There are several methods for forming clusters. In this study, a hierarchical agglomerate method was used to compute initial cluster seeds for a non hierarchical method (K-means clustering). In this way, the advantages of hierarchical methods are complemented by the ability of the non hierarchical methods to ‘‘fine-tune’’ the results by allowing the switching of cluster membership (Hair et al., 1998). To prevent different scale intervals from affecting the clustering procedures, data were standardized (with a mean of 0 and a standard deviation of 1).