By using the number of clusters generated from the above analysis
as an input, a non-hierarchical (K-means) cluster analysis was undertaken
as the second analysis stage to produce the final clusters, which
include the details of the firms within each cluster, and their associated
centroids for each IDR factor.