Residuals vs. Fit Values Plots
- To produce a scatterplot of residuals by fit
values for the Linear model, from the menus choose:
Figure 1. Chart Builder - Select the Scatter/Dot gallery and choose Simple Scatter.
- Select Error for sales with advert from CURVEFIT, MOD_1 LINEAR as the y variable and Fit for sales with advert from CURVEFIT, MOD_1 LINEAR as the x variable.
- Click OK.
Figure 2. Scatterplot of residuals by fit values for linear modell This plot reinforces your suspicions from the curve fit plot. There is a clear "inverted U" shape to the points, which means that there is a pattern in the data that is not captured by the Linear model.
Figure 3. Chart Builder - To produce a scatterplot of residuals by fit values for the Quadratic model, recall the Chart Builder.
- Click Reset to restore the default settings.
- Select the Scatter/Dot gallery and choose Simple Scatter.
- Select Error for sales with advert from CURVEFIT, MOD_1 QUADRATIC as the y variable and Fit for sales with advert from CURVEFIT, MOD_1 QUADRATIC as the x variable.
- Click OK.

These residuals do not show a pattern, thus the Quadratic model is acceptable in the sense the residuals are independent of the fit values.