Effort and cost estimations are the basis for project management to calculate the costs of software projects in order to create a budget offer long before the project's realization starts. At this time details about the project are rarely good enough for correct estimations. The basic requirement for a good estimation is the knowledge of the cost driving factors within the software development cycle. The outcome of an analysis regarding the cost driving factors are the expected demands for an effort estimation. A recalculation of project effort estimations using data from previous projects can provide pros and cons of current estimation procedures. In my thesis I work in this direction. For this purpose data from the LOGIM GesmbH company were taken. Function Point methods like IFPUG, Mark II and Full Function Points were tested. Furthermore, the Object Point and the COCOMO II method were analyzed with respect to the available data. This data comprises time recordings and source lines of code for various projects and thus yield to a comparison of the different methods with respect to their ability of computing good cost estimations. The analysis in my work leads to the main conclusion that the used method itself does not have an huge influence of the quality of estimation. More important is the handling of the chosen method which can be ensured by using clear definitions, a well educated personal, pre-adjusted parameters, and influence factors.
|Effective start/end date||1/09/02 → 31/03/03|