Chenyu Zhang, Sihua Xu
Value-based decisions often require the integration of information that is not directly observable but internally constructed from memory, simulation, and affective processes. Although previous research has extensively examined valuation for options with explicit attributes, less is known about how such internally generated information influences the dynamics of evidence accumulation and its neural implementation. In this study, we integrated functional near-infrared spectroscopy (fNIRS) with Hierarchical Drift Diffusion Modeling (HDDM) to examine decision processes in experiential versus material contexts, which differ in their reliance on internally constructed versus externally specified value information. Behaviorally, experiential choices were associated with higher willingness to upgrade and reduced sensitivity to price. Computationally, HDDM revealed that this preference was driven by a drift bias toward upgrading, while decision thresholds remained stable. At the neural level, fNIRS showed increased activation in a prefrontal network including the frontopolar cortex (FPC), inferior frontal gyrus (IFG), and dorsolateral prefrontal cortex (dlPFC) during experiential evaluation. Dynamic causal modeling (DCM) further indicated condition-dependent modulation of intrinsic connectivity within dorsolateral prefrontal cortex, characterized by reduced self-inhibitory coupling. Together, these findings suggest that differences in value construction are associated with systematic variation in the efficiency of evidence accumulation and with corresponding changes in prefrontal functional dynamics. The observed links between drift rate, prefrontal dynamics, and individual differences in subjective effort and evaluation strategies provide a coherent bridge between computational mechanisms and their neural substrates, offering a network-level account of how internally generated information is integrated during value-based decision-making.