i_ Agglomerative hierarchical cluster analysis uses a stepNOTE
wise clustering algorithm to gatber firms into clusters. At
each stage of agglomeration, the researcher is able to determine
tbe degree to wbicb cluster positions bave altered
as a result of successive cluster combinations. However, a
disadvantage of bierarchical clustering is tbat firms are
not able to switch clusters once assigned, and early suboptimal
cluster assignment persists to the end of the clustering
process. A clustering procedure that allows for case
reassignment is the k-means metbod.