Estimating time performance for building construction projects in Vietnam

Abstract

Time performance is one of the most important aspects to be benchmarked in project management. Many factors arise during the project implementation process affecting the project time. Identifying the problems that are significant to executing construction schedule and incorporating them into a comprehensive model will be useful in project management. Based on a data set of 70 projects (outliers discarded), the multiple regression technique is applied to integrate significant variables into a model used to predict or evaluate the project time index. The result suggests that six variables, namely underground site condition, project management works, estimating works, competency of subcontractor(s), accuracy and completeness of design, and owner’s project financing emerged as important factors in regression model. The constructed regression model has a good performance when compared with an artificial neural network model developed for purpose of cross validation.