A 10–15 minute investigation
A chart can be accurate and still tell a different story.
You are joining the Night School Data Desk. Test three fictional cases: a graph frame, a sample, and a correlation. Your notebook stays in this browser tab.
Investigation 01 · Graph framing
Same cups. Different visual emphasis.
A fictional kiosk recorded cups sold from Monday to Friday. Move the vertical axis minimum. The values and the calculated change must stay the same.
Choose the frame
Full context · zero baseline
Adjustable frame
Try the slider, then record what changed and what did not.
Data stays fixed
| Day | Mon | Tue | Wed | Thu | Fri |
|---|
Investigation 02 · Sampling
Who made it into the sample?
A fictional book fair had 120 visitors: 30 club members who stayed 60–89 minutes, and 90 walk-ins who stayed 10–39 minutes. Compare a convenient club sample with a simple random sample.
Population baseline
What did your sample show?
Hint: a larger sample reduces random noise, but it cannot repair a method that leaves people out.
Investigation 03 · Correlation
A pattern is not yet a cause.
These fictional fair records pair banner size (metres) with visitor count. The fair's scale is a plausible third variable: larger fairs can have both bigger banners and more visitors.
Circle: small · Square: medium · Triangle: large
Hint: Ask what could influence both variables before calling one a cause.
Filter the data, then choose a cautious conclusion.
Records behind the plot
Final case · Transfer
The school garden headline.
Full context · 0–30 kg
Close-up · 20–30 kg
This is fictional synthetic data: two average harvest reports from a 20-of-100 volunteer-club plot selection. The earlier and later reports are not a census of all plots.
Build a careful response
A strong answer names the observed difference, the sampling limit, and what the close-up frame cannot prove.
Self-review checklist
Use this checklist for your own wording. The page does not pretend to score open text.
Your record