K-means cluster analysis uses a squared Euclidean distance
to place firms closest to the cluster center witb cbaracteristics
most similar to the firm's. Two key issues that
arise wben k-means cluster analysis is employed are the
specification of initial cluster centers and the number of
clusters to be specified. Altbough k-means clustering allows
for repeated case reassignment to maximize between
centroid distance and minimize within cluster differences,
k-means clustering does not incorporate a procedure
for specifying initial cluster centers.