A Study on the Use of COSMIC BFCs for Effort Estimation

Some studies claim that Base Functional Components (BFCs) contribute to effort at different levels and thus using BFCs instead of Function Points (FP) is better for effort estimation. This study examined the claim with sound filtration and extra-sample error, which were lacked in the past study. As a result, we confirmed that some BFCs did not have significant effect for better estimation. We also demonstrated that a BFCs-based model could achieve better predictive performance than FP-based models. The result contributes to understand importance of BFCs-based models.