Fuzzy set theory approach for measuring the performance of relationship-based construction projects in Australia

Abstract

Research into performance measures for relationship-based construction projects becomes crucial because of an increasing trend of client organizations to adopt a relationship-based (or relational contracting) approach to their construction projects worldwide over the last decade. However, few, if any, comprehensive and systematic research studies focus on developing a comprehensive, objective, reliable, and practical performance evaluation model for relationship-based construction projects. A performance index (PI), which comprises eight weighted key performance indicators (KPIs) and a set of corresponding quantitative indicators (QIs) for measuring the performance of relationship-based construction projects have been developed in Australia. The PI and QIs can assist in developing a benchmark for measuring the performance of relationship-based construction projects. However, the establishment of a set of QIs cannot fully solve the subjectivity of performance evaluation. To remedy this deficiency, the aim of this paper is to adopt a fuzzy set theory (FST) approach to establish a well-defined range of quantitative requirements (QRs) for each QI within each of the five performance levels. By using the modified horizontal approach, fuzzy membership functions (FMFs) have been constructed through three various methods, namely constrained regression line with the vertical error method (VEM), the horizontal error method (HEM), and the bisector error method (BEM). It was shown that the results derived from the three methods were similar, and it seems that the BEM is the best technique to construct the FMFs because it considers both the errors created by the residual sum of squares by both vertical and horizontal distances. The newly developed performance evaluation model is not only innovative in nature, but it can also improve the objectiveness, reliability, and practicality in evaluating the performance of relationship-based construction projects.