How Individuals Assign Personality Traits to QTrobot Based on Its Design and Behavior : Study on Robot-Assisted Learning Tasks with University Students

Abstract

The rapid advancement of technology has led to the integration of robots into various domains of everyday life, including education, healthcare, manufacturing, and customer service. As a result, human-robot interaction (HRI) has become a rapidly developing field that combines psychology, design, and engineering. Recent studies emphasize the importance of robot’s personality in fostering meaningful and engaging interactions between humans and robots. For instance, a robot designed to be friendly, engaging, and inquisitive can help improve children's participation and interest in classroom activities. Personality, which plays a role in human-human interaction (HHI), similarly impacts HRI by enhancing user experience (UX) and user satisfaction. Therefore, understanding how people assign personality traits to robots is important for designing robots that deliver a positive UX.

This qualitative study explores how university students assign personality traits to QTrobot, a humanoid social robot, based on its design and behaviors, particularly in its role as a teaching assistant. The study addresses three research questions: (1) What personality traits do the university students expect from the QTrobot as a teaching assistant? (2) What design elements affect the university students’ perceptions of QTrobot’s personality traits? and (3) What are the emotional reactions of the university students when interacting with the QTrobot as a teaching assistant? Utilizing a human-centered design (HCD) approach informed by participatory design (PD), two workshops were conducted at RoboStudio, Tampere University, involving eight university students. During these workshops, participants interacted with QTrobot while it delivered a lesson related to “social media ethics.” University students’ insights were gathered using Mural canvas tasks, online questionnaires, and post-task interviews. The collected data were primarily analyzed using the affinity diagramming method, with thematic analysis first applied to interview data and open-ended questionnaire responses.

Findings from the study revealed that students expect QTrobot to exhibit social, supportive, and professional personality traits, highlighting characteristics such as friendliness, politeness, happiness, engagement, patience, cooperation, empathy, and curiosity. Students' perceptions of robot personality traits were strongly influenced by nonverbal cues (e.g., facial expressions, gestures, and robot movements), verbal communication style (e.g., language used and responses provided), and the robot's appearance (e.g., child-like). During the interaction, students felt positive emotions, including engagement and enjoyment, when QTrobot's verbal and nonverbal behaviors matched their expected traits. In contrast, negative emotions such as boredom, confusion, and feeling overwhelmed occurred when the robot's verbal communication did not align with its nonverbal cues.

This research provides practical design implications aimed at improving the UX by enhancing the perception of personality traits in robots. The recommendations emphasize the importance of enhancing verbal and nonverbal communication, improving robot appearance, shaping personality traits, and enhancing the learning experience. Additionally, the thesis emphasizes ethical considerations that need to be addressed when designing robots, including trust, privacy, the humanization of robots, and concerns about replacing human roles in sensitive domains such as education and healthcare. Overall, this thesis provides valuable insights to the design of educational social robots, aiming to foster effective learning environments, enhance UX, and facilitate meaningful human-robot interactions through identifying robot’s personality traits.