A Bayesian belief network predictive model for construction delay avoidance in the UK

Abstract

Purpose

The purpose of this research is to advise on UK construction delay strategies. Critical delay factors were identified and their interrelationships were explored; in addition, a predictive model was established upon the factors and interrelationships to calculate delay potentials.

Design/methodology/approach

The critical causes were identified by a literature review, verified by an open-ended questionnaire survey and then analysed with 299 samples returned from structured questionnaire surveys. The model consisted of factors screened out by Pearson product–moment correlational coefficient, constructed by a logical reasoning process and then quantified by conducting Bayesian belief networks parameter learning.

Findings

The technical aspect of construction project management was less critical while the managerial aspect became more emphasised. Project factors and client factors present relatively weak impact on construction delay, while contractor factors, contractual arrangement factors and distinctively interaction factors present relatively strong impact.

Research limitations/implications

This research does not differentiate delay types, such as excusable vs non-excusable ones and compensable vs non-compensable ones. The model nodes have been tested to be critical to construction delay, but the model structure is mostly based on previous literature and logical deduction. Further research could be done to accommodate delay types and test the relationships.

Originality/value

This research updates critical delay factor list for the UK construction projects, suggesting general rules for resource allocation concerning delay avoidance. Besides, this research establishes a predictive model, assisting delay avoidance strategies on a case-by-case basis.

Citation

Wang, P., Fenn, P., Wang, K. and Huang, Y. (2022), "A Bayesian belief network predictive model for construction delay avoidance in the UK", Engineering, Construction and Architectural Management, Vol. 29 No. 5, pp. 2011-2026. https://doi.org/10.1108/ECAM-10-2020-0873