This graph plots the residuals versus the fitted values. The residuals should be scattered randomly about zero. There should be no special features or patterns in the residuals, such as:

·    Outliers - An outlier is an unusually large or small value. Outliers in residuals plot can indicate outliers in the data set. Usually, you can more easily spot an outlier in a residual plot than in a plot of the data.

·    Nonconstant variance - Analysis of variance assumes that the residuals for each level have the same population variance. Thus, the spread of the residuals about zero should be similar for each fitted value.

For the door lock data, the residuals for the fitted usability and quality values seem scattered randomly about zero.