Aged to Perfection? Analyzing the Impact of Years of Experience on Code Quality

The association between developer experience and code quality is a significant debate within the software engineering community. To explore this relationship, we quantitatively evaluated 401 GitHub software repositories in JavaScript, PHP, and Python, maintained by 98 developers with Workana profiles – a freelancing platform. For quality assessment, we rely on SonarQube, a widely adopted tool in both industry and academia. We analyzed several dimensions, such as the programming languages and severity (high, medium, and low) of maintainability and reliability issues. We observed that developers with the highest level of experience presented code with fewer issues. However, developers with intermediate level of experience presented more code quality issues than their novice and experienced counterparts, revealing a more complex and nuanced relationship between level of experience and code quality. Despite that, our analysis did not indicate significant statistical differences between the density of the issue at varying levels of experience, suggesting that other factors may also contribute to code quality outcomes. This study contributes to an open question with implications for software engineering recruitment and code quality assurance.We provide a new dataset of developers and their projects based on recent data. All extracted data and accompanying scripts are available, aiming to enable further replications of our study.