Estimating the time needed for execution of two phases in software development

Matija Krajnc (2010) Estimating the time needed for execution of two phases in software development. MSc thesis.

	HTML Download (3396Kb)
	PDF Download (3396Kb)

Abstract

The following Masters thesis studies the acceptance of simplified use case point method for time (effort) estimation of execution, analysis and design phase in software development. With this method I wanted to estimate the required time for execution of a particular phase. The aim of this thesis is to provide a simple and fast way for efficient time estimation at the beginning of software development project for particular development phases, like the analysis and design phase. WBS (Work Breakdown Structure) was used as a blueprint for implementation of the project and as a basis for estimation of the required time for analysis and design phase. With WBS the smallest units of work which need to be implemented in the software were identified and counted. These units of work were used as a basic input for time (effort) estimation. In the next step theoretical background for different, generally used methods for time (effort) estimation were analyzed and at the end acceptable methods for effort estimation in software development projects were analyzed in greater detail. Three acceptable methods were identified within analysis: function point method, use case point method and simplified use case point method. The simplified use case point method was proposed and used in this research because of the following features: the ability to estimate effort in the earlier phases of project, the simplicity of use and the ability to estimate the effort needed for particular phase of software development project. The main result of my thesis is a confirmation that the simplified use case point method can provide acceptable effort estimation for analysis and design phase in software development projects. In the most projects used in this research the difference (error) between effort estimation and actual effort for analysis and design phase was smaller than 25 percent. Additional result of my thesis is also a finding that the quality of effort estimation with the simplified use case method more depends on the number of used projects (or projects history) than the similarity between the used projects in this research. It's recommended that the influence of the similarity factor is to be researched in greater detail in the future. The research was carried out in the period of three years on five successfully executed software projects. In the last, the fifth project the method provided acceptable effort estimation for analysis and design phase. These estimations were then used by the project manager for quality and efficient project implementation. The used method has been proven as acceptable.

Actions (login required)