For the paint-sprayer data, use some suitable graphics to identify any observations that you think are outliers. Construct chi-square plots of the generalized distances after removing the outlying observations both for the raw data and log-transformed data.