An Online Survey on the Perception of Mediated Social Touch Interaction and Device Design

Tracking changes in touch desire and touch avoidance before and after the COVID-19 outbreak

Touch is essential for survival, social bonding, and overall health. However, the COVID-19 pandemic calls for an abrupt withdrawal from physical contact, and the prolonged lockdown has left many people in solitude without touch for months. This unprecedented dissociation from touch has cast a shadow on people's mental and physical well-being. Here we approached the issue by examining COVID-19's impact on people's touch attitudes. We analyzed people's desire and avoidance for animate and inanimate targets based on large-scale Japanese Twitter posts over an 8-year span. We analyzed the impact of the COVID-19 outbreak with the difference-in-differences estimation method, which can estimate the impact while accounting for other changes over time such as seasonality or long-term effects. As a result, we found that people's desire for touching the human body and pet animals increased significantly after the COVID-19 outbreak and remained high afterward. In contrast, the avoidance of touching everyday objects (e.g., doorknobs and money) increased immediately after the outbreak but gradually returned to the pre-COVID-19 levels. Our findings manifest the impact of COVID-19 on human touch behavior. Most importantly, they highlight the sign of "skin hunger," a public health crisis due to social distancing, and call attention to the trend that people are becoming less aware of infection control as COVID-19 persists.

Linking Haptic Parameters to the Emotional Space for Mediated Social Touch

Social touch is essential for creating and maintaining strong interpersonal bonds amongst humans. However, when distance separates users, they often rely on voice and video communication technologies to stay connected with each other, and the lack of tactile interactions between users lowers the quality of the social interactions. In this research, we investigated haptic patterns to communicate five tactile messages comprising of four types of social touch (high five, handshake, caress, and asking for attention) and one physiological signal (the pulse of a heartbeat), delivered on the hand through a haptic glove. Since social interactions are highly dependent on their context, we conceived two interaction scenarios for each of the five tactile messages, conveying distinct emotions being spread across the circumplex model of emotions. We conducted two user studies: in the first one participants tuned the parameters of haptic patterns to convey tactile messages in each scenario, and a follow up study tested naïve participants to assess the validity of these patterns. Our results show that all haptic patterns were recognized above chance level, and the well-defined parameter clusters had a higher recognition rate, reinforcing the hypothesis that some social touches have more universal patterns than others. We also observed parallels between the parameters' levels and the type of emotions they conveyed based on their mapping in the circumplex model of emotions.