Cause-and-Effect Relations Cannot Run Backwards In Time

The researchers based their study on the reasoning that cause-and-effect relations cannot run backwards in time. If one factor causes another, using the cause to predict future values of the effect should yield more successful predictions than using it to predict past values of the effect. In this way, they could calculate strengths of associations between past and future values of body mass index and personality and infer which is likely to be the cause and which to be the consequence.