Effects of Cognitive-driven Development in the Early Stages of the Software Development Life Cycle

The main goal of software design is to continue slicing the code to fit the human mind. A likely reason for that is related to the fact that human work can be improved by a focus on a limited set of data. However, even with advanced practices to support software quality, complex codes continue to be produced, resulting in cognitive overload for the developers. Cognitive-Driven Development (CDD) is an inspiration from cognitive psychology that aims to support the developers in defining a cognitive complexity constraint for the source code. The main idea behind the CDD is keeping the implementation units under this constraint, even with the continuous expansion of software scale. This paper presents an experimental study for verifying the CDD effects in the early stages of development compared to conventional practices. Projects adopted for hiring developers in Java by important Brazilian software companies were chosen. 44 experienced software engineers from the same company attended this experiment and the CDD guided part of them. The projects were evaluated with the following metrics: CBO (Coupling between objects), WMC (Weight Method Class), RFC (Response for a Class), LCOM (Lack of Cohesion of Methods) and LOC (Lines of Code). The result suggests that CDD can guide the developers to achieve better quality levels for the software with lower dispersion for the values of such metrics.