Abstract
The construction industry (CI) is one of the most hazardous where the specific behaviours of different stakeholders can be contrary to the conditions and safety behaviours in the construction process. Understanding the nature of this kind of uncertainty can reduce the level of inconsistency that reinforces the safety management system in CI. The purpose of this paper is to present an interval-valued intuitionistic fuzzy-improved score function and weighted divergence based approximation (IVIF-ISF-WDBA) approach for ranking main stakeholders with clear duties that can influence workers' unsafe behaviours as well as risk reduction in final decisions. In this method, a score matrix is compiled based on improved score performance and IVIF decision matrix, and then a linear programming model is developed to determine the weight of unknown criteria. In addition, a divergence measure has been developed to find out the degree of performance of the alternatives. Next, we perform comparative and sensitivity analyses with different proposed methods and different weight criteria sets to determine the stability of the developed approach. The results indicated that the most effective safety factors (ESFs) are safety climate factor3 (SCF3) (safety management rules and regulations) at 45%, and safety perceptual factor2 (SPF2) (workers' preferences and behaviours related to safety category) at 25%. The LP-model ranked client, designer, construction manager, and contractor as the authoritative bodies that should carry out extensive inspections of ESFs, namely, SCFs, SPFs, and workplace condition factors (WCFs) in the preconstruction phase. However, these priorities can change during the construction phase among the contractor, construction manager, supervisor, client, and designer. This study concluded that workers' unsafe behaviours could be linked to stakeholder duties in complex and dynamic conditions of the CI.