Early evaluation of technical debt impact on maintainability

• Published in: The Journal of systems and software Vol. 142; pp. 92 - 114
• Main Authors: Conejero, José M., Rodríguez-Echeverría, Roberto, Hernández, Juan, Clemente, Pedro J., Ortiz-Caraballo, Carmen, Jurado, Elena, Sánchez-Figueroa, Fernando
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.08.2018
• Subjects: Empirical evaluation; Maintainability; Modularity anomalies; Requirements; Technical Debt indicator; Technical Debt indicator; Modularity anomalies; Maintainability; Empirical evaluation; Requirements
• ISSN: 0164-1212; 1873-1228
• DOI: 10.1016/j.jss.2018.04.035

•This work presents an empirical analysis that relates modularity anomalies with Technical Debt.•The work is based on the application of a framework to identify modularity anomalies.•The study has been applied to three different software product lines.•The study provides information to anticipate refactoring decisions (reducing interest). It is widely claimed that Technical Debt is related to quality problems being often produced by poor processes, lack of verification or basic incompetence. Several techniques have been proposed to detect Technical Debt in source code, as identification of modularity violations, code smells or grime buildups. These approaches have been used to empirically demonstrate the relation among Technical Debt indicators and quality harms. However, these works are mainly focused on programming level, when the system has already been implemented. There may also be sources of Technical Debt in non-code artifacts, e.g. requirements, and its identification may provide important information to move refactoring efforts to previous stages and reduce future Technical Debt interest. This paper presents an empirical study to evaluate whether modularity anomalies at requirements level are directly related to maintainability attributes affecting systems quality and increasing, thus, system's interest. The study relies on a framework that allows the identification of modularity anomalies and its quantification by using modularity metrics. Maintainability metrics are also used to assess dynamic maintainability properties. The results obtained by both sets of metrics are pairwise compared to check whether the more modularity anomalies the system presents, the less stable and more difficult to maintain it is.