If a study reports a correlation between two variables, what is the safest inference?

If a study reports a correlation between two variables, what is the safest inference?

• A. X causes Y
• B. Y causes X
• C. They are related, but causality is not established
• D. They are not related

Answer: (C)

The main idea here is that correlation shows association, not causation. When two variables correlate, they move together in a related way, but that does not tell you which one, if any, causes the other. There could be a third factor influencing both, the direction could be reverse, or the relationship could be coincidental or non-linear. Therefore, the safest inference is that the variables are related, but causality is not established. To make a causal claim, stronger evidence from experimental manipulation or methods that address direction and confounding would be needed.

The main idea here is that correlation shows association, not causation. When two variables correlate, they move together in a related way, but that does not tell you which one, if any, causes the other. There could be a third factor influencing both, the direction could be reverse, or the relationship could be coincidental or non-linear. Therefore, the safest inference is that the variables are related, but causality is not established. To make a causal claim, stronger evidence from experimental manipulation or methods that address direction and confounding would be needed.