PROJECT COST PREDICTION MODEL USING PRINCIPAL COMPONENT REGRESSION FOR PUBLIC BUILDING PROJECTS IN NIGERIA

Abstract

Major problem in Nigeria construction industry is that building contracts are completed at sums much higher than estimated cost, hence the need to develop predictive cost model that capture factors affecting project cost using principal components regression, through set objectives: to identify factors contributing to project cost; examine the importance of the factors and develop cost predictive model. Literature review on the study indicated that nature of clients, professional involved in a project and their decision regarding design, function, duration, technology and implementation have significant effect on the overall project cost. Data for the study are obtained through random sampling of public building projects completed in Nigeria after 1995. The study identifies six most significant factors to project cost among the design related variables as: Level of design complexity; level of construction complexity; level of technological advancement; percentage of repetitive element; presence of special issues and scope of work. Three factors among time/cost related factors as Importance for project to be delivered; time allowed by the client and his representative for bid evaluation; need for the project to be completed. Client, consultant and contractor’s experience on similar project; adequacy of contractor’s plants and equipments are most significant among project parties experience related factors. The selected factors were used for cost predictive model.