Paradigm:
Our novel paradigm combines VR, motion tracking and fNIRS recordings from the bilateral inferior parietal cortex. Participants were instructed to perform two tasks with a participant in another room (whose behaviour was actually generated by the VR system). The first task was to describe a picture or listen to their partner’s description; this task sets a conversation context. The second task was to make 5 simple hand movements in a 30 second block and watch their partner’s hand movements. Different virtual partners were programmed to mimic the participant in various ways (see Experiments 1 to 3 below). Polhemus motion tracking markers tracked the movements of the participants’ hands, allowing a monitoring of their motor performance and a systematic implementation of mimicry. After interacting with each partner via this VR system, participants rated trust and liking and then repeated the procedure with a new virtual partner. They also engaged in variations of a Virtual Maze test to measure implicit trust. In Experiment 1, two independent variables were manipulated with two levels each: (1) congruency and (2) temporal contiguity of mimicked actions. In Experiments 2 and 3 these factors are manipulated independently. The experimental setup in Experiments 2 and 3 were similar to Experiment 1 with a few major differences: Here, (1) participants were instructed to initiate their movements themselves, hence no auditory cues were used to prompt their movements and a more complex mimicry algorithm using proximity sensors of the Polhemus motion trackers was employed. We also added the fNIRS measurement to investigate the neural correlates of being mimicked: In Experiment 2 we used an Artinis Medical systems Oxymon Mk III fNIRS system with 4 channels over left inferior parietal cortex and 4 over right inferior parietal cortex, in Experiment 3, we used a Shimadzu Lightnirs with optodes arranged in two by four arrays, resulting in 20 channels and larger coverage of the IPL and pSTS bilaterally. In both studies, oxy-Hb signals were analysed with a general linear model approach in SPM.
Results:
In Experiment 1, the final sample was 23 (12F:11M, mean age 26.17 yrs). Results show that there was a significant main effect of congruency of the averaged social judgement ratings. Partners moving congruently tend to be judged more favourably than those who move incongruently. We found a significant disordinal interaction between congruency and contiguity for the likeability ratings. This suggests that partners who move congruently are judged more favourably than those who move incongruently in the slow condition, but the opposite is true for the fast condition. In the Virtual Maze task, which measured implicit trust towards an interaction partner. However, we found a trend for a main effect of contiguity. Participants ask more often for advice those partners who moved slow, rather than fast, irrespective of congruency. This suggests that for an implicit measure of trust towards others, congruency is irrelevant and the delay of movement is again important to elicit more trust (irrespective of congruency of movement).
In Experiment 2 data were collected from 39 participants (24F:15M, mean age 26.20 yrs). There were 19 datasest for fNIRS analysis. Behavioural results show a main effect of manipulation with the partner mimicking at a fast timing being judged as significantly less trustworthy compared to the one mimicking slow or the two control conditions. No such effect was found for rapport judgements or for the implicit trust measured by the virtual maze. fNIRS results show stronger activation in the IPL to the fast-mimicry partner. This is consistent with the role of IPL in differentiating self from other. This suggests that contiguity is important for understanding mimicry responses.
Data for Experiment 3 is still being collected/analysed.
Results have been presented at various national and international scientific conferences and meetings. Publications are in preparation.
