COMPUTATIONAL AND NEURAL MECHANISMS OF JUDGMENT AND INTROSPECTION

Abstract

This thesis aimed to investigate the mechanisms underlying decision-making and subjective introspection, in both perceptual and economic domains. Using behavioral, computational, and neuroimaging methods, our research not only confirms that expectation, even arbitrary, could bias the perceptual judgment, but also makes the distinct theoretical point that this prediction bias could facilitate metacognitive sensitivity of the judgment by serving as a cue to engage on deeper reasoning. In the economic decision-making domain, we provide a mechanistic link among choices, speed, and confidence in assessing metacognitive variations, and the neuroimaging results suggest that metacognitive representation of subjective decisions might reflect an automatic monitoring process in the brain. Altogether, our findings suggest that human decision-making is a self-correcting system in which individuals can flexibly adjust their behaviors and evaluate their actions based on accessible resources in the context such as expectations, reaction speed, and behavioral associations.