The graph below displays the heights of children from age 5-15. Why shouldn't we use the regression line shown to predict a child height of age 13? The shape is non-linear, so we should not use a line...

Extracted text: The graph below displays the heights of children from age 5-15. Why shouldn't we use the regression line shown to predict a child height of age 13? The shape is non-linear, so we should not use a line to predict values. An outlier is skewing our regression line, making it inaccurate. 13 is not in our data's range/domain so this would be extrapolation. We could use the line to accurately predict height.