01When two variables change together, they are correlated. Yet the pattern may arise from coincidence, a hidden variable, or a causal relationship.
02A causal claim needs a plausible mechanism and evidence that rules out competing explanations. Controlled experiments can help, but they also have limits.