Please help
Have you given some thoughts?
i thought it was A
One single excessive residual residual does not spoil the soup. Often this is taken care of by rejecting it as an outlier. Obviously we cannot say that the residuals are nearly zero either. When the line has equal number of points on each side, with a systematic deviation, it's time to think of an alternative. By systematic I mean that the points are over at the beginning, under in the middle, and over at the end, which indicates some kind of a curve.
|dw:1406080510320:dw| Points here are also on each side of the regression line. It does not show systematic deviation, so does not indicate a curve as a better solution.
I am confident that you can pick your choice. If you have further discussions, kindly post or tag me.
Join our real-time social learning platform and learn together with your friends!