Classifying construction contractors: A case study using cluster analysis

It is widely accepted in construction management literature that superlative contractor selection criteria are: contractor ability to complete a project on time, within budgeted cost and to expected quality standards. Hence, contractor evaluation and selection models with the ability to highlight these attributes (i.e. help the selection decision) should be fully exploited. To date, such models have evolved based predominantly on multi-attribute analysis, case-based reasoning, and discriminant analysis, but there is scope for investigation of alternative strategies including: fuzzy set theory; neural networks; regression techniques; and cluster analysis. This paper concentrates on the latter by applying cluster analysis to real-life contractor selection data. Results indicate that the technique will simultaneously classify large numbers of contractors while identifying the most significant discriminating criteria among them. These characteristics offer potential for rationalization of contractor evaluation, classification and selection.