Jari K. Hietanen

Research has shown that dyadic eye contact elicits psychophysiological responses related to affect and attention. In daily life, we also frequently notice eye contact occurring between other people, even when we are not directly involved. This study investigated whether such vicarious eye contact elicits affect‐ and attention‐related psychophysiological responses in observers. In two experiments, we measured participants' skin conductance (affective arousal), facial muscle activity (affective valence/affiliation), and heart rate deceleration (attention orienting) responses while they viewed gaze interactions between two agents. In Experiment 1, participants observed eye contact between two humans, whereas in Experiment 2, participants saw a human and a humanoid robot NAO. We found novel evidence for a vicarious eye contact effect : viewing eye contact between two humans elicited greater affective arousal in observers compared to non‐eye contact conditions. Somewhat surprisingly, facial muscle responses indicated increased zygomaticus activity and decreased corrugator activity during vicarious eye contact relative to non‐eye contact conditions when participants viewed eye contact between a human and a humanoid robot, but not when viewing eye contact between two humans. In both experiments, vicarious eye contact failed to elicit a greater heart rate deceleration response, suggesting that it lacks the attention‐capturing characteristics of dyadic eye contact. We discuss these findings in the context of mind attribution in gaze interactions and argue that using artificial agents as social stimuli can offer valuable insights into the role of mind attribution in social cognition.

Research has shown that dyadic eye contact elicits psychophysiological responses related to affect and attention. In daily life, we also frequently notice eye contact occurring between other people, even when we are not directly involved. This study investigated whether such vicarious eye contact elicits affect- and attention-related psychophysiological responses in observers. In two experiments, we measured participants’ skin conductance (affective arousal), facial muscle activity (affective valence/affiliation), and heart rate deceleration (attention orienting) responses while they viewed gaze interactions between two agents. In Experiment 1, participants observed eye contact between two humans, whereas in Experiment 2, participants saw a human and a humanoid robot NAO. We found novel evidence for a vicarious eye contact effect: viewing eye contact between two humans elicited greater affective arousal in observers compared to non-eye contact conditions. Somewhat surprisingly, facial muscle responses were greater to vicarious eye contact relative to non-eye contact conditions when participants viewed eye contact between a human and a humanoid robot, but not when viewing eye contact between two humans. In both experiments, vicarious eye contact failed to elicit a greater heart rate deceleration response, suggesting that it lacks the attention-capturing characteristics of dyadic eye contact. We discuss these findings in the context of mind attribution in gaze interactions and argue that using artificial agents as social stimuli can offer valuable insights into the role of mind attribution in social cognition.

Eye gaze is a fundamental nonverbal cue in human social interactions which is used to signal interest. When eye gaze is not received, individuals feel the painful effects of ostracism (being excluded and ignored). This is between humans – what happens in the ever-increasing human-robot dyads? Previous research has demonstrated that a humanoid robot’s eye gaze toward the observer elicits psychophysiological responses related to attention orienting and affective arousal and valence. Our study (N = 183) investigated the psychological, more complex reactions of ostracism in response to a humanoid robot’s averted versus direct gaze. We also compared these reactions to the responses elicited by a human’s gaze. Our results showed that both a human’s and a humanoid robot’s averted, compared to direct, gaze increased participants’ feelings of ostracism and negative mood, as well as lowered basic needs satisfaction, perceived closeness of the relationship (i.e., relational evaluation), and implicit self?esteem (i.e., indirect assessment of self-worth). Moreover, the negative socioemotional reactions to a humanoid robot’s averted gaze seemed to be equally as powerful as those to a human’s averted gaze. Our results indicate that individuals respond to humanoid robots as if they are a human.

We investigated whether trait inference about others’ social desirability influences perceivers’ judgments of eye contact with them, measured by the cone of direct gaze (CoDG) width—the range within which slight gaze deviations are perceived as direct. We also explored how the effect is modulated by participants’ gender and gazers’ sex. In Experiment 1, participants read descriptions of a likeable female and an unlikeable male (or vice versa). Subsequently, they judged whether they were being looked at or not by these individuals displaying various gaze directions. In Experiment 2, participants inferred trait desirability and performed eye contact tasks sequentially for two faces of the same sex, either beginning with a likeable face followed by an unlikeable face or vice versa. In Experiment 3, participants first performed eye-contact judgment for an individual without trait manipulation. They then inferred trait desirability for a second same-sex individual whose traits were manipulated as likeable or unlikeable, followed by eye-contact judgment for that individual. Results showed that CoDG width was wider for female faces associated with socially desirable traits than for those associated with socially undesirable traits. Desirable traits inferred from women’s social behaviors may align with individuals’ expectations of stereotypical women, enhancing their perceived likeability. The widened CoDG may result from individuals’ approach-related action tendency, self-referential positivity bias, or a combination of both mechanisms.

The present study (N = 367) investigated the association of the cone of direct gaze (CoDG; a range within which slightly averted gaze is accepted as direct gaze) width with the need to belong (NTB) and the possible further influence of NTB on self-esteem and loneliness. Results showed that: (i) men’s CoDG width predicted the NTB, self-esteem, and loneliness; (ii) men’s narrow CoDG predicted high loneliness via a high NTB and low self-esteem; (iii) a narrow CoDG was associated with low self-esteem via high loneliness only in high-NTB men. Among women, the CoDG width was not associated with the NTB, self-esteem, and loneliness. The findings are discussed considering the formation and maintenance of relationships together with gender differences.

1. Gaze cueing effect (GCE) refers to attention orienting towards the gazed-at location, characterised by faster responses to gazed-at than non-gazed at stimuli (Friesen & Kingstone, 1998). Despite...

We measured participants’ psychophysiological responses and gaze behavior while viewing a stimulus person's direct and averted gaze in three different conditions manipulating the participants’ experience of being watched. The results showed that skin conductance responses and heart rate deceleration responses were greater to direct than averted gaze only in the condition in which the participants had the experience of being watched by the other individual. In contrast, gaze direction had no effects on these responses when the participants were manipulated to believe that the other individual could not watch them or when the stimulus person was presented in a pre-recorded video. Importantly, the eye tracking measures showed no differences in participants’ looking behavior between these stimulus presentation conditions. The results of facial electromyography responses suggested that direct gaze elicited greater zygomatic and periocular responses than averted gaze did, independent of the presentation condition. It was concluded that the affective arousal and attention-orienting indexing autonomic responses to eye contact are driven by the experience of being watched. In contrast, the facial responses seem to reflect automatized affiliative responses which can be elicited even in conditions in which seeing another's direct gaze does not signal that the self is being watched.