Indian Journal of Science and Technology

Abstract

Objective: we aim to enhance the agile sizing units to lessen the subjectivity of estimations and the dependency of the personal experience by introducing factors that would provide a clear basis and estimation guidance for size estimation, especially Epics estimation yet maintaining the agility of size estimation. Methods/Analysis: we surveyed a number of approaches that used to estimate software size in traditional software development. However, these size approaches have limitations and may not be suitable for agile software projects. In agile projects, size is estimated based on the experience of the team by Story Point (SP). SP is the common sizing unit that is assigned based on the relative size of the User Story. However, Story Point as it stands is subjective and is not defined in a standard way, and is ill-suited to large projects. Findings: in this research, we propose a framework and introduce a new sizing unit for estimating Epics called Enhanced Story Point sizing unit (ESP). Enhanced Story Point is calculated based on three factors affecting size estimation especially in the early phase of the software projects i.e. requirements phase. These factors are Uncertainty, Functionality, and Complexity of project requirements. We applied our framework on three case studies that used an agile process in order to manage their work and each team used Story Points unit as a unit of the project size. We evaluated our results by using two measures; Effort Variance and Magnitude Relative Error (MRE) for each Epics. Then we compared the results before and after using our framework i.e. when using SP and ESP. Improvement: an empirical evaluation demonstrates that our proposed work outperforms traditional estimation by Story Point in effort variance, especially in case of low-experience in estimation using existing Story Points of the team. 

Keywords: Complexity, Epic, Estimation size, Story Points, Uncertainty, User Stories