The paper will present a model for the verbal interaction between humans and robots, the "CAT Technology Equivalence Model". The model is based on the CAT-model bei Giles and others, which war focused on human-to-human communication. To test the model we conducted an experimental study integrating the ChatGPTo chatbot into a NAO robot. NAO is a small, humanoid robot equipped with a range of physical and linguistic capabilities. By integrating ChatGPTo, we enabled NAO to generate verbal responses, thereby facilitating direct verbal interactions between human participants and the robot. The experimental setup took place in a laboratory environment ("Human-Robot Lab") at our university. The results of 9 conversations between the robot and human participants will be contexualized in our model and we can demonstrate, on which verbal levels both interlocutors adapted to each other.

Human-Autonomy teamwork (HAT) model approach has been adapted to a variety of contexts and challenges of human-machine collaboration, but has so far widely neglected the dimensions of (verbal and non-verbal) communication between humans and (their) machines. In order to colect data on communicaotin practice in the field of archietcture, we tested the HAT-model through HRI-enabled fabrication processes with sand and clay. In these processes, a human teammate physically interacts with the robot teammate to dynamically adapt robotic actions to the emergent material states. Results show that certain co-creation modes which reduce human insecurities could lead to more precision in human-robot teamwork.

This study extends the Computers-Are-Social-Actors (CASA) framework and Proteus effect theories by examining how people are influenced when a media agent becomes their avatar. The CASA framework suggests that people respond socially to media agents, while the Proteus effect posits that users conform to avatar characteristics. This research explores the intersection of these theories, investigating whether embodying a previously encountered media agent enhances its persuasive influence. Across two experiments using virtual reality, participants first interacted with an agent endorsing an attitude and then either embodied that agent or a non-endorsing one. Results from the Pilot Study (N = 170) suggest that participants who controlled the attitude-endorsing agent reported greater attitudinal agreement and behavioral intention. The Main Study (N = 442) examines team membership, agent embodiment, and agent control. Main Study analyses are in progress. Findings will advance theoretical understanding of media influence and offer practical implications for persuasive design in digital environments.