Software benchmarks using function point analysis

The alternative to using source lines of code (SLOC) for costing software projects is to use function points. Functional point analysis (FPA), which was first introduced in 1979, has now been widely accepted as the industry standard for estimating software size and costs. International standard bodies like the International Function Point Users' Group (IFPUG) has been maintaining a repository of data based on projects that were drawn from 14 countries in North America, Europe and Asia Australasia. This paper presents the results of the analysis that was performed on these data together with recommendations such as benchmarks for software projects. Project measures such as functional size, work effort, project delivery rate, speed of delivery and project duration were analysed using the statistical package, Systat. Apart from project duration, all distributions were log-normal, thus suggesting that the results can be used for software benchmarking. Regression analysis on the data further identified correlation of functional size with other project measures that can be used as a basis for planning and estimating software projects. Further work on the repository involves other project measures such as team size and software defects.