A hybrid PLS-SEM-ANN approach to COQ optimization

Construction industry faces challenges in making objective decisions due to the complex financial implications of quality management systems (QMS). To address these, QMS integrates cost of quality (COQ), a strategy derived from the manufacturing industry. However, accurately estimating hidden factors (HF) required for COQ design and execution remains a major challenge, impacting visible factors (VF). Therefore, a conceptual model is developed to overcome such challenges by transforming traditional categories of COQ and analyzing the interrelationships between these aspects through the perspective of complexity theory. Data was gathered from 142 quality experts through purposive sampling and a semi-structured questionnaire. A two-stage analysis was conducted using partial least squares structural equation modeling (PLS-SEM) and artificial neural networks (ANN) to address both linear and non-linear regression. The proposed relationships were statistically significant, as assessed through the PLS-SEM model. Additionally, ANN analysis further elucidated these findings, indicating that internal failure (HF) represents the strongest predictor in achieving optimal quality, followed by prevention (HF), and external failure (HF). In contrast, both external failure (HF) and preventive actions (HF) were found to have a comparatively lesser impact. The study identified compensatory, non-compensatory, linear, and non-linear relationships among HF, VF, and quality, and advancing effective QMS strategies.