On the effects of programming and testing skills on external quality and productivity in a test-driven development context

Background: In previous studies, a model was proposed that investigated how the developers' unit testing effort impacted their productivity as well as the external quality of the software they developed. Goal: The aim of this study is to enhance the proposed model by considering two additional factors related to the expertise of developers: programming and unit testing skills. The possibility of including such skills in a model that represents the relationship that testing effort has with the developer's productivity and the product's external quality was investigated. Method: Data collected from a test-first development task in academic setting was used in order to gauge the relationship between testing effort, external quality, and productivity. Furthermore, Analysis of Covariance (ANCOVA) was utilized to check the impact of developers' skills on productivity and quality. Result: The results obtained in previous studies were confirmed: there exists a positive effect of testing effort on productivity, but not on quality. Moreover, the developers' skills have an impact on productivity but none on external quality. Conclusion: Productivity improves with testing effort, a result consistent across previous, similar studies. The role of existing skills is a relevant factor in studying the effects of developers' unit testing effort on productivity. Nevertheless, more investigations are needed regarding the relationship between unit testing effort and external quality.