Investigating the effect of heartrate synchrony on prosocial behaviour towards a robot

Abstract

Robots are being designed to alleviate the burden of social isolation and loneliness, particularly for older adults for whom these issues are more widespread. While good intentions underpin these developments, the reality is that many of these robots are abandoned within a short period of time. To encourage the longer-term use and utility of such robots, researchers are exploring ways to increase robot likability and facilitate attachment. Results from experimental psychology suggest that interpersonal synchrony (the overlap of movement/sensation between two agents) increases the extent to which people like one another. To investigate the possibility that synchrony could facilitate people’s liking towards a robot, we undertook a between-subjects experiment in which participants interacted with a robot programmed to illuminate at the same rate, or 20% slower, than their heart rate. To quantify the impact of cardio-visual synchrony on prosocial attitudes and behaviours toward this robot, participants completed self-report questionnaires, a gaze-cueing task, and were asked to strike the robot with a mallet. Contrary to hypotheses, results revealed no differences in self-reported liking of the robot, gaze cueing effects, or the extent to which participants hesitated to hit the robot between the synchronous and asynchronous groups. The qualitative data collected in semi-structured interviews provided rich insights, though, and call into question the use of the broad “Likability” measurement, and the appropriateness of the ‘hesitance to hit’ paradigm as a measure of attachment to a robotic system.

“Hit the Robot on the Head With This Mallet” – Making a Case for Including More Open Questions in HRI Research

“…To further investigate whether these undeniably prosocial feelings persist, or change over time, or whether it is possible for those reporting no emotional attachment toward the robot to develop attachment, longitudinal data collection would be required (c.f. Riddoch and Cross, 2020). To better understand individual differences in attachment to robots, it would also be insightful to collect more information from participants regarding their personality and background.…”

“…It could be suggested that if a robot were deployed in a certain setting, and programmed to behave in a consistent way, that there could be wide scale acceptance and usage, but without an appreciation for individual differences and changes in attitudes over time, we argue that this is unlikely to be the case. Despite being a complicated venture, a better understanding how humans perceive robots, and how these feelings evolve across time (e.g., Riddoch and Cross, 2020) is becoming increasingly important, as in some instances, it could be inappropriate/ harmful to for humans to feel attached to robots -e.g., military personnel with bomb-disposal robots, or vulnerable people with care robots (Sharkey and Sharkey, 2012;Darling, 2015;Lin et al, 2017).…”