Motives and risks of self-disclosure to robots versus humans
Augustine A, Eyssel F (Accepted)
ACM Transactions on Human-Robot Interaction.
The extent to which people self-disclose depends on the valence and intimacy of that information. We developed a 16-item measure that features both dimensions to assess participants’ inclination to self-disclose to humans and robots across three studies, i.e., the VISS scale. We hypothesized that participants would prefer a human over a robot conversation partner, independent of the valence and intimacy of the topic. Using the VISS, participants reported equal willingness to self-disclose to a human and a robot. However, in one out of three studies, participants were more willing to self-disclose positive topics that are low in intimacy (PTLI) to a fellow human, whereas in two out of three studies, they were less prone to self-disclose negative topics that are high in intimacy (NTHI) to a human compared to a robot. Two out of three studies provide a plausible explanation for these differences: participants indicated more motives to self-disclose PTLI to a human than to a robot, with relationship development being one central motive that was more strongly indicated in relation to a human. Furthermore, participants perceived more risks to self-disclose NTHI to a human than to a robot, with the risk of rejection being most prevalent for a human.