Further comparison of cross-company and within-company effort estimation models for web applications

This paper extends a previous study, using data on 67 Web projects from the Tukutuku database, investigating to what extent a cross-company cost model can be successfully employed to estimate effort for projects that belong to a single company, where no projects from this company were used to build the cross-company model. Our within-company model employed data on 14 Web projects from a single Web company. Our results were similar to those from the previous study, showing that predictions based on the within company model were significantly more accurate than those based on the cross-company model. We also found that predictions were very poor when the within-company cost model was used to estimate effort for 53 Web projects from different companies. We analysed the data using two techniques, forward stepwise regression and case-based reasoning. We found estimates produced using stepwise regression models were better for the within company model while case-based reasoning predictions were better for the cross-company model.