Research
Research
What is our research about?“What would I remove if I had a magic wand? Overconfidence”
Daniel Kahneman’s famous fairy-tale wish (The Guardian, 18 Jul 2015) conveyed a deeply seated pessimism in cognitive scientists that overconfidence is hardwired into human cognition. It is true that this bias is pervasive, costly and the root cause of many human failures. But, previous failed attempts at reducing overconfidence always targeted individual decisions.
We propose to reduce this bias at SOCIAL level of decision making by determining the underlying mental, neural and social processes involved in overconfidence and testing these models by causal interventions in group situations.
We focus on 3 main directions of research:
1. Communication of Subjective Uncertainty
In honest communication of uncertainty (eg 2 doctors disagreeing over diagnosis), overconfidence impairs joint decisions. The problem here is that different people mean different things when they express the same level of confidence (see panel 1 below). Panel 1. Equality bias and misunderstanding of uncertainty in honest communication But when they make decisions together […]
1. Communication of Subjective Uncertainty
In honest communication of uncertainty (eg 2 doctors disagreeing over diagnosis), overconfidence impairs joint decisions. The problem here is that different people mean different things when they express the same level of confidence (see panel 1 below).
Panel 1. Equality bias and misunderstanding of uncertainty in honest communication
But when they make decisions together they use the simplifying heuristic that everybody means the same thing. This leads to a systematic misunderstanding of confidence across individuals (see panel 2).
Panel 2. Diversity of subjective notions of confidence leads to miscommunication of uncertainty
To solve this problem, We want to combine computational analysis of behaviour and brain response to develop a real-time neuro-feedback loop that allows weighting one’s opinion by an individually-tailored signature of one’s own uncertainty that calibrates different individuals’ subjective confidences against each other and removes overconfidence.
1) Mahmoodi, A., et al (2015) Equality bias impairs collective decision-making across cultures. PNAS (2015) 12, 3835–3840.
2) Navajas, J., et al (2017). The idiosyncratic nature of confidence. Nature Human Behaviour, doi:10.1038/s41562-017-0215-1.
3) Bang, D., et al (2017). Confidence matching in group decision-making. Nature Human Behaviour, 1 (6), 0117. doi:10.1038/s41562-017-0117
2. Uncertainty and Social Influence
By focusing on advising and consulting, we use a novel laboratory model that we have recently developed (ie Advising Game) to understand overconfidence helps us influence others when there is conflict of interest. We want to understand the mental and neural processes underlying strategic manipulation of others. Advising Game is played between a client and two […]
2. Uncertainty and Social Influence
By focusing on advising and consulting, we use a novel laboratory model that we have recently developed (ie Advising Game) to understand overconfidence helps us influence others when there is conflict of interest. We want to understand the mental and neural processes underlying strategic manipulation of others.
Advising Game is played between a client and two rival advisors. The Client (red), wants to buy a lottery ticket but does not know which one (Black or White) is better.
The client
Two advisers (blue and black) have some private information about the chances of the lotteries. They offer their advice to the client and, importantly compete with each other for the client’s limited attention.
The advisers
This game allows us to study the advisers’ behaviour and investigate how they transform their Facts, which the experimenter controls, to Opinions which they use to attract and maintain the client’s attention. We are interested in the psychological, cognitive and neurobiological processes at work in the adviser’s brain (see Question Mark below). We believe that this model system allows us to understand what goes on in today’s exuberant politician’s minds when they try to win the public opinion by hyperbole and lying.
Understanding the Cognitive Basis of Social Influence
Hertz, U., Palminteri, S., Brunetti, S., Olesen, C., Frith, CD., Bahrami, B. Neural computations underpinning the strategic management of influence in advice giving. Nature Communications https://doi.org/10.1101/121947
3. Wisdom of Small, Structured Crowds
The aggregation of many independent estimates can outperform the most accurate individual judgment. Such “wisdom of crowds” has recently been applied to problems ranging from the diagnosis of cancer to financial forecasting. It is widely believed that social influence undermines collective wisdom by reducing the diversity of opinions within the crowd. Our research shows that […]