Authors : O.O. Olugbara , O.O. Ekabua and M.O. Adigun
Abstract: Software Cost Estimation (SCE) is an arduous task in software engineering that has become even more arduous for Business Software Development (BSD) due to changing requirements. The SCE models proposed in literature incorporate software size, approximated by Source Lines of Code (SLoC) or Function Point (FP). However, size estimation for a large portion of software takes time and is as arduous as cost estimation. This study investigates Adaptive Prototyping (Adapro) method that uses probability distribution to estimate software size, given an initial size of its prototype and the number of changing requirements. The method demonstrates the usefulness of Change Impact Analysis (CIA) in SCE. The accuracy of the method was validated with Rayleigh, Exponential and Logistic distributions on real-life projects. The result shows that Logistic and exponential distributions are suitable size estimators when compared to Rayleigh distribution, but Logistic distribution converges faster to the actual software size than exponential distribution.
O.O. Olugbara , O.O. Ekabua and M.O. Adigun , 2007. Estimating Software Cost from Prototyping, Change Impact Analysis and Experience using XP Methodology . Asian Journal of Information Technology, 6: 1010-1014.