HMeta-d: Hierarchical Bayesian estimation of metacognitive efficiency from confidence ratings

Fleming, Stephen M.

doi:10.1101/068601

Cited by 91 publications

(206 citation statements)

References 47 publications

Supporting

Mentioning

202

Contrasting

Order By: Relevance

“…The resulting metric indicates normalized (z) evidence criteria needed to commit to a "high confidence" decisions; lower values therefor indicate more liberal type-2 criteria (less evidence is needed). The model was estimated from choice and confidence ratings for each participant using a Bayesian model called HMeta-d (Fleming, 2017) wherein the calculation of d' follows the model of (Lee, 2008). This open-source software is available at (https://github.com/metacoglab/HMeta-d).…”

Section: Methodsmentioning

confidence: 99%

The positive evidence bias in perceptual confidence is not post-decisional

Samaha

Denison

2020

Preprint

View full text Add to dashboard Cite

Confidence in a perceptual decision is a subjective estimate of the accuracy of one's choice. As such, confidence is thought to be an important computation for a variety of cognitive and perceptual processes, and it features heavily in theorizing about conscious access to perceptual states. Recent experiments have revealed a "positive evidence bias" (PEB) in the computations underlying confidence reports. A PEB occurs when confidence, unlike objective choice, overweights the evidence for the chosen option, relative to evidence against the chosen option.Accordingly, in a perceptual task, appropriate stimulus conditions can be arranged that produce selective changes in confidence reports but no changes in accuracy. Although the PEB is generally assumed to reflect the observer's perceptual and/or decision processes, postdecisional accounts have not been ruled out. We therefore asked whether the PEB persisted under novel conditions that eliminated two possible post-decisional accounts: 1) post-decision evidence accumulation that contributes to a confidence report solicited after the perceptual choice, and 2) a memory bias that emerges in the delay between the stimulus offset and the confidence report. We found that even when the stimulus remained on the screen until observers responded, and when observers reported their choice and confidence simultaneously, the PEB still emerged. Signal detection-based modeling also showed that the PEB was not associated with changes to metacognitive efficiency, but rather to confidence criteria. We conclude that once-plausible post-decisional accounts of the PEB do not explain the bias, bolstering the idea that it is perceptual or decisional in nature.

show abstract

Section: Methodsmentioning

confidence: 99%

The positive evidence bias in perceptual confidence is not post-decisional

Samaha

Denison

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Here we used metacognitive efficiency (meta-d' / d') to represent participants' metacognitive ability independent of primary decisionmaking performance. We quantified metacognitive ability through a hierarchical Bayesian Meta-d' model (Fleming, 2017). This computational model gives rise to more precise meta-d' estimation at both individual-and group-level (log-transformed) level, which allow direct comparison and correlation analyses of metacognitive abilities across conditions.…”

Section: Behavioural Analysismentioning

confidence: 99%

Diffusion property and functional connectivity of superior longitudinal fasciculus underpin human metacognition

Zheng

Wang

et al. 2020

Preprint

View full text Add to dashboard Cite

Metacognition as the capacity of monitoring one's own cognition operates across domains.Here, we addressed whether metacognition in different cognitive domains rely on common or distinct neural substrates with diffusion tensor imaging (DTI) and inter-areal functional connectivity techniques. After acquiring DTI and resting-state fMRI data, we asked participants to perform a temporal-order memory judgement task and a perceptual discrimination task, followed by trial-specific confidence judgments. DTI analysis revealed that the structural integrity (fractional anisotropy) in the anterior portion of right superior longitudinal fasciculus (SLF) was associated with both perceptual and mnemonic metacognitive abilities. When the mnemonic metacognitive ability was disrupted by TMS, the mnemonic metacognition scores did not correlate with SLF structural integrity anymore, revealing the causal relevance of this tract in memory metacognition. Furthermore, taking the dorsolateral PFC and IPL (both of which are connected by SLF) as seeds, we found perceptual and mnemonic metacognitive abilities to be associated with functional connectivity between DLPFC and VLPFC, whereas mnemonic metacognitive ability was selectively associated with connectivity between IPL and precuneus. These results illustrate the importance of SLF and a distinct white-matter grey-matter circuitry that supports human metacognition. Introduction:The capacity of reflecting on one's own cognitive process is known as metacognition (Flavell, 1979;Fleming and Dolan, 2012;Yeung and Summerfield, 2012). Metacognition has been considered as the most crucial function emerged during evolution (Hayes, 2016). Given its crucial functions, researchers have endeavored to understand how a metacognitive judgment is computed (Fleming and Daw, 2017;Kepecs et al., 2008;Zylberberg et al., 2016), how it is disrupted in psychiatric disorders (Hauser et al., 2017a;Rouault et al., 2018a), and how its accuracy can be improved (Carpenter et al., 2019). Metacognition is an umbrella term for the higher-level cognition about the lower-level cognition in various domains (e.g., perception and memory). As cognitive processes are implemented in the brain, an interesting and important question is whether the neural circuit supporting metacognition is the same or distinct across different cognitive domains.A large body of fMRI work have increasingly demonstrated a nuanced picture for this domaingenerality issue of metacognition. For example, the dorsolateral prefrontal cortex (DLPFC) might be involved in reading out the information of primary decision-making and using it for computing both perceptual and mnemonic metacognitive judgements (Chua and Ahmed, 2016;Fleming and Dolan, 2012;Kwok et al., 2019;Rounis et al., 2010). Morales et al. (2018) reported that the dorsal anterior cingulate cortex (dACC) was active during metacognitive evaluation in both memory and perception tasks but also found the activation patterns decoded from a perception task in the posterior medial frontal cortex and ven...

show abstract

“…We also checked our results with M.Ratio which has an equal-variance assumption. Furthermore, we repeated the analyses through hierarchical Bayesian estimation of meta-cognitive efficiency, known as HMeta-d method Fleming (2017). The Fleming's method provides opportunities to enhance statistical power Fleming (2017).…”

Section: Meta-cognitive Accuracymentioning

confidence: 99%

Higher Meta-cognitive Ability Predicts Less Reliance on Over Confident Habitual Learning System

Ershadmanesh

Miandari

Vahabie

et al. 2019

Preprint

View full text Add to dashboard Cite

Many studies on human and animals have provided evidence for the contribution of goal-directed and habitual valuation systems in learning and decision-making. These two systems can be modeled using model-based (MB) and model-free (MF) algorithms in Reinforcement Learning (RL) framework.Here, we study the link between the contribution of these two learning systems to behavior and meta-cognitive capabilities. Using computational modeling we showed that in a highly variable environment, where both learning strategies have chance level performances, model-free learning predicts higher confidence in decisions compared to model-based strategy. Our experimental results showed that the subjects' meta-cognitive ability is negatively correlated with the contribution of model-free system to their behavior while having no correlation with the contribution of model-based system. Over-confidence of the model-free system justifies this counter-intuitive result. This is a new explanation for individual difference in learning style. * Corresponding author (saraershad@ipm.ir) † These two authors contributed equally

show abstract

HMeta-d: Hierarchical Bayesian estimation of metacognitive efficiency from confidence ratings

Cited by 91 publications

References 47 publications

The positive evidence bias in perceptual confidence is not post-decisional

The positive evidence bias in perceptual confidence is not post-decisional

Diffusion property and functional connectivity of superior longitudinal fasciculus underpin human metacognition

Higher Meta-cognitive Ability Predicts Less Reliance on Over Confident Habitual Learning System

Contact Info

Product

Resources

About