A Controlled Experiment for the Assessment of Aspects: Tracing in an Industrial Context

For years, the Aspect Oriented Software Development (AOSD) community has promoted aspects as a solution to the complexities of implementing crosscutting concerns. But while a lot of research focuses on advancing the state of the art of AOSD, there are few reports on the industrial adoption of AOSD practices. In this paper we report on an experiment we performed-in an industrial setting- to quantify the effort and errors involved in an aspect-based approach to Tracing. We argue that Tracing is a stereotypical aspectoriented problem and as such the results of this experiment are very likely generalizable to other aspects. Participants of the experiment were requested to carry out five simple maintenance scenarios, all related to Tracing. The results of this experiment show an overall 6% reduction in development effort, and 77% reduction in errors, when using an aspectbased approach. Statistical significance analysis confirmed these reductions for a subset of the individual scenarios.