Partial Residual Plots

In applied statistics, a partial residual plot is a graphical technique that attempts to show the relationship between a given independent variable and the response variable given that other independent variables are also in the model.

Background

When performing a linear regression with a single independent variable, a scatterplot of the response variables against the independent variable provides a good indication of the nature of the relationship. If there is more than one independent variable, things become more complicated. Although it can still be useful to generate scatterplots of the response variable against each of the independent variables, this does not take into account of the effect of the ofther independent variables in the model.

Definition

Partial residual plots are formed as
Residual + $\beta$i*Xi versus Xi
where
Residual = residuals from the full model
$\beta$i = regression coefficient from the ith independent variable in the full model
Xi = the ith independent variable
Although partial residual plots can offen be useful, they can also fail to indicate the proper relationship. In particular is Xi is highly correlated with any of the other independent variables, the variance indicated by the particial residual plot can be much less than the actual variance.

CCPR plot

The CCPR (component and component-plus-residual) plot is a refinement of the partial residual plot, adding
$\beta$i*Xi versus Xi
This is the “component” part of the plot and is intended to show where the “fitted line” would lie.

Linear relationship analysis based on the plots

A significant difference between the residual line and the component line indicates that the predictor does not have a linear relationship with the dependent variable.