PhD Theses in Experimental Software Engineering, Band 66
Hrsg.: Frank Bomarius, Peter Liggesmeyer, Dieter Rombach; Fraunhofer IESE, Kaiserslautern
2020, 243 S., num. illus. and tab., Softcover
Effort estimation is used for planning and managing all phases of software development. Traditional estimation methods rely on having a mostly “complete” and “fixed” specification of a system. In agile software development, cost and effort estimation is challenging, as the requirements cannot be specified entirely upfront and are evolved as development progress and users interact with the product. Current practice in this context relies heavily on human judgment which has several limitations. It is based on limited information (subjective opinion), hampered by wishful thinking and prone to human judgment bias (individual and group effects). Estimates produced by these methods are often inaccurate. Experts recognize the need and importance of more objective information that includes historical data from previous iterations. Moreover, current estimation methods do not objectively consider the potential impact of a change on existing software and custom factors that contribute to the effort overhead. Thus, considerable improvement potential exists concerning systematic effort estimation in this environment and marks the contribution of this research to the body of knowledge.