Understanding the relationships between the perceptions of burnout and instability in Software Engineering

TL;DR

This study investigates how software engineers' perceptions of environmental instability—task, team, and technological changes—relate to burnout. Using a positivist, cross-sectional survey of 411 Portuguese-speaking software engineers and CFA-validated measures for burnout and instability, it finds positive associations between exhaustion and cynicism with all instability dimensions, while efficacy shows weak or no negative associations. Exhaustion is the most prevalent burnout symptom, and task instability is the most perceived instability type. The work provides actionable insights for industry to curb burnout by reducing perceived instability and offers directions for academia to refine instability measurements and further explore its consequences on well-being and performance in software development.

Abstract

Changes are inherent in software development, often increasing developers' perception of instability. Understanding the relationship between human factors and Software Engineering processes is crucial to mitigating and preventing issues. One such factor is burnout, a recognized disease that impacts productivity, turnover, and, most importantly, developers' well-being. Investigating the link between instability and burnout can help organizations implement strategies to improve developers' work conditions and performance. This study aims to identify and describe the relationship between perceived instability and burnout among software developers. A cross-sectional survey was conducted with 411 respondents, using convenience sampling and self-selection. In addition to analyzing variable relationships, confirmatory factor analysis was applied. Key findings include: (1) A significant positive relationship between burnout (exhaustion and cynicism) and team, technological, and task instability; (2) A weak negative relationship between efficacy and technological/team instability, with no correlation to task instability; (3) Exhaustion was the most frequently reported burnout symptom, while task instability was the most perceived type of instability. These results are valuable for both industry and academia, providing insights to reduce burnout and instability among software engineers. Future research can further explore the impact of instability, offering new perspectives on monitoring and mitigating its effects in software development.