Twenty-five new subjects, with normal vision or corrected to normal vision with contact lenses, participated in this experiment (M = 20.11 years, SD = 1.28, 5 male). Participants received a monetary compensation for participation. All participants gave written informed consent for participation and this experiment was performed in accordance with the declaration of Helsinki. This study was approved by the local ethical committee of the Faculty of Social Sciences of Utrecht University.
Procedure and stimuli
In addition to the pupil-size measurements, skin conductance response (SCR) was measured during each trial. Both the dorsal (D) and ventral (V) sides of the right hand were stroked. The following sequence was repeated 15 times (90 trials in total): 3 cms−1 V, 30 cms−1 D, 0.3 cms−1 D, 30 cms−1 V, 3 cms−1 D, and 0.3 cms−1 V.
Analysis of subjective pleasantness
The analysis of the subjective pleasantness of touch was similar to Experiment 1, where TPT items were categorized in representing either positive or negative affect. Repeated-measures ANOVAs were conducted for both categories, with stroking velocity and side (D or V) as within-subject factors. Post-hoc tests for pairwise comparisons were performed, with a Bonferroni correction (α = 0.05). In addition, three post-hoc paired samples t tests were conducted, in which we compared side for each velocity (i.e., 0.3 cms−1 D vs 0.3 cms−1 V, 3 cms−1 D vs 3 cms−1 V, etc.). We applied a Bonferroni correction, with α = 0.0167.
Pupillometry collection and analysis
We measured and analyzed pupil size in the same way as for Experiment 1, with the exception that our statistical analysis now contained an additional effect: hand side (ventral/ dorsal). Therefore, we conducted linear mixed-effects analyses (again for every 33 ms sample) using pupil size as dependent measure, stroking velocity, hand side, and the stroking velocity × hand side interaction as fixed effects, and by-participant random intercepts and slopes for all fixed effects. Hand side was dummy coded, such that − 1 corresponded to ventral and 1 to dorsal, so that the reference value (0) was the average of ventral and dorsal.
SCR collection and analysis
SCR recordings were made using the Biosemi ActiveTwo system (Biosemi, Amsterdam, The Netherlands) at 16 Hz sampling rate from two passive AG/AgCl electrodes attached to the palmar surface of the left middle and index finger. Saline conductor gel was used to improve signal-to-noise ratio. The ground reference point consisted of the active common mode sense (CMS) and passive driven right leg (DRL) electrode placed on the dorsal side of the left hand.
A low-pass filter of 3 Hz was applied offline to the raw SCR data to reduce interference. For each trial, − 3000 ms to + 15,000 ms response windows were selected time-locked to stroke onsets. Baseline correction was applied by subtracting the averaged SCR activity 3000 ms pre-stroke onset period from the post-stroke onset values. Data reduction was performed using Brain Vision Analyser 2 (Brainproducts, Munich, Germany). Post-stroke onset signals were entered into the same per-sample linear-mixed-effects model as we used for the pupil-size data, with the sole exception that SCR samples were taken every 62.5 ms.
Results and discussion
Repeated-measures ANOVAs revealed that stroking velocity had a significant effect on TPT-item scores representing positive affect, F(2, 48) = 47.325, p < 0.001, partial η2 = 0.664. Positive items were rated higher during 3 cms−1 stroking compared to both 0.3 and 30 cms−1 stroking, p < 0.001 for all comparisons, Bonferroni corrected. In addition, 0.3 cms−1 stroking was rated significantly more pleasant than 30 cms−1 stroking, p = 0.042. No significant effect was found for side, F(1, 24) = 2.465, p = 0.130, η2 = 0.093. An interaction effect of side × stroking velocity was found, F(2, 48) = 4.973, p = 0.011, η2 = 0.172. Post-hoc paired samples t tests revealed a significant effect of side for 3 cms−1 stroking, in which dorsal stroking was rated as more pleasant than ventral stroking; t(24) = 2.657, p = 0.014. For 0.3 and 30 cms−1 stroking, no effect of side was found (see Figs. 3 and 4).
For words representing negative affect, stroking velocity also had a significant effect on TPT-item scores, F(2, 48) = 18.266, p < 0.001, η2 = 0.432. Negative items were rated lower during 3 cms−1 stroking compared to both 0.3 cms−1 and 30 cms−1 stroking. In addition, an interaction effect of side × stroking was found, F(2, 48) = 3.372, p = 0.043, η2 = 0.123. Post-hoc paired samples t tests with Bonferroni correction, however, revealed no significant differences in side for each velocity. Again, no significant effect was found for side, F(1, 24) = 1.977, p = 0.173, η2 = 0.076.
As shown in Fig. 5, there was again a clear effect of stroking velocity, such that the pupil was larger for higher stroking velocities, and no evidence that the pupil dilated most strongly for the intermediate (affective touch) stroking velocity. There was also a small effect of hand side, such that the pupil was slightly larger when the ventral side of the hand was stroked. Importantly, there was no significant interaction between hand side and stroking velocity.
As shown in Fig. 6, the SCR results largely resemble the pupil-size results. There was a clear effect of stroking velocity, such that skin conductance was higher for higher stroking velocities, and no evidence that the pupil dilated most strongly for the intermediate (affective touch) stroking velocity. There was also an effect of hand side, such that skin conductance was higher when the ventral side of the hand was stroked. Finally, there was a stroking velocity × hand side interaction, such that the effect of stroking velocity was most pronounced when the ventral side of the hand was stroked.