Sensitivity Analysis of Key Cost Performance Determinants Using Extreme Bound Analysis Method

Abstract

The complex nature of construction projects makes achieving positive outcomes challenging, and cost overruns, which frequently occur in all three construction phases, are one of the major impediments to success. When indicators of cost overruns are determined early in a project, however, strategies for reducing or eliminating them can be adopted. Therefore, this study aimed to identify primary cost performance indicators (CPIs) and assess their robustness during the engineering, procurement, and construction (EPC) phases of heavy industrial projects. To accomplish this, an extensive survey was designed and distributed to experienced construction professionals to gather information on tried and true cost performance indicators they had successfully employed. The chi-square test, two-sample t-test, and the Kruskal-Wallis test were used to evaluate their responses, and a final list of cost performance indicators was developed using the all-possible combination regression analysis approach. The robustness or fragility of the indicators was investigated using extreme bounds analysis (EBA), and the results were analyzed using Sala-i-Martin and Leamer’s EBA methods. Because Leamer’s approach concentrates exclusively on the indicators’ extreme boundaries and Sala-i-Martin’s approach considers the entire distribution of the variables, the conclusions were drawn using the Sala-i-Martin method. The results revealed that the three determinants of cost performance indicators in all three construction phases are a project’s alignment, effective management strategies, and the staff turnover rate. Two more robust indicators in all three phases were the fundamental essence of a project and the number of designer/engineering organizations. The study findings will help project managers of construction projects effectively allocate their financial, human, and machinery resources.