When using the model to make predictions, care must be taken to only choose values of X that are in the range of X values of the data. In the rainfall/sales example, the values of X range ...When using the model to make predictions, care must be taken to only choose values of X that are in the range of X values of the data. In the rainfall/sales example, the values of X range from 10 to 40 inches of rainfall. Choosing a value of X outside this range is called extrapolation and could lead to invalid results. For example, if we use the model to predict sales for a city with 80 inches of rainfall, we get an impossible negative result for sales: ˆY=45.647−.767X
