Neural basis of metacognition

Columbia Univ New York Morningside, New York NY

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Humans are able to reflect upon ongoing cognition, an ability known as metacognition. Metacognition can be measured as one's ability to introspectively distinguish between effective and ineffective cognitive processing, for instance, by rating confidence to distinguish between correct and incorrect trials in a task. In this project we have two specific aims. First, we try to address the question of whether there is a single common mechanism for metacognition across different task modalities (e.g., memory and vision), or specific independent, domain C specific mechanisms. We take an individual differences approach to investigate the structural brain differences (assessed via structural magnetic resonance imaging, MRI) that facilitate metacognitive efficiency. We develop original psychophysical measurements to quantify an individual s metacognitive efficiency, and investigate the neural bases for metacognition in both verbal memory and visual perception. We measure metacognition in two tasks (memory and vision) and apply structural equation modeling (SEM) to analyze anatomical variability in healthy individuals. We test hypotheses regarding which brain regions facilitate metacognitive efficiency generated from these analyses in patients with lesions in those target regions. Second, we investigate the functional mechanisms supporting metacognition in order to explain, and investigate why normal human subjects often do not perform at optimal levels in metacognitive tasks. There are likely two sources of such suboptimalities. On the sensory side, we test a model according to which perceptual confidence is heuristically based on only part of the perceptual signal. We test this by using intracranial electrocortigraphy (ECoG) in human (surgical epileptic) subjects. On the response output side, one candidate model suggests that subjects evaluate the efficacy of their cognitive processes based on retrospective monitoring of their behavioral response systems. If this model is correct, then subjective confidence may depend on the level of noise or conflict evoked in the motoric response system as perceptual judgments are formed. We test this model by directly injecting noise into the motor system at different times during a perceptual task via transcranial magnetic stimulation (TMS). This project is timely as it involves testing novel hypotheses about metacognitive mechanisms, inspired by recent studies and our own preliminary findings. Some of these tests are also only made possible due to our development of novel psychophysical and modeling approaches, which will be of use to other researchers in the field. Finally, the use of neuropsychology, TMS, and intracranial electrophysiology to test formal models of the functional mechanisms of metacognition will help us to arbitrate between existing views in the literature, which, in human studies, currently remains largely confined to behavioral rather than neurobiological investigations. Our project will help to bridge the gap between these two related fields, fostering the development of a mechanistic account of metacognition that can inform clinical and everyday applications (such as cognitive training).