Goodbye and good luck to Marion!

We are sad to say goodbye to Marion Rouault who is leaving the lab to take up a new position at the Ecole Normale Supérieure in Paris. Marion was one of the founding members of the MetaLab, and had a fantastically productive few years here in London - we will greatly miss having her around, and wish her all the best for her next steps!

We had a lab picnic on a lovely spring evening on Primrose Hill to give Marion a fitting send off:

marion_picnic.jpeg

New paper on cognitive offloading published in Cognition

Xiao’s paper on a role for metacognition in cognitive offloading to the environment has now been published in Cognition. A Twitter thread summarising the findings can be found here:

xiao_offloading.jpeg

New paper published in Nature Communications

Marion’s paper entitled “Forming global estimates of self-performance from local confidence” has been published in Nature Communications.

A blog post by Marion summarizing the findings is here.

Brain .png

In the paper we developed a new paradigm to investigate how external feedback and local decision confidence relate to global self-performance estimates (SPEs) , and whether local fluctuations in decision confidence inform SPEs when external feedback is unavailable. We showed that fluctuations in confidence contribute to global SPEs over and above effects of objective performance and reaction times. Surprisingly, we found that humans tend to underestimate their performance in the absence of feedback, compared with a condition with full feedback, despite objective performance being similar in the two cases. We hope that study of global self-beliefs formation will advance our understanding of self-efficacy distortion in psychiatric disorders.

New paper published in Current Biology

Max’s paper entitled "Metacognitive Failure as a Feature of Those Holding Radical Beliefs" has been published in Current Biology.

Figure 4. Individual differences in radicalism are captured by a choice bias model.

Figure 4. Individual differences in radicalism are captured by a choice bias model.

We found that radical participants (as measured by scores on dogmatism and authoritarianism questionnaires) showed reduced insight into the correctness of their choices and less sensitivity to post-decision evidence, indicating a generic resistance to revising mistakes. The effect sizes were small, but robust and replicable in a second independent dataset. Given that the tasks were far removed from real-world issues we think it’s striking that basic difference in metacognition predict answers to questions indicative of radical beliefs. A blogpost unpacking these findings further can be found here.

New paper published in Personality Neuroscience

A paper entitled "Human Metacognition Across Domains: Insights from Individual Differences and Neuroimaging" has been published in Personality Neuroscience. Congratulations to Marion and Andy!

The article reviews recent literature on the domain-generality of human metacognition, drawing on evidence from individual differences, behavioural studies and neuroimaging. A meta-analysis of behavioural studies found perceptual metacognitive ability being correlated across different sensory modalities, but not between metacognition of perception and memory. Evidence from neuroimaging studies provided a complementary perspective on the domain-generality of metacognition, revealing co-existence of neural signatures that are common and distinct across tasks. 

Figure 5. Different methodologies for quantifying brain structure and function shed light on the underpinnings of metacognition across domains.

Figure 5. Different methodologies for quantifying brain structure and function shed light on the underpinnings of metacognition across domains.

Steve awarded a Philip Leverhulme Prize

Steve has been awarded a 2018 Philip Leverhulme Prize for Psychology from the Leverhulme Trust, together with Prof. Nichola Raihani (also at UCL), Prof. Emily Cross, Dr. Claire Haworth and Dr. Harriet Over. Leverhulme awards are granted annually in recognition of the international impact and future potential of the work. Steve plans to use the £100k award to support work in the lab developing new models of conscious awareness. Read more here!

Steve receiving the award from Astronomer Royal Lord Martin Rees

Steve receiving the award from Astronomer Royal Lord Martin Rees

Leverhulme_Trust_CMYK_black.jpg

MetaLab at Latitude

Last weekend we pitched up at Latitude Festival as part of the BrainHub, organized together with James Kilner's group. We used this as an opportunity to get some user feedback about the latest prototype version of "MetacogMission" - an app that Andy McWilliams is developing together with London-based technology company DamnFine to study metacognition via a series of decision-making games. Thanks for everyone who took part - we got tons of useful feedback and pilot data that we are now starting to sift through.

Here's a brilliant timelapse video of the BrainHub put together by Alex Galvez-Pol in James' lab. We are also very grateful to WCHN Public Engagement coordinator Cassie Hughill who held everything together and kept everyone fed and watered, and to Ben from DamnFine for being on hand for remote tech support all the way from New York! And last but not least, Max, Matan, Xiao and Alisa who worked hard to make it a great event.

The MetaLab team

The MetaLab team

Paper on decision confidence in PNAS

Dan Bang's paper isolating decision confidence from its component parts has recently been published in PNAS - congratulations Dan! An open access preprint can be found here, and code/data supporting the study are on our lab GitHub site.

In this paper we devised a novel psychophysical task that allowed us to independently manipulate the certainty about the perceptual input (the direction of a cloud of moving dots) and the difficulty of the decision (how close the dots were to a reference line). Both factors influenced confidence. Using fMRI, we found that some brain regions previously implicated in the encoding of decision confidence actually tracked either perceptual certainty (such as the parietal cortex) or decision difficulty (such as pre-SMA), whereas confidence was uniquely tracked in medial PFC (see image below). This may explain why prefrontal damage is regularly associated with dysfunction in metacognition (higher-level estimates of task performance), without affecting "lower level" encoding of beliefs.

BangPNAS2018.png