Conflict resolution in the context of an audiovisual temporal order judgment task in rat and human

Abstract Conflict resolution is fundamental to adaptive decision-making, yet how species differ in resolving sensory and response conflicts remains poorly understood. Here, we compared conflict resolution in humans and rats using an analogous audiovisual temporal order judgment task, where the location of the first stimulus was either congruent or incongruent with the instructed response. Hierarchical Bayesian modeling was applied across three complementary analysis streams: distributional analyses of conditional accuracy and delta plots, psychometric functions, and drift–diffusion models. Distributional analyses revealed no congruency effect on accuracy in humans, but a facilitative effect on reaction time (faster responses on incongruent trials) that was stable across reaction-time bins. In contrast, rats exhibited a positive congruency effect on both accuracy and reaction time, with the accuracy difference decreasing across reaction-time bins in auditory-first trials but increasing in visual-first trials. Psychometric modeling showed that incongruent trials did not reliably alter temporal sensitivity in humans, though a modest increase in the lower lapse rate was observed. In rats, however, conflict substantially impaired temporal sensitivity, accompanied by dramatic increases in both lower and upper lapse rates—particularly in visual-first trials. Drift–diffusion modeling indicated that in humans, the effect of conflict on drift rate and non-decision time across SOA depended on modality-order, while response caution was not reliably affected by conflict. In rats, conflict reduced evidence accumulation efficiency—particularly in visual-first trials—and diminished response caution, with asymmetric non-decision time effects: slowing in auditory-first trials but accelerating in visual-first trials. These findings suggest that humans may adopt a flexible strategy for resolving conflict that preserves overall evidence quality, whereas rats use a more rigid strategy that prioritizes speed over accuracy, leading to compromised evidence quality and less efficient conflict resolution. This cross-species comparison provides computational description into potential evolutionary divergence in decision-making under conflict, although these interpretations should be considered in light of the substantial differences in training history between species.