FIGURE 3. (a) The GLM [successful trials > failed trials] contrast in the second stage of feedback reveals activated clusters, such as the rPG. (b) The right Sub-gyral was found correlated with the individual reciprocity index in the [successful trials > failed trials] contrast. The other two indices (greed and total reward) were not shown correlated with any brain regions. (c) The PPI results showed more negative functional connectivity in the MNS (such as the rPG, the left IPL, and the bilateral IFG,) with the rTPJseed in the successful reciprocity trials. (d) The bar graph shows the beta values extracted from some of the regions shown in (d) for illustration purposes only. These regions are negatively correlated with the seed region, the rTPJ [from left to the right: rTPJseed (M = 0.13, SD = 1.07), rPG (M = -3.29, SD = 0.51), left IPL (M = -2.98,SD = 0.38), and left IFG -3 (M = -3.15, SD = 0.43)]. (e) The rTPJseed-rPG, -rAG, and -rIPL connectivities, correlated with individual total reward indicates that one’s rTPJ (thinking of others) activity increases, with the depressed activity of the right PG for coordinated action, and the right IPL and right AG for relating themselves and detecting incongruence, in the successful reciprocity trials. See further detailed interpretation in Discussion. (f) The scatter plot shows the correlations between the total reward and one of the three ROIs (the right IPL chosen for illustration purposes; See the other two in Supplementary Figure S3) found via PPI. The three types of participants are: reciprocal (red), dominant (green), and submissive (blue). The bars on the x axle show the mean total reward of the three types: reciprocal (NT$ 307.82), dominant (NT$ 274), and submissive (NT$ 247.28). The scatter plot suggests the reciprocal individuals (in red) with the most reward (reciprocal-dominant, p = 0.014; reciprocal-submissive, p< 0.001) and negative connectivity between the rTPJseed and the right IPL identified from PPI.
rTPJseed-to-whole brain frequency-domain coherence
Comparisons among various pairing combinations
The rTPJseed-whole brain interpersonal coherence of the WIPs in the second stage feedback time is shown in the Supplementary Table S6. Brain regions show significant coherence with the rTPJseed during the event of feedback time under the cluster k-threshold of 40 voxels and the applied map threshold of 0.215. Furthermore, we compared the 36th frequency bin with the rest, excluding the first 3 bins, and the 35th/the 37th; the peak of 36th bin is higher than the rest by 1.5 of the standard errors. The ROIs with significant coherence with the rTPJseed are the rTPJ and among others. In the WNPs, the rTPJseed are significantly coupled with the rSTG and others under the cluster k-threshold of 40 voxels and the applied map threshold of 0.215 (see Supplementary Table S7). Also the peak of the 36th frequency bin is higher than the rest by 1.5 standard errors.
To further investigate inter-regional couplings, Figure 4 shows the coherence spectra between rTPJseed-rTPJp(posterior right TPJ) in the WIPs pairs (Figure 4a) and rTPJseed-rSTG in the WNPs (Figure 4b). The independent two-sample t-tests are applied among every possible pair at the 36th frequency bin. The rTPJseed-rTPJpcoherence value of the WIPs is stronger than those of the WNPs and BPPs (all ps < 0.001). The rTPJseed-rSTG coherence value of the WNPs (in yellow) is stronger than those of the WIPs (p < 0.016) and BPPs (p < 0.001).
We ran the same analyses and applied the same peak threshold (the 36th frequency bin is higher than the rest by 1.5 standard errors) on the first feedback (in the planning stage) to compare with the results of the second feedback stage in coherence. Brain regions show significant coherence with the rTPJseed during the event of first feedback time under the cluster k-threshold of 40 voxels and the applied map threshold of 0.225. Bilateral Lingual Gyri, the left PG, and the right Precuneus are identified in the WIPs. In the WNPs, the bilateral Lingual Gyri were found coupling with the rTPJseedduring the event of first feedback time under the cluster k-threshold of 40 voxels and the applied map threshold of 0.22 (see Figure 4c and Supplementary Table S8-9).
As the interacting pairs had real interactions with each other in the same group, we calculated the three neural differences in coherence when 1) reciprocal people interacted with reciprocal partners and 2) with non-reciprocal ones (the dominant and submissive types). Furthermore, we calculated the coherence values among 3) the non-reciprocal pairs. After applying the stringent threshold (the peak of the designated frequency bin, the 36th, should be higher than the rest by 1.5 standard errors), we found the left TPJ and right IFG coupled with the rTPJ among the reciprocal pairs (under the cluster k-threshold of 40 voxels and the applied map threshold of 0.228). When the reciprocal individuals interacted with the non-reciprocal partners, no coupling was found with the rTPJ. As for the non-reciprocal pairs, their left IFG and the right amygdala were in sync with the rTPJ (under the cluster k-threshold of 40 voxels and the applied map threshold of 0.226). (See Figure 4d-e and the Supplementary Table S10-11)
To examine the brain-behavior relationship, we calculated the correlation between the coherence values of the regions found coupled with the rTPJ in the WIP. No significant results were shown (see the Supplementary Table S12).