A residual plot graphs the residuals (on the y-axis) against the fitted values (on the x-axis). Residual plots can be produced with the rvfplot command. This is a postestimation command, so you need to order it right after your regression analysis.

If the points in the plot are evenly/randomly dispersed around the x-axis, it means that a linear regression is appropriate. If not – and there is some type of pattern (e.g. cone-shaped) emerging in the plot – then you most likely have problems with heteroskedasticity. If the pattern is such that the points are not following the regression line (e.g. showing a curve-linear pattern), you may have problems with non-linearity. Moreover, you will quite clearly see if there are any outliers in the plot.

More information help rvfplot

Practical example

Usually, we would conduct model diagnostics for the full model, so we go back to the example from the multiple linear regression analysis. The quietly option is included in the beginning of the command to suppress the output.

quietly reg gpa cognitive bullied ib1.skipped if pop_linear==1

We then order the residual plot:

rvfplot, yline(0)

It looks pretty OK – apart from the points in the upper left corner and the lower right corner. This suggests that this model might not have any massive problems with heteroskedasticity, non-linearity, or outliers.