Pupil size
We found a main effect of Hand type [F(1,119) = 10.196, p = 0.002, η2 = 0.079], indicating a stronger pupil dilation when participants received a touch from a Human hand compared to an Artificial hand [t(119) = 3.193,p = 0.002]. Crucially, we also found a significant Hand type by Touch type interaction [F(1,119) = 7.402, p = 0.007, η2 = 0.059], indicating that the magnitude of increase in pupil dilation during the touch promoted by a Human hand differed depending on the type of touch. Specifically, post-hoc t-tests showed that only during Dynamic touch participants exhibited a stronger pupil dilation when receiving a touch from a Human hand compared to an Artificial hand [t(119) =4.023; p < 0.001], indicating that pupil dilation specifically encodes skin-stroking caress only when promoted by a Human hand. Furthermore, we observed that a touch promoted by a Human hand elicited a significant increase in pupil dilation for Dynamic compared to Static touch [t(119) = 2.966; p = 0.006]. During Static touch participants did not show any difference between a Human and Artificial hand [t(119) =0.213; p = 0.832]. More importantly, we did not observe any difference in pupil dilation between Dynamic and Static conditions when the touch was promoted by an Artificial hand [t(119) = 1.079; p = 0.283] (Figure 2b and 2c ). Supralinearity analyses showed that 70% (n=21) of participants displayed a supralinear effect, that is a larger pupil size in the Dynamic_Human condition alone than in the Dynamic_Artificial plus Static_Human conditions summed together [t(29) = 1.781, p = 0.043] (Figure 2d ). Our results show a stronger pupil dilation when participants received a dynamic touch delivered by a human hand. This kind of touch invoked a supralinear enhancement of pupil dilation indicating that the combination of these two features induced a significantly stronger physiological activation than the summed effects of each delivered separately.
Lastly, we did not find any differences in blink rates across conditions (all ps > 0.270), suggesting that participants did not show differences in blinking activity depending on the Hand type nor on Touch type.