Notes:

• An outlier is often an extreme value in the response variable. One can often observe its existence as a large or small value in the residual plot.
• An influential observation is often an extreme value in the explanatory variable.
• An observation is considered an influential observation if it’s effect is to pull the regression line towards it. Consequently it’s existence will not show up in a residual plot.

** One observation’s inclusion or exclusion from the regression should not change the fit of the regression line dramatically.