A q-q plot of regression residuals, as in Figure 4.13, is often a part of regression analysis.

Why might a q-q plot of the residuals be of interest?

What kind of regression residuals should be used in a q-q plot? Why?

Figure 4.13:

Transcribed Image Text:

-0.15 Residuals -0.05 0.05 0.08 0.00 Cook's distance 0.04 0 o -0.15 Residuals vs Fitted -0.05 0.05 0.15 Fitted values Cook's distance 54 3 50 100 Obs. number 150 10 Standardized residuals 2 0 [ Standardized residuals 4 T 2- 0 2 063 062 -2 -1 Normal Q-Q 0 1 Theoretical Quantiles Residuals vs Leverage 038 054 640 10 2 705 +0.5 -- Cook's distance T 0.00 0.05 0.10 0.15 0.20 Leverage