Risk factors in the refurbishment design process: the case of Malaysia

Abstract

PurposeRefurbishment projects are mainly due to building physical improvements and extensive repair work. However, it is more difficult to manage refurbishment compared to new building projects, due to risk factors. Therefore, this paper aims to identify the risk variables in the refurbishment design process and how they influence the design performance variables.Design/methodology/approachThe study has been designed using predominantly the triangulation technique, which consists of quantitative and qualitative approaches to ensure appropriate validation of the data and enabled cross-verification of multiple sources. The research subject was 82 refurbishment projects with the contract value of more than RM 500,000.00. This provided a better view of the risk aspects that affect the design performance of refurbishment projects. Meanwhile, the method of procurement used for the selected refurbishment projects were either traditional or design-and-build. These procurement methods were found to be the most commonly used in refurbishment projects in Malaysia.FindingsThis study concludes that the design phase of refurbishment projects is adversely impacted by the risk; whereby variables consisting of services, structural aspects, design fees and clients' attributes affect the design process. Moreover, a prediction model is formulated to estimate the provisional sum of the refurbishment projects. There were two prediction models that have been generated using Statistical Package for Social Science (SPSS). One independent variable of model 1 is not a significant predictor. Thus, another model was produced by using the stepwise method.Originality/valueDespite the fact that the topic of refurbishment has been thoroughly explored by academics and professionals in the past, there is a clear gap which requires clarification in the field of design process and its link to performance of a project. Hence, this paper adds new value to the body of knowledge and widens awareness among the subject audience. There were two prediction models that have been generated using SPSS. Both prediction models were presented in the research.