Social robots are on the rise in various service environments, prompting questions about their readiness for integration into human settings. Effective verbal communication is crucial for user acceptance of social robots. This thesis focuses on testing a humanoid robot's verbal communication abilities, specifically in senior care environments. The goal is to identify key aspects of verbal communication in human-robot interactions and develop methods to assess these aspects.

A literature review explored key communication aspects, covering human and robotic contexts, resulting in four criteria: Content, Interaction Flow, Emotion, and Voice. Drawing measurement items from the literature, the test procedure, inspired by previous work, featured three role-play scenarios adapted to care contexts. Participants (N=11) received situational briefings for free interaction, and they quantitatively assessed communication with the humanoid robot using the developed criteria and measurement items. The evaluation was complemented by a qualitative assessment of the test method.

The findings show that the proposed testing method was positively accepted. The participants claimed the test setting was enjoyable and easy to understand, and the suggested testing criteria were effective at assessing the robot’s verbal communication. The average rating of Content, Interaction Flow, Emotion, and Voice showed a positive association with the overall rating of the robot's verbal interaction, emphasizing the method's holistic effectiveness. Moreover, the testing methode allows to see exactly what parts of the communication are good and what parts need improvement. When it came to the robot's verbal interaction, the results show that participants had varied opinions. While acknowledging the robot's communicative abilities, they expressed a need for a more creative, inspiring, and flexible robot. This thesis effectively creates assessment standards that identify verbal communication development opportunities and serves as the initial study for upcoming tests on verbal interaction of humanoid social robots.