Whenever there is ANY spread about the *linear regression line*, for any above , the values will increase by less than the increase in .

For example, for one standard deviation above , the corresponding predicted value (y-hat) will be less than one standard deviation above .

Similarly, for one standard deviation less than , the corresponding predicted (y-hat) will decrease by less than one standard deviation below .

This is called the *regression effect*.

It happens because part of the *unexplained* variation in is *explained* by the change of (the regression).

The regression line is ALWAYS flatter than the SD-line. (Both lines pass through the point of averages, .