Association is not causation
For a group of athletes, an observational data set contains weekly training distance (km) and fitness test total (points)
The observations show a strong positive association: higher weekly training distance tends to occur with higher fitness test total