Neural Mechanisms

Question 1

What are the issues with looking at neural mechanisms for metacognition?

Answer 1

Difficult to separate from activity that’s part of cognitive processes, e.g. does activity relate to evidence (part of decision process) or confidence (metacognitive judgement)

Question 2

Give a study showing that cognition and metacognition are dissociable processes

Answer 2

Shimamura and Squire 1986 - korsakoff’s amnesiacs (damage to MTL memory system from alcoholism) relatively matched to alcoholics without memory issues for recognition of non recalled facts (memory performance) but worse correlated feeling of knowing judgements (metamemory performance)

Question 3

What’s the PFC got to do with metacognition? Give some studies

Answer 3

Strongly involved in metacognition. E.g. Fleming et al. 2014 - manipulated performance on which circle had more dots in it to be equal, found deficit in confidence judgement accuracy in aPFC lesion patients compared to healthy and lesions elsewhere
Fleming et al. 2010 aPFc grey matter Volume correlated with metacognitive ability invidual difference level when control perceptual performance

Question 4

Review the ways people have tried to dissociate metacognitive processes. Why is the evidence a mess?

Answer 4

Process account (different types, prospective versus retrospective) - suggests domain generaltiy, versus content account (different domains, memory versus perceptual). Evidence to support both - selective impairment with lesions and difference in activity/connection strength with fmri, but also many exceptions to both. Metacognition likely not a pure process as use multiple cues and which cues using (and therefore which areas getting cues from) depends on particular judgement.

Question 5

Explain the posterior anterior dissociation in confidence and give some evidence

Answer 5

Levels of metacognitive judgement. Basic confidence judgement from evidence strength assesment in posterior and context based more nuanced judgement in anterior incorporating other cues e.g. confidence heuristic.
Bang and Fleming 2020 - Posterior activity varies with confidence regardless of context, anterior sensitive to expression of confidence from decision partner (degree to which need to adjust judgement based on whether partner in extremes of confidence)
Shekhar and Rahnev 2018 - theoretical model where posterior receives sensory signal and reads out its strength for the anterior to make final judgement.
Shekhar and Rahnev 2018 - TMS to posterior reduces confidence. TMS to anterior doesn’t reduce confidence but does improve accuracy. Less confident if can’t assess evidence, Improved accuracy if only using evidence for judgement when not using cues from elsewhere?