Correlation is a measurable statistical tendency for two things to occur together, or to rise and fall together, more (or less) often than chance alone would predict. “Correlation is not causation” is a cliché precisely because it’s so often ignored: a real correlation can come from one thing directly causing the other, from both sharing a hidden common cause, from pure chance in a small sample, or from how the data was collected in the first place.
People can also perceive a correlation that isn’t statistically present at all — a well-documented effect called illusory correlation, in which an association matching a prior expectation gets noticed and remembered far more readily than the disconfirming cases that would show it isn’t really there.