A Comparative Evaluation of Firms’ Behavioral Environments Based on Two Forms of Cluster Analysis

Abstract

This paper utilizes hierarchical and non-hierarchical Cluster Analysis to evaluate entrepreneurs' perception of regional economic milieux, and tests hypotheses concerning differences in entrepreneurial perception within an undifferentiated population of firms and among groups of product-related firms. The attitudes of 107 manufacturers in Calgary. Alberta, Canada toward fifty statements deemed to reflect the essence of their regional manufacturing environment are evaluated for evidence of the relative merits of these two forms of Cluster Analysis for application in human-geographical research.The paper concludes that non-hierarchical Cluster Analysis has many advantages for evaluation of multivariate data sets in human geography that have not been clearly recognized until now, and that non-hierarchical Cluster Analysis overcomes some of the limitations such as “chaining”, small-group isolation, and intra-group dissimilarity which are inherent to hierarchical Cluster Analysis. The paper suggests that many of the strengths of non-hierarchical Cluster Analysis tentatively identified here merit extensive additional elucidation and investigation.