Correction to: Scientific Reportshttps://doi.org/10.1038/s41598-024-57582-1, published online 21 March 2024 The original version of this Article contained errors in the legend of Figure 4. “How does it feel? Significant effects predicted by Valence and Arousal models; n_participants = 142; n_observations = 1115. Panel (A) visualises participants’ clicks on the EmojiGrid, averaged by experimental condition (scenario) and touch phase. Results suggest that the characters of the observed interaction are perceived to be in a less negative affective state when touch is reciprocated by the receiver. That effect is moderated by the Partner factor, with the reciprocity effect being smaller in the Robot condition (B). A more aroused state is reported when touch happens between a human and the robot rather than two humans. More specifically, perceived arousal is higher when the robot is the one expressing vulnerability (C). Arousal also increases when touch is reciprocated by the receiver, with no significant difference between human and robot partner (D). Individuals’ propensity to trust others is positively associated with perceived arousal, only in human-to-human interactions (E). Individuals’ touch aversion is associated with reduced arousal.” now reads: “How does it feel? Significant effects predicted by Valence and Arousal models; n_participants = 142; n_observations = 1115. Panel (A) visualises participants’ clicks on the EmojiGrid, averaged by experimental condition (scenario) and touch phase. Results suggest that the characters of the observed interaction are perceived to be in a less negative affective state when touch is reciprocated by the receiver. That effect is moderated by the Partner factor, with the reciprocity effect being smaller in the Robot condition (B). A more neutral arousal state (closer to 0, which represents the center of the EmojiGrid) is reported when touch happens between a human and the robot rather than two humans. This is especially evident when the robot is the one expressing vulnerability (C). Arousal also becomes more neutral when touch is reciprocated by the receiver, with no significant difference between human and robot partner (D). Individuals’ propensity to trust others is associated with more neutral perceived arousal, only in human-to-human interactions (E). Individuals’ touch aversion is associated with more neutral arousal (F).” The original Article has been corrected.