A Proposed Model for Embedding Risk Proportion in Software Development Effort Estimation

Abstract

The risk of project implementation failure has always been a primary concern for software development team managers. Previous studies have recommended that risk should be considered as a significant parameter in software development effort estimation. The aim of this research is to create a modelling of risk proportion in software development projects based on the risk classification from previous studies. The analysis method used in this research is quantitative, based on the observation of the risk proportion values against the actual effort. Seventeen software projects were analysed to derive the risk proportion model. After the experimental process of calculating risk occurrence, the accuracy of the risk proportion in relation to the real effort in the 17 projects was found to be 91.97%. This risk proportion model is then incorporated as an additional formulation in estimating the economics of software development projects using commonly encountered methods, especially those related to non-technical or environmental complexity factors. In future research, the integration of risk proportion with complexity factor data will be pursued.