What are the possible effects of outliers on a regression analysis?
I know that outliers can dramatically change the magnitude of the regression coefficients and even change the direction of the coefficient signs. I am wondering however if there are set effects (a list) which are effects of outliers?
May have severe influence on fitted line ⇒ systematic lack of fit at other observations
May inflate residual variance
⇒ low R2
⇒ non-significant slope β1
Observations with large standardized residuals (numerically larger than 2) are flagged by Minitab