Increasing developers’ code accountability perceptions in open source software development

Open source (OS) software projects rely on high-quality code, making developers’ participation in reviewing their code essential to improve its quality. In non-OS domains, formal consequences (i.e., sanctions and rewards like fines or bonuses) are usually applied to raise developers’ perceived accountability for their code, thus motivating them to review it. However, in the OS domain, such formal consequences are often absent or only weakly pronounced, raising questions of whether and how developers perceive accountability for their code and how it affects their motivation to improve its quality. Using a developmental mixed-method research approach, we conducted a qualitative study with semi-structured online interviews to develop a research model (n = 29) and validated it through a follow-up scenario-based factorial survey (n = 480). Our findings indicate that in the OS domain, developers often perceive themselves accountable for their code. Project- and individual-related factors, such as legal liability for contributed code, the integration of the OS software project into other projects, the work status of developers, and gaining a reputation for code contributions, appear as mechanisms to influence and increase developers’ perceived code accountability. Furthermore, we observe that perceived code accountability promotes developers’ intention to review their code before contributing it to OS software projects and, thus, contribute code with higher quality. Overall, our paper contributes to information systems (IS) research on accountability perceptions in OS software development, indicating that in the OS domain, formal consequences such as sanctions and rewards are not exclusively necessary for evoking developers’ perceived code accountability and increasing code quality. • Developers perceive code accountability within the OS domain. • Perceived code accountability motivates developers to review their own code. • Project- and individual-related factors shape perceived code accountability.