German Title: Scrum und Selbstkontrolle

Preview

PDF, English - main document Download (2MB) | Terms of use

Abstract

Scrum is an agile software development methodology that suggests highly structured and iterative processes of software development by small autonomous teams (Schwaber & Sutherland, 2011). It is hypothesized that Scrum can be implemented in a way that supports and potentially improves Development Team members’ self-control. High self-control yields positive effects in countless life domains (Tangney, Baumeister, & Boone, 2004). One correlative study with 23 Development Teams and a total of 171 team members from an international software company investigated relevant conditions of Scrum to support and improve the self-control of Development Team members. Findings indicate a potential self-control improvement by the creation of a moderate concrete plan, the Sprint Backlog, at the beginning of a development cycle (Sprint), and performing an active team internal progress monitoring during the Sprint in short daily meetings (Daily Scrum Meetings). Planning and monitoring correlated simultaneously with high trait self-control and with low state self-control, indicating a possible self- control improvement through Scrum. Frequent team internal discussions targeting process improvements in Sprint Retrospective Meetings were related to high state self-control. In addition to correlations found with team members’ self-control, high team performance correlated with finishing committed Sprint Backlog Items by the end of the Sprints and reviewing these in the Sprint Review Meetings. Development Team members’ good health and low stress correlated with high team members’ self- control. High team autonomy was an essential requirement of the potential self-control improvement process. Moreover, high team autonomy correlated positively with high team performance and low experienced stress of Development Team members. Adding to these results, one laboratory experiment found indications that Sprint Backlog Item specificity may affect self-control during the processing of that item, suggesting a preference for moderate as opposed to low or highly specific Sprint Backlog Items. A second laboratory experiment failed to reproduce and extend this finding, probably because of the plan specificity operationalization. Overall, theoretical considerations and empirical indications are presented that Scrum could be implemented in a way that supports and potentially improves Development Team members’ self-control and health, reduces Development Team members’ experienced stress, and improves performance of the Development Team.