choose Year 7 (初一) first.
- Open Interactive exploration (互動探索), then choose Sampling and charts (抽樣與圖表).
- Apply the settings below manually. Opening a link does not load these settings or replace your saved inquiry.
Device / Python fallback
The existing activity runs in the browser with JavaScript. If a device is unavailable, use the population and group counts below for a paper comparison; that is not a new random draw.
Keep this evidence
- Two saved comparisons with method, size, seed and A percentage.
- A revised claim and a lunch-survey plan with one limitation.
Use Save evidence, Download record and Print in the original inquiry. Its local record stays in this browser; this classroom page does not read, save or import it. Paper responses can be kept separately.
Exit question
If the complete activity group gives 90%, can we claim the whole school is 90%? Why?
No. The observed group is not the whole school. Here the invented school proportion is 50%; increasing the group sample does not include the other pupils.
Misconceptions to discuss
- A bigger sample automatically fixes selection bias.
- A lucky random draw proves the sampling method is always better.
- Changing the chart axis changes the observations.
These statements are misconceptions, not conclusions. Ask for the missing condition or evidence.
Source and further reading
Sequence adapted from the existing Year 7 sampling inquiry (content g7-1), including its four comparison cases, layered hints and transfer question. The 45-minute pacing is an editorial lesson plan, not a measured learning outcome.