The independent variable (IV) in a natural experiment is naturally occurring, such as a flood or earthquake, and the behaviour of people affected is either compared to their own behaviour beforehand, when possible, or with a control group who have not encountered the IV.