The Use of Nonparametric Statistical Inference for Studying the Effects of Construction Waste

Abstract

Abstract: Minimizing construction waste can help achieve the environmental, economic, and social benefits of sustainable construction. Types of waste may include those known as non-value adding activities. Studies on the effects of construction waste on project performance are important to enable mitigation actions. Most of such studies, however, are based on perception surveys. This has led to problems in deriving valid information using parametric methods during the statistical analysis of the response. These problems are mainly related to the assumptions concerning the underlying distribution and the categorical nature of the data. This paper explores a class of nonparametric methods for analyzing survey data concerning the effects of construction waste on project performance. It includes a number of nonparametric tests and post-hoc procedures for repeated measures. Data concerning seven types of construction waste on the generation of material waste from past study are used for this purpose. The results show that consistent outcomes and inferences can be made using different nonparametric methods. A recommendation on which nonparametric methods to use is given.