Integrated DEMATEL and ANP-Based Framework to Model Construction Labor Productivity

Abstract

Accurately modeling productivity is essential for ensuring that the results of construction process simulation models align with actual practice. Since collecting quantitative data is challenging and expensive, productivity models are often built using the information provided by industry experts. The subjectivity of this information, however, commonly results in oversimplified or inadequate productivity models. To address this challenge, this research has developed a novel framework that reduces the subjectivity associated with labor productivity modeling by identifying interrelationships between factors affecting productivity that individual subject experts may have overlooked. A Decision-Making Trial and Evaluation Laboratory (DEMATEL) is used to identify relationships (i.e., dependencies) between factors, which is integrated with an Analytic Network Process (ANP)-based approach to determine the strength (i.e., weight) of each relationship. Results can support decision-making or feed productivity data to simulation, empirical, or dynamic models of construction systems. Outputs of the proposed method yield higher-quality inputs for productivity modeling-based decision-support systems compared to traditional input preparation approaches. The effectiveness of the framework is demonstrated through an illustrative example.