The closer the line of best fit is to the points the more accurate it is. The correlation coefficient gives a numerical rank, where 1 is a perfect fit and the closer to 0 it is the less accurate it is.
I need the values of the line at the same x values as the scatterplot points. After you have that take the difference of the value of the line of best fit and the corresponding scatterpoint. if the point is higher than the line then the resiual will be negative and if the point is lower the residual will be positive. Do you have the equation of the line of best fit?
