MODELLING REWORK COST FOR TETFUND BUILDING CONSTRUCTION PROJECTS IN NIGERIA

Abstract

Accurate predictions of rework cost enable contractors to evaluate how successful their projects are likely to be, thus, improve their operations and productivity. Studies have developed project scope-based and non-scope-based rework cost predictive models. However, researchers have criticized the accuracy of these models because of their weaknesses in giving individualistic treatments to project scope-based and non- scope-based rework cost influencing factors. This study therefore aimed at developing rework cost predictive models that combine the synergistic effects of both project scope-based and non-scope-based rework cost factors. 440 sample size was drawn from 2,638 population of Tertiary Education Trust Fund (TETfund) building projects constructed between 2009 to 2015.  Data of 287 (65%) of 440 (sample size) of  completed TETfund building projects on project scope factors as well as the extent of influence of the non-scope factors on cost of rework in TETfund building projects was collected. Principal Component Analysis (PCA) and Multiple Linear Regression (MLR) were used in developing an integrated rework cost predictive model. More accuracy was achieved in the integrated model through the incorporation of impact of project scope and non-scope influencing factors into rework cost prediction model, this increased the variability in cost of rework to R² = 0.759 (75.9%), while errors in rework cost predictions reduced by as much as 9.75%.The model serve as useful tool used to enhance accuracy of rework cost predictions for TETfund building contractors, thus reduce cost of rework for building construction projects.