Trust facilitates conditions for safe sharing of valued resources---a social setting vital to success in a wide range of socio-technological networks. With an increasing reliance of economic initiatives on trust-assured interactions, the need to inquire into the mental processes of trust has emerged. This led to a proliferation of functional magnetic resonance imaging (fMRI) studies focusing on domain-specific measures. However, inherent metric deficits of fMRI have resulted in discrete outcomes highlighting further need for constructing a comprehensive neurocognitive model of trust. Here, a domain-general methodology aims at overcoming the negative tendencies in prior fMRI studies by applying a series of coordinate-based "Activation Likelihood Estimation" (ALE) meta-analyses of the fMRI data and a data-driven Multivariate Granger Causality (MVGC) connectivity analysis of hyperscan-fMRI data---an approach not undertaken in trust studies prior to this dissertation. To determine the effects on behavior of cross-study variability in brain activation during a trust-inducing investment game (IG) task, the meta-analysis aims at revealing the extent of neurocognitive differentiation during trust, learning to trust and reciprocity. One-shot IG, implicating unconditional trust, is compared to multi-round IG implicating the conditional trust. In the MVGC study, the neurocognitive differences in the effective connectivity of interpersonal ("brain-to-brain") trust are discerned. Meta-analysis revealed a strong differential response between unconditional trust (ambiguity, insula) and conditional trust (reward, ventral striatum). Learning to trust engaged a goal-guided (rostrolateral PFC) transition between decision-making (dorsal striatum, action-valuation) and feedback processing (ventral striatum, reward reinforcement). Reciprocating trust was linked to insula-mediated norm-compliance tendency to avoid breaking trust. For the effective connectivity analysis, a steady increase in trust and reciprocity engaged a mentalizing network as evidenced in the observed dorsal PFC connectivity with the parietal cortex. Within-trustor, dorsomedial (dmPFC) bidirectional connectivity with posterior cingulate cortex was key to guiding trust-valued choices (hypothalamus). Within-trustee, the key motive of norm-compliance---trustworthiness (lateral orbitofrontal cortex) was mediated by dorsomedial and dorsolateral PFC in Stage 1 and by precuneus in Stage 2. For the brain-to-brain exchange, the trustor's dmPFC was most active, but the trustee's dmPFC was virtually absent indicating dissociable patterns of other-regarding preferences for the trustor and trustee. Collectively, this dissertation lends evidence consistent with the putative socio-cognitive, economic-utility and reinforcement-learning models of trust and opens new perspectives by applying an effective domain-general data-driven dynamic approach.
展开▼
