When a Number Needs a Neighbor
An impressive increase becomes a different story when the total walks into the room.
A closed case, not a coffee verdict
This case is entirely fictional: both cafés, their respondents and every number were invented. You can finish all three clues and read the resolution without registering. Your first-choice tally records the reasoning you tried, not your intelligence or whether you like either café.
The resolution
Café A has more happy responses in absolute numbers. Café B has a larger happy-response share among its respondents. Those statements are compatible. The recruitment methods differ, so neither share establishes which café serves better coffee or has happier customers overall. The poster’s missing information was the context.
Your next investigation
A fairer comparison would start by defining the question and collecting comparable samples using the same method and period. Before sharing a persuasive count, ask: out of how many, and who was asked? Follow the linked numbers story for another calculation, or the sampling story to examine who a large online response can leave out.
Sources & editorial notes
- Pew Research Center: What are nonprobability surveys?
- Seeing Theory, Brown University: Basic Probability (archived)
Rights: Original Oddoscope text and code-authored SVG illustration. External references are linked for attribution, not reproduced or licensed by the presence of a link.
