Multivariate Discriminant Analysis for Assessing Residential Development Projects from a Contractor's Perspective

摘要

The Korean construction market is constantly observing residential development projects performed together by developers and contractors. Although the development cost of the overall project is high, there are no standard methods to help contractors make decisions as to whether to participate in these residential development projects or not. This study proposes a simple and efficient model to assist contractors to determine successful projects with a more standardized and objective decision process. To achieve this goal, 31 project assessment items were selected and their weights were updated by AHP method with more sample sizes. Several independent samples t-tests were also carried out to identify the significant factors that affect the contractor's decision for residential development projects. Using multivariate discriminant analysis (MDA), a prediction model was developed with 109 residential development projects. The results indicated that 82% of the original grouped cases had been classified correctly. Cross validation showed that 81.5% of cross-validated grouped cases had been classified correctly. The simple and efficient model suggested in this study will help contractors make better decisions regarding whether to participate in the residential development projects or not.