Overall, outliers can have a large effect on your model results. How is your ability to make a prediction when a model is produced containing an outlier? Think about the models produced with and without the outlier. As you have seen with the Outlier and Outlier II worksheets, the predictive capabilities of the model can be very distorted or biased by the presence of an outlier.
Consider this assessment question: If the point indicated in Figure 9 with the arrow were to be measured again and found to be (1.4, 0.10) instead of (1.4, 0.20), how would the regression line respond?
The answer is shown in Figure 10.