Correlation is not causation
SAT glossary · Course Foundations
Definition
An association can arise from causation in either direction, a third variable influencing both, or coincidence; only a randomized controlled experiment licenses a causal claim. Tested on both sections: an option asserting causation from an observational study is wrong because of the design, not the plausibility of the story.
Correlation is not causation in plain English
Just because two things tend to happen together (correlation) doesn't mean one is causing the other. There are actually several reasons an association could show up in data: A could genuinely cause B, B could cause A, some hidden third factor could be secretly causing BOTH of them, or it could simply be a coincidence. The only way to responsibly claim 'A causes B' is if researchers ran a proper randomized controlled experiment, where they actively controlled who got what treatment. So on the test, if an option claims something caused something else, but the study described was just observing people without controlling anything, that option is automatically wrong — not because the idea is unbelievable, but purely because of how the study was designed.
Sign up free: all 157 SAT terms as flashcards, plus practice questions with full explanations →