When psychologists use the term “predictor” they usually mean a correlation between two variables measured at dierent points of time. Suppose you measured variable A two years ago and variable B just recently. When you calculate the correlation between A and B, you nd out that it is suciently large and statistically signicant. In this case you will be able to claim that A is a predictor of B, or that A predicts B. You will also be able to quantify the predictive power (that would be the size of the correlation).

Now suppose that two years ago you also measured C, D and F. You will be able to compare correlations between A and B, C and B, D and B, F and B. You will make inferences, for example: “A predicts B better than D”.

Given what you know about correlations, does the fact that “A is a predictor of B” imply that “A causes B”?