A Hybrid Framework For Accurate Software Effort Estimation In Agile And Traditional Development Models

Abstract

The success of software development projects depends on accurate software effort estimation. Plan-driven frameworks frequently employ conventional models like COCOMO and Function Point Analysis (FPA), however these models frequently lack the adaptability needed to change with the needs of a project. Although agile estimation methods, such as T-shirt sizing and Story Points, offer flexibility, they may compromise early accuracy. The hybrid estimating framework proposed in this paper combines adaptive and deterministic techniques to increase accuracy while preserving adaptability. A major government project and a mobile application development endeavor serve as two case studies used to assess the framework. The findings show that when conventional and agile estimate techniques are combined, accuracy and flexibility are improved over when either strategy is used alone. Performance indicators including usability, accuracy, and adaptability are examined. Finally, guidelines for selecting appropriate estimation techniques based on project characteristics are presented.