You can display Mahalanobis distance using either Principal Components or Discriminant Analysis.

In discriminant analysis, Minitab uses the pooled covariance matrix to calculate the Mahalanobis distance. This considers the classification that each observation is grouped into. Because principle components analysis does not classify the observation into groups, it uses the covariance matrix of all the data.

- Display the Mahalanobis distance between an observation and the centroid using Principal Components.
- Choose Storage. and click
- In Distances, enter the column that you want to store the distances in.
- Click OK in each dialog box.

- Display the Mahalanobis distance between an observation and the group centroid using Discriminant Analysis.
- Choose Options. and click
- Under Display of Results, choose Above plus complete classification summary. Click OK.

In the Summary of Classified Observations table, the Squared Distance is the Mahalanobis Distance (D squared) statistic, calculated for each observation from each group centroid.