We present results of a quasi-experimental study investigating how user preference might change after direct interaction with two different types of robot regarding morphology and affordances: a machine-like that interacts through touch, and a human-like that interacts verbally. The study was performed in an art exhibition setting, where visitors had the opportunity to interact with the robots voluntarily, and were asked to fill out questionnaires before and after the experience. Post interactions, visitors preferred to touch the machine-like “hard” robot despite initial stated preference for soft materials, preferred mutual contact despite initial preference of subject to initiate touch, and preferred communication with a robot that could touch rather than initial preference for a robot that could “see”. Overall, users showed a significant constant preference for the machine-like robot, reportedly feeling a stronger connection with it than with the human-like one as it met their expectations, and they found its movements more appealing. Social conditioning can render people reluctant to touch a robot with very human-like appearance, and set the expectations for interaction too high to meet. Our results, lastly, indicate that interaction with more than one type of social robot can affect the interaction experience for each of the robots.

The human perception of cognitive robots as social depends on many factors, including those that do not necessarily pertain to a robot’s cognitive functioning. Experience Design offers a useful framework for evaluating when participants interact with robots as products or tools and when they regard them as social actors. This study describes a between-participants experiment conducted at a science museum, where visitors were invited to play a game of noughts and crosses with a Baxter robot. The goal is to foster meaningful interactions that promote engagement between the human and robot in a museum context. Using an Experience Design framework, we tested the robot in three different conditions to better understand which factors contribute to the perception of robots as social. The experiment also outlines best practices for conducting human-robot interaction research in museum exhibitions. Results from the study indicate that perceived social presence can be evaluated using a combination of HRI and Experience Design methods that measure co-presence and co-experience.