Title: Control Styles in Robotic Scrum Facilitation: Insights from a Video Vignette-Based Experimental Study
Begin: Tuesday, 16. of November 2021 (4:30 PM)
Room: I 1.407
Given the trajectory of artificial intelligence (AI), autonomous conversational agents such as social robots will likely soon join teams in the workplace. A potential role such agents will play is adopting the role of facilitators, for example, facilitating certain tasks like a Scrum master in agile software development projects. In this role, they need to enact a certain degree of control over the Scrum team. It is unclear how human team members would react to a robotic Scrum master’s control style, which could be more enabling or more authoritative. Based on a vignette-based experiment supported with rich video animations, our findings suggest that a robot’s enabling control style results in higher team members’ job satisfaction because the robot’s enabling style leads to more trust. Thus, trust positively mediates the relationship between control style and job satisfaction. No mediation effect was found for monitoring. These findings have implications for control theory and conversational agents design. In this talk, I will also complement our findings with first insights we gained from a laboratory experiment, where subjects interact with a NAO robot as a Scrum Master in a role-playing game.