HomeEventsPhD Defence Christian Willemse

PhD Defence Christian Willemse

Social touch technologies - how they feel and how they make you feel

Christian Willemse is a PhD student in the research group Human Media Interaction. His supervisors are prof.dr. D.K.J. Heylen and prof.dr. J.B.F. van Erp from the Faculty of Electrical Engineering, Mathematics and Computer Science.

People tend to touch each other on a regular basis during their social encounters. When such a touch, like for example a pat on the back or a hug, leads to systematic changes in the recipient’s thoughts, feelings, and/or behavior, we speak of social touch. Social touches can come in many different forms and can elicit a vast range of responses. Touches can, among others, have strong effects on bonding and attachment between people. This is easy to imagine when you think about the intimate physical contact between lovers or a gentle caress from mother to child. Moreover, touch can have a severe impact on one’s emotional state and physiological well-being. Receiving touch can for example help in the direct decrease of stress symptoms. The third effect of touch that is of particular importance for this dissertation, is the effect of touch on the recipient’s behavior and attitude. Receiving a touch can for example enhance one’s helpful behavior, one’s generosity, and one’s willingness to comply with a request. This is also known as the Midas Touch effect. Aforementioned effects of touch do not occur with all forms of physical contact and in any situation per se. There are various aspects that may modulate the responses, which include the physical composition of the touch (i.e., how the touch feels), the setting in which the touch takes place, the relationship between toucher and touchee, and several personality characteristics of the recipient of the touch, such as his or her attachment style and touch receptivity.

Despite the importance of social touch in co-located interpersonal communication, this form of communication is underexposed in current communication technologies. In this dissertation, we provide a next step in the research on Social Touch Technologies; technologies in which human social touches are emulated through the use of various haptic actuators. We question whether the use of social touch technologies can elicit similar effects in the human recipient as actual human touches can, and if so, under which specific circumstances. Contrary to co-located human touch, social touch technologies enable the disentanglement of the various dimensions that may modulate the perception of a touch and to investigate each of these social touch building blocks in isolation. On this premise, we identified three broad main challenges that need to be addressed to advance the understanding of the use of social touch technologies. (1) We need to understand how a social touch technology should feel. What are essential physical qualities, and to what extent should it resemble an actual human touch? (2) We need to understand the importance of attributing the physical stimulus to a social source (i.e., the sender). Are the (anticipated) effects of a social touch technology mainly caused by the feel of the touch, or by the cognitive processes that are involved once the recipient knows who sent the touch? (3) We need to understand which contextual, relational, and personal elements may modulate the perception of a touch through a social touch technology, and how. In this dissertation, we addressed these challenges by researching two specific instances of social touch technologies. The research is intended to contribute to the understanding of the opportunities, limitations, and boundary conditions of interactions through social touch technologies, with the intention to eventually develop meaningful and effective interactions.

In Part II of this dissertation, we focused on mediated social touch; a sub-category of social touch technologies in which the recipient receives an artificial touch and attributes it to a remotely located other person. More specifically, we focused on the role that physical warmth can play in such interactions. On the premise that warmth is an important physical quality of a human touch, and on the premise that perceptions of physical warmth – either through warm objects, or via the ambient temperature – can activate perceptions of social warmth (a physical-social warmth link), we carried out three studies. In these studies, participants engaged in a form of communication with another person, in which they either received warmth or a stimulus at room temperature. The warmth was either presented through a heated chair (Study 1), or through a custom built mediated touch interface that dissipated warmth (Studies 2 and 3). When receiving a warm physical stimulus, this was either attributed to the sender (as (mediated) body heat), or to a non-social feature of the haptic interface (in order to better understand the role of attribution). Over the course of these studies, we also investigated the role of various possible modulators of the perceptions of warm mediated touches, such as the ambient temperature, the moment of presentation of the warmth during the communication, the duration of the exposure to the warmth, the relation between the sender and recipient, and the setting in which the warmth was perceived. Neither the outcomes of the studies, nor accompanying meta-analyses provided support for the suggestion that the physical-social warmth link is also of practical use in mediated social touch interactions. As a consequence, we were not able to find insights in the role that attribution of the warmth or other possible modulators may play during such communication either.

In Part III of this dissertation, we focus on another sub-category of social touch technologies, namely simulated social touch. In this category, the touch that is perceived through a haptic interface is attributed to a non-human social entity such as a virtual agent or a social robot. We specifically focused on robot-initiated touches, and questioned whether such touches could elicit similar effects as human touches. Moreover, we investigated some of the possible boundary conditions within which such robot-initiated touches could elicit beneficial effects with regard to the human recipient’s emotional state, his or her perceptions of the robot, and his or her pro-social behavior. In three studies, participants were invited to watch a scary movie (to elicit stress) together with a social robot that either did or did not provide supposedly soothing touches occasionally. Throughout these studies, we varied the social presence of the robot (i.e., the extent to which the robot is actually considered a social entity rather than a non-social machine) and the bond the recipient had with the robot prior to the touch interaction. Although the majority of the dependent variables that we applied were not affected by the touch manipulations, we did find some beneficial effects of robot-initiated touch on one’s heart rate during stressful circumstances, the perceived intimacy that was felt towards the robot, and the attachment one had with the robot. These effects imply that robot-initiated touches indeed can have beneficial effects, but only within specific boundary conditions. To be able to exactly define these conditions, more research is warranted.

Since people are very well capable of imagining how a touch would feel, solely based on observing a touch or by merely thinking about it, we carried out three additional studies in which visual stimuli related to robot-initiated touch were employed in online crowdsourcing studies. From these studies, several guidelines for the design of the physical appearance of robots were derived, in order to optimize them for robot-initiated social touch interactions. Robots should for example appear friendly and non-threatening, but only moderately humanlike and perhaps somewhat caricatured.

Although the nine studies as presented in this dissertation do not unequivocally support the idea that social touch technologies can elicit similar responses as human touches can, this work contributes to the research area in several ways. Firstly, the results provide clear insights in the limitations of social touch technologies. Physical stimulation by means of haptic actuators does not necessarily lead to beneficial effects. The most important contribution of this dissertation is however the contribution to the maturation of the research field. Based on an extensive literature discussion we developed a working model for social touch and we defined several challenges that need to be addressed to advance the understanding of social touch. Moreover, we developed and applied various research methodologies such as the lab studies with the scary movie, but also the online experiments with visual materials. These aspects, combined with the outcomes of the studies and our suggestions for improvement may form a solid basis for upcoming research and will hopefully inspire other researchers. Progressing from a young research area with some haphazard studies and mostly anecdotal evidence, to a mature field with clear, coherent, and consistent insights is necessary to be able to eventually develop the meaningful interactions that social touch technologies may have to offer.