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.