5 questions0m 0s
Learn the skill
When two groups differ, ask how the groups were formed and what else differed. A measured association can be real while its explanation remains uncertain. The classroom studies here are fictional.
Worked example
If volunteers choosing a guide score higher, the guide might matter, but prior interest could also differ. Choice alone does not isolate a cause.
Original fiction: The Two Reading Corners
In a fictional school trial, more books were borrowed from Corner A than Corner B. Corner A had new signs, but it was also beside the busiest entrance and open for an extra hour. Corner B had older signs and shorter hours. The report counted loans but did not separately test location, hours or signs. It cannot determine which factor caused the difference.