Building Better Research Software (Are you sure?): Why Technical Refactoring Does Not Lead To Software Quality

Paper i proceeding, 2026

Software is fundamental to modern research. This paper reports the analysis of source code produced from a Software Carpentry tutorial on Building Better Research Software to understand whether the recommendations and practices advocated led to improvements in the quality of the source code. The analysis was undertaken as part of an educational exercise by fourth year Immersive Software Engineering (ISE) post-graduate students. Their observations suggest (i) the adoption of standard practices in software development such as the use of meaningful variable and function names, removing dead code, the use third-party libraries, separating units of functionality, and using comments and docstrings can lead to code that is easier to read and understand; but also (ii) examination of the code through the lens of core software engineering design principles, i.e. SOLID, highlighted significant violations, which have practical and measurable effects on the quality of the code in relation to key characteristics of maintainability including modularity, modifiability, reusability, and testability. The implications of violating SOLID principles directly undermines maintainability as a core quality that contributes to the overall technical sustainability of all codebases.