Prefrontal Contributions to Metacognition in Perceptual Decision Making

Prefrontal Contributions to Metacognition in Perceptual Decision Making Fleming Huijgen Dolan

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Prefrontal Contributions to Metacognition in Perceptual Decision Making 2012 Journal of Neuroscience 10.1523/jneurosci.6489-11.2012 Fleming2012-prefrontal-contributions-to-metacognition-in Fleming2012-prefrontal-contributions-to-metacognition-in.txt, the project's staged copy 1404160f10df0702f9a4c34f0cfe7df5beb28f28 not-redistributable neuroscience

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them the neural basis of metacognitive awareness of one's own perceptual decisions Quotation withheld because this record is not cleared for redistribution. which brain region supports metacognitive report, as opposed to the decision itself Quotation withheld because this record is not cleared for redistribution. identifies right rostrolateral PFC by three pre-stated criteria, and proposes it re-represents object-level uncertainty in a form usable for report Quotation withheld because this record is not cleared for redistribution. mechanism Quotation withheld because this record is not cleared for redistribution. metacognition of perceptual decisions Quotation withheld because this record is not cleared for redistribution. fmri fmri Quotation withheld because this record is not cleared for redistribution. mechanisms rostrolateral PFC re-represents object-level decision uncertainty at the meta level, which is what makes an explicit confidence report possible Quotation withheld because this record is not cleared for redistribution. mechanisms object-level uncertainty signals computed within the decision circuitry are re-coded for use in report, rather than confidence being read off the decision itself Quotation withheld because this record is not cleared for redistribution. mechanisms metacognitive accuracy arises from trial-by-trial access to internal decision uncertainty, not from monitoring one's own reaction time Quotation withheld because this record is not cleared for redistribution. a staircase equates objective performance across individuals, so metacognitive ability is not confounded by task performance Quotation withheld because this record is not cleared for redistribution. the Follow condition is yoked to the Report condition, matching the motor act of moving the cursor without any confidence judgment Quotation withheld because this record is not cleared for redistribution. the starting cursor position is jittered to prevent motor preparation from carrying the confidence signal Quotation withheld because this record is not cleared for redistribution. no feedback is given, so confidence cannot be updated from knowledge of accuracy Quotation withheld because this record is not cleared for redistribution. the proposal: object-level uncertainty represented elsewhere is re-represented in rlPFC in a form suitable for explicit report Quotation withheld because this record is not cleared for redistribution. metacognitive ability is measured as area under the type 2 ROC, distinct from type 1 sensitivity Quotation withheld because this record is not cleared for redistribution. The region-wise correlations linking confidence-related activity to metacognitive accuracy were not corrected for multiple comparisons. what this limits: the region-wise correlation results only: TESTING FOR REGIONAL CORRELATIONS WAS CARRIED OUT WITHOUT CORRECTION FOR MULTIPLE COMPARISONS, AND THUS A POSITIVE RESULT HERE SHOULD BE TEMPERED BY THIS CAVEAT Quotation withheld because this record is not cleared for redistribution. The study cannot separate the respective contributions of rostrolateral (BA10) and dorsolateral (BA46) prefrontal cortex to metacognitive ability. what this limits: any attribution to a specific prefrontal subregion: THE RESPECTIVE ROLE OF RLPFC (BA10) AND DLPFC (BA46) IN METACOGNITIVE ABILITY IS NOT YET CLEAR Quotation withheld because this record is not cleared for redistribution. Whether the rlPFC account generalises beyond perceptual decisions to mnemonic metacognition is not settled by these data. what this limits: the domain-generality claim: FURTHER COMPARISON OF PERCEPTUAL AND MNEMONIC TASKS IS REQUIRED to separate domain-specific from domain-general substrates Quotation withheld because this record is not cleared for redistribution. the regional correlations were tested without correction for multiple comparisons, and the authors say the positive result should be tempered by it what this limits: the same scope as limitations[0] - the uncorrected regional correlations Quotation withheld because this record is not cleared for redistribution. some connectivity activations did not survive correction and are reported uncorrected what this limits: the visual-cortex activations: ADDITIONAL ACTIVATIONS WERE OBSERVED IN PRIMARY AND EXTRASTRIATE VISUAL CORTEX THAT DID NOT SURVIVE CORRECTION FOR MULTIPLE COMPARISONS (p < 0.001, UNCORRECTED) Quotation withheld because this record is not cleared for redistribution. the respective roles of rlPFC and dlPFC in metacognitive ability are stated to be unclear what this limits: the same scope as limitations[1] - the respective roles of rlPFC and dlPFC in metacognitive ability cannot be separated here Quotation withheld because this record is not cleared for redistribution. whether the region serves metacognition outside perceptual decisions is an open question what this limits: the domain of the rlPFC finding: WHETHER rlPFC ACTIVITY PLAYS A ROLE IN METACOGNITION IN DOMAINS OUTSIDE OF EVALUATION OF PERCEPTUAL DECISION MAKING IS AN OPEN QUESTION Quotation withheld because this record is not cleared for redistribution. a localiser scan was collected and then dropped because the main analysis found no visual cortex effects what this limits: the visual-cortex question specifically, which was closed by the main analysis rather than by the data collected for it: AS OUR INITIAL WHOLE-BRAIN ANALYSES DID NOT FIND EFFECTS IN VISUAL CORTEX, WE DO NOT ANALYZE THESE DATA FURTHER HERE Quotation withheld because this record is not cleared for redistribution. three subjects excluded for performance, one more from the intelligence analysis what this limits: the sample every effect rests on: WE EXCLUDED THREE SUBJECTS FROM FURTHER ANALYSIS DUE TO TYPE 1 d' SCORES >2 SDs FROM THE GROUP MEAN. TWENTY-THREE SUBJECTS WERE INCLUDED Quotation withheld because this record is not cleared for redistribution. metacognitive ability did not correlate with fluid intelligence, reported as a null what this limits: any account linking metacognition to general ability: METACOGNITIVE ABILITY DID NOT SIGNIFICANTLY CORRELATE WITH FLUID INTELLIGENCE (g) Quotation withheld because this record is not cleared for redistribution. the type 1 versus type 2 distinction, which is what makes metacognition measurable separately from performance Quotation withheld because this record is not cleared for redistribution. the object-level versus meta-level model, imported as the framework for perceptual metacognition Quotation withheld because this record is not cleared for redistribution. a MATCHED control condition against which self-report activity was compared eliminates: that right rostrolateral prefrontal activity reflects the act of reporting, the button press or the task's difficulty rather than metacognition · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. metacognition metacognition Quotation withheld because this record is not cleared for redistribution. metacognitive ability is defined so as to be independent of task performance - measured by type 2 ROC after a staircase equates type 1 sensitivity across people Quotation withheld because this record is not cleared for redistribution. object level and meta level are distinguished as monitoring and controlling versus being monitored, which is what makes the two separable in principle Quotation withheld because this record is not cleared for redistribution. confidence is instructed as RELATIVE rather than absolute, because the task is too hard for certainty Quotation withheld because this record is not cleared for redistribution. the paper distinguishes object-level uncertainty, which other studies measure, from metacognitive report, and argues the earlier work cannot separate them without concurrent reports Quotation withheld because this record is not cleared for redistribution. claim_type: existence a second reading DIFFERING from the recorded value. Recorded: mechanism — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. method: fMRI during a face/house discrimination with staircased difficulty and concurrent confidence ratings, against a yoked motor-matched control condition a second reading DIFFERING from the recorded value. Recorded: fmri — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. phenomenon: metacognition of decision performance a second reading DIFFERING from the recorded value. Recorded: metacognition of perceptual decisions — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. extends → Rounis et al. (2010); Fleming et al. (2010b); Yokoyama et al. (2010) Quotation withheld because this record is not cleared for redistribution. replicates → Song et al. (2011); Maniscalco and Lau (2012) Quotation withheld because this record is not cleared for redistribution. builds-on → Fleming et al. (2010b); Kornbrot (2006) Quotation withheld because this record is not cleared for redistribution. extends → structural and TMS evidence that anterior PFC supports metacognitive accuracy Quotation withheld because this record is not cleared for redistribution. qualifies → studies of decision confidence that manipulate difficulty without collecting concurrent confidence reports Quotation withheld because this record is not cleared for redistribution. distinguishes-from → left rlPFC, dACC and right PPC, which track confidence but do not predict metacognitive ability differs in: THE AXIS IS TRACKING VERSUS PREDICTING. 'Quotation withheld.' All these regions carry a confidence signal; only one predicts how good the metacognition is. Two different tests, and the regions separate on the second - so a study that only measured tracking would report them as equivalent. Quotation withheld because this record is not cleared for redistribution. contradicts → the expectation that metacognitive ability tracks fluid intelligence Quotation withheld because this record is not cleared for redistribution. metacognitive ability —operationalised-as→ type 2 metacognitive ability type 2 strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. meta —distinct-from→ object meta object strength of the relation: 1 · note: the model that licenses treating them as separable at all Quotation withheld because this record is not cleared for redistribution. metacognitive ability —distinct-from→ fluid intelligence metacognitive ability fluid intelligence strength of the relation: 0.8 · note: argued from a null, which the paper reports plainly Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but metacognitive ability did NOT CORRELATE with fluid intelligence; distinct-from rests on a null correlation. Quotation withheld because this record is not cleared for redistribution. rlPFC —distinct-from→ dACC rlPFC dACC differs in: THE AXIS IS WHETHER CONFIDENCE-RELATED ACTIVITY PREDICTS METACOGNITIVE ABILITY. 'Quotation withheld.' All these regions TRACK confidence; only one predicts how good the metacognition is. Tracking and predicting are different tests and the regions separate on the second, not the first. · strength of the relation: 0.9 · note: both track confidence; only rlPFC predicts metacognitive ability, which is the paper's third criterion doing the separating Quotation withheld because this record is not cleared for redistribution. signal detection theory signal detection theory Quotation withheld because this record is not cleared for redistribution. receiver operating characteristic / area under the curve receiver operating characteristic / area under the curve Quotation withheld because this record is not cleared for redistribution. t-test (paired) t-test (paired) Quotation withheld because this record is not cleared for redistribution. t-test (one-sample) t-test (one-sample) Quotation withheld because this record is not cleared for redistribution. correlation correlation Quotation withheld because this record is not cleared for redistribution. partial correlation partial correlation Quotation withheld because this record is not cleared for redistribution. regression regression Quotation withheld because this record is not cleared for redistribution. general linear model general linear model Quotation withheld because this record is not cleared for redistribution. random-effects group analysis random-effects group analysis Quotation withheld because this record is not cleared for redistribution. multiple-comparison correction multiple-comparison correction Quotation withheld because this record is not cleared for redistribution. standard deviation standard deviation Quotation withheld because this record is not cleared for redistribution. type 1 signal detection theory type 1 signal detection theory Quotation withheld because this record is not cleared for redistribution. area under the type 2 ROC area under the type 2 ROC Quotation withheld because this record is not cleared for redistribution. psychophysiological interaction psychophysiological interaction Quotation withheld because this record is not cleared for redistribution. orthogonalisation against reaction time orthogonalisation against reaction time Quotation withheld because this record is not cleared for redistribution. Right rostrolateral prefrontal cortex met all three of the paper's criteria for a role in metacognition: greater activity during self-report, correlation with confidence, and prediction of metacognitive ability across people. Quotation withheld because this record is not cleared for redistribution. Confidence-related activity was exclusively negative: dACC, right posterior parietal cortex and bilateral rlPFC scaled inversely with reported confidence and no region scaled positively with it. Quotation withheld because this record is not cleared for redistribution. Metacognitive ability was statistically independent of perceptual task performance, replicating an earlier dissociation. Quotation withheld because this record is not cleared for redistribution. Simply moving the confidence cursor in the control (Follow) condition did not modulate activity in the confidence-sensitive regions, showing the effect is specific to genuine self-report. Quotation withheld because this record is not cleared for redistribution. Functional connectivity of right rlPFC with contralateral prefrontal cortex and occipital cortex rose during metacognitive report relative to the control condition. Quotation withheld because this record is not cleared for redistribution. The confidence effect in right rlPFC was not an artefact of reaction time, since it survived orthogonalizing confidence with respect to RT. Quotation withheld because this record is not cleared for redistribution. The individual-differences effect was lateralized: left rlPFC showed no relationship between confidence-related activity and metacognitive ability. Quotation withheld because this record is not cleared for redistribution. right rlPFC activity was greater during confidence report than in the yoked control Quotation withheld because this record is not cleared for redistribution. the strength of the confidence-activity relation in right rlPFC predicted metacognitive ability across people, r = -0.47 Quotation withheld because this record is not cleared for redistribution. no such correlation in left rlPFC, r = 0.024 Quotation withheld because this record is not cleared for redistribution. confidence-related activity was absent in the Follow condition, so using the scale is not sufficient to produce it Quotation withheld because this record is not cleared for redistribution. activity did not differentiate correct from incorrect trials, so it is not an error signal Quotation withheld because this record is not cleared for redistribution. the confidence effect survived orthogonalising against reaction time Quotation withheld because this record is not cleared for redistribution. connectivity between right rlPFC and contralateral PFC and occipital cortex rose during report Quotation withheld because this record is not cleared for redistribution. metacognitive ability did not correlate with fluid intelligence Quotation withheld because this record is not cleared for redistribution. A roc A roc Quotation withheld because this record is not cleared for redistribution. psychophysiological interaction (PPI) psychophysiological interaction (PPI) Quotation withheld because this record is not cleared for redistribution. metacognitive ability metacognitive ability Quotation withheld because this record is not cleared for redistribution. three criteria three criteria Quotation withheld because this record is not cleared for redistribution. decision confidence decision confidence Quotation withheld because this record is not cleared for redistribution. fMRI time series fMRI time series Quotation withheld because this record is not cleared for redistribution. metacognitive ability metacognitive ability Quotation withheld because this record is not cleared for redistribution. type 1 reaction time type 1 reaction time Quotation withheld because this record is not cleared for redistribution. motion correction parameters motion correction parameters Quotation withheld because this record is not cleared for redistribution. reported confidence reported confidence Quotation withheld because this record is not cleared for redistribution. n = 23 — number of participants retained for analysis after exclusions, with age range and mean age Quotation withheld because this record is not cleared for redistribution. 70.6 ± 3.2 — mean percentage of correct face/house decisions held near threshold by the staircase during scanning Quotation withheld because this record is not cleared for redistribution. r = −0.47, p = 0.025 — across-subject correlation between the right rlPFC confidence beta and metacognitive ability (A roc), the paper's key individual-differences result Quotation withheld because this record is not cleared for redistribution. mean R 2 = 0.98 — goodness of fit of the type 2 signal detection model to the confidence rating data Quotation withheld because this record is not cleared for redistribution. 1087 ms vs 971 ms — mean reaction time on incorrect versus correct perceptual decisions Quotation withheld because this record is not cleared for redistribution. mean r = −0.32 ± 0.09 — within-subject correlation between reaction time and reported confidence Quotation withheld because this record is not cleared for redistribution. 26 recruited, 3 excluded for performance, 23 analysed — the sample after exclusions Quotation withheld because this record is not cleared for redistribution. r = -0.47, p = 0.025, partial r = -0.46 controlling for d prime — the third criterion in right rlPFC Quotation withheld because this record is not cleared for redistribution. r = 0.024, p = 0.91 — the same test in left rlPFC Quotation withheld because this record is not cleared for redistribution. t(22) = 2.35, p = 0.028 — the confidence effect after controlling for reaction time Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.1523/jneurosci.6489-11.2012.