01 · Learn the idea
A data set is more than one mean
The following diameters are invented for analysis practice, not measured results. Surface P: 5, 6, 5, 4, 5 mm. Surface Q: 8, 7, 9, 8, 13 mm. A notebook note says Q’s last trial accidentally received two drops. P’s mean is 5 mm and range 2 mm. Q’s mean is 9 mm and range 6 mm.
Keep the original record. Report a second analysis excluding the documented procedural error: Q’s first four trials average 8 mm and have range 2 mm. Explain the exclusion rule transparently and repeat the failed trial if possible. Never delete an inconvenient result merely because it disagrees with your prediction.
02 · Build understanding
Challenge the claim respectfully
Claim 1: “Surface Q always produces a diameter of 9 mm.” Claim 2: “Under these practice conditions, Q’s valid trials show greater spreading than P’s, but more controlled trials are needed.” Compare these claims with the observations. The second accounts for variation and the documented error.
Individually write a claim-evidence-reasoning paragraph. In a group, present a graph with the unusual point visibly marked. A listener asks one evidence question; the speaker paraphrases it and responds with a number or procedural detail. Use wording such as “Which measurement supports that?” and “Could drop size explain part of the difference?” Critique the explanation, not the person.
03 · Build understanding
Improve the next test
Propose a consistent drop-volume method, a level tray, a fixed measurement time, and a way to reduce observer differences. Explain which source of error each change addresses. Greater sample size alone cannot correct a biased measurement method.
Apply and explain
Your turn
Calculate both means and ranges, show the original and justified adjusted Q analysis, then defend a revised explanation in writing and a small-group presentation. Record a peer’s evidence-based challenge and your response. Propose two design improvements.
Use paper or explain your thinking aloud before opening the solution.
Check your reasoning
P mean 5/range 2; Q original mean 9/range 6; Q valid subset mean 8/range 2. The documented double drop warrants a transparent separate analysis and repeat, not silent deletion. A conditional spreading comparison fits better than an “always 9” claim.
- Correct calculations and retained original data
- Exclusion justified by procedure rather than preference
- Claim cites evidence and acknowledges limits
- Respectful argument and response documented
- Improvements address named errors
Ready to practice?
Try the questions, get feedback, and review your reasoning.
Evaluate trials and revise an explanation practice →