Investigation of the effort data consistency in the ISBSG repository

To develop adequate software project estimation models using statistical techniques, the consistency of historical data is important. This paper investigates this issue by looking into the consistency of the information contained in one of the most important fields in the International Software Benchmarking Standards Group (ISBSG) repository, that is, the project effort data field. This paper also presents an example of how effort data from projects that include a large number of project phases can be used for extrapolation, through a normalization process, to projects with fewer phases. The ISBSG organization has attempted to tackle this issue on the variability of phases included in the project effort field by deriving a normalized work effort field. This paper investigates this problem and reports on a number of related issues.