Although correlational studies cannot demonstrate causality, on level they are often preferable to , because they do not involve people and possibly exposing them to risk of harm. And on level, even without establishing causation, correlational studies can be useful for .