Behavioral Software Engineering - Organizational Change Refocused

Doctoral thesis, 2018

Background The development of software is not only a technical endeavor; it is significantly affected by the behaviors of the people involved. Since social scientists have been studying humans for over a century, it is likely that insights they have developed could be used to increase software development effectiveness. There are, nevertheless, indications that software engineering researchers seldom use theories developed and proven within the social sciences. Overall, software engineering research that emphasizes human aspects is still limited compared to studies with technology or process focus.

Objective Given the importance of human aspects in software development, we argue that knowledge from the social sciences should be used more extensively to improve software engineering research. Therefore, the primary objective of our research was to advance software engineering by more profoundly considering humans and their behavior. For in-depth insights into such interdisciplinary research efforts, we chose to explore a specific phenomenon: organizational change. Our secondary objective was thus to create understanding to help improve software companies’ organizational change efforts.

Method To address our research objectives, we used a variety of research designs and data collection methods, including literature reviews, surveys, interviews, focus groups, and quantitative analysis of project data. This diversity allowed us to examine phenomena from different perspectives.

Results We provide directions for future research on behavioral aspects of software engineering by outlining the behavioral software engineering (BSE) research area, reviewing contemporary research, and identifying industrial needs. Moreover, our findings suggest that software engineers form their attitudes toward change collectively and according to their teams’ social norms, which are governed by their distinct professional identity. Our results also indicate that misalignment of organizational values between groups adversely affects change efforts and overall performance.

Conclusions Our research concludes that in order to effectively manage organizational change efforts, software companies must strengthen their organizational identity and reduce misalignment of organizational values. By providing such concrete advice on how to improve organizational change, our research confirms the usefulness of and need for additional BSE research to create novel and in-depth insights into software engineering phenomena.