Notes
Robot-Supported Decision-Making Processes: Effects of Embodiment, Recommendation Timing and Gender
307With the rise of elaborated AI-systems the application of these technologies in decision-making processes broaden and robot-supported decisions will be a crucial part of future workplaces. However, with the use of robotic advice problems like overtrust (Ullrich, Butz & Diefenbach, 2021) or algorithmic aversion (Mahmud et al., 2022) might occur, which inhibit the advantages of robot-supported decisions. Therefore, it is necessary to investigate important characteristics that affect human decision making. Prior research demonstrated that the timing of robotic advice effects how humans make use of this advice (Straßmann et al., 2020). As the embodiment of robotic systems has an important role on interaction variables, we assume that an embodied robot with a physical body has a stronger effect on in the decision-making process than a voice assistant without one. Moreover, it was found that men and women perceive and trust robotic systems differently (Gallimore et al., 2019). Accordingly, this work investigates the influence of the users’ gender in robot-supported decision-making. In an online experiment with a 2x2 between-subject design (N = 288 participants (143 women, 142 men and 3 non-binary)), two video-recorded robot systems presented a decision task and their recommendations for the decision. The decision task and associated recommendation was either presented by a physically embodied robot (Pepper) or an embodied speech-assistant (Alexa). Moreover, the time at which the recommendation was presented varied: The robotic system either presented its recommendation together with the decision task or first presented the task, then gave the user the chance to state their own opinion and then the robotic recommendation was presented. In a questionnaire, participants rated the robotic system, the decision-making process, and their decision. The study results give valuable insights on psychological effects in robot-supported decision-making and thereby deliver design recommendations for the use of robotic support systems.
308Robot-supported decision-making / gender effects / decision process / robot / speech assistant