London Judgment & Decision Making Group
Winter term 2012 – 2013
Organizers
Emmanouil Konstantinidis
University College London
Contact details:
Department of Cognitive, Perceptual & Brain Sciences
Room 204c, 26 Bedford Way, London, WC1H 0AP
UK
Telephone: (+44) 020 7679 5364
E-mail: emmanouil.konstantinidis.09@ucl.ac.uk
Neil Bramley
University College London
Contact details:
Department of Cognitive, Perceptual & Brain Sciences
Room 201, 26 Bedford Way, London, WC1H 0AP
UK
E-mail: neil.bramley.10@ucl.ac.uk
LJDM website
http://www.ljdm.info
Web administrator:
Dr Stian Reimers (Stian.Reimers.1@city.ac.uk)
LJDM members’ (Risk & Decision) list
Contact: Dr Marianne Promberger (marianne.promberger@kcl.ac.uk)
Seminar Schedule
October – December 2012
5:00 pm in Room 313, 26 Bedford Way, UCL Psychology
10th October
The Clash of the (Formal Probability Theory) Titans
Emmanuel Pothos
City University London
17th October
Microinsurance Decisions: Evidence from Ethiopia
Daniel Clarke
University of Oxford
24th October
Risk Sensitivity as Ecologically Valid Inference
Jerker Denrell
University of Warwick
31th October
Competition and Cooperation between Model-based and Model-free Decisions
Peter Dayan
Gatsby Computational Neuroscience Unit / UCL
07th November
NO SEMINAR – UCL READING WEEK
14th November
The Importance of Meta-Cognition for Group Decision Making
Chris Frith
University of Oxford / UCL
21st November
Do we have Bayes’ theorem hardwired in the cortico-basal-ganglia circuit?
Rafal Bogacz
University of Bristol
28th November
Title TBA
Adam Harris
UCL
05th December
Kindernomics: Other-regarding Preferences and Game Behaviour in Children
Liam Pollock
Queen Mary, University of London
12th December
Modelling legal arguments using Bayesian idioms
Martin Neil
Queen Mary, University of London and Agena Ltd
Abstracts
10.10.2012
Emmanuel Pothos
City University London
The Clash of the (Formal Probability Theory) Titans
One of the dominant traditions in cognitive modelling is classic (Bayesian) probability (CP)
theory. Yet considerable evidence has accumulated that human judgment often goes against
classical principles. We discuss quantum probability (QP) theory as an alternative formal
probabilistic framework for understanding cognition. In QP theory, probabilistic assessment is
often strongly context and order dependent, individual states can be superposition states (which
are indefinite with respect to some specific judgment), and composite systems can be entangled
(they cannot be decomposed into simpler subsystems). We review two fundamental empirical
findings (the conjunction fallacy, violations of the sure thing principle in prisoner’s dilemma
games), which have persistently challenged classical theory, yet have natural instead of national
accounts within quantum theory.
17.10.2012
Daniel Clarke
University of Oxford
Microinsurance Decisions: Evidence from Ethiopia
We review experimental evidence collected from a framed microinsurance lab experiment using
poor subjects in rural Ethiopia. In line with results from recent field experiments the demand for
index insurance is increasing in wealth and decreasing with risk aversion at low levels of wealth.
However, this relationship is nonlinear, consistent with the hump-shape prediction of decreasing
absolute risk averse expected utility theory (DARA EUT). In contrast with recent field
experiments where it is not possible to demonstrate that low demand for indexed insurance is 'too
low', use of a laboratory experiment with an objectively known joint probability distribution
allows normative statements to be made about the observed level of demand. The observed level
of demand for index insurance in the experiment is higher than the DARA EUT upper bound,
strengthening the case for research methods that allow comparison between the observed demand
for indexed insurance products and demand which would be consistent with objective financial
advice, without which it is not possible to make normative statements about the level of observed
demand.
24.10.2012
Jerker Denrell
University of Warwick
Risk Sensitivity as Ecologically Valid Inference
Existing explanations of why humans and animals are sensitive to risk attribute it to cognitive
imperfections such as diminishing sensitivity and heuristic choice processes. This paper argues
that risk sensitivity can be explained as ecologically valid inferences about the expected value of
alternatives. Theoretically, I show that even a risk neutral Bayesian decision maker, who is
uncertain about the reliability of observations, should use variability in observed outcomes as an
indicator of low expected value for outcomes above a reference level and as an indicator of high
expected value for outcomes below a reference level. Empirically, I demonstrate that combining
past outcomes using an s-shaped value function leads to accurate predictions about future values.
These theoretical and empirical results suggest the hypothesis that risk sensitivity can be
explained as ecologically valid inferences about the attractiveness of uncertain alternatives.
31.10.2012
Peter Dayan
Gatsby Computational Neuroscience Unit / UCL
Competition and Cooperation between Model-based and Model-free Decisions
Substantial recent work has explored multiple mechanisms of decision-making in humans and
other animals. Functionally and anatomically distinct modules have been identified, and their
individual properties have been examined using intricate behavioural and neural tools. I will
discuss the background of these studies, and describe a number of recent experiments that have
cast a stronger, albeit more confusing, light on the interactions between these mechanisms. I will
also talk about other related dichotomies of control.
07.11.2012
UCL Reading Week
14.11.2012
Chris Frith
University of Oxford / UCL
The importance of meta-cognition for group decision making.
Using simply psychophysical tasks we have shown that, through discussion, two people working
together can make better decisions than the best member of the pair working alone. Analysis of
the discussions suggests this advantage depends upon explicit meta-cognition since successful
pairs introspect on their perceptual experience and report this to each other. But group decisions
are not always advantageous. In particular, good group decisions require a degree of independence
between the participants. I shall discuss some of the factors that can compromise this
independence.
21.11.2012
Rafal Bogacz
University of Bristol
Do we have Bayes’ theorem hardwired in the cortico-basal-ganglia circuit?
This talk will present a model assuming that during decision making the cortico-basal-ganglia
circuit computes probabilities that considered alternatives are correct, according to Bayes’
theorem. The model suggests how the equation of Bayes’ theorem is mapped onto the functional
anatomy of a circuit involving the cortex, basal ganglia and thalamus. The talk will also present
relationship of model’s prediction to experimental data, ranging from detailed properties of
individual neurons in the circuit, to the effects of disruptions of computations in this circuit on
behaviour.
28.11.2012
Adam Harris
UCL
TBA
05.12.2012
Liam Pollock
Queen Mary, University of London
Kindernomics: Other-regarding Preferences and Game Behaviour in Children
Human decision-making in social scenarios is often guided by other-regarding preferences, a
phenomenon that makes us unique as a species. A common means of exploring such preferences
in adults experimentally is through the use of behavioural economics. Although this
methodology’s use with adults is well established, its use with children is comparatively rare,
something that needs to be rectified in order to properly chart the ontogenetic development of
other-regarding preferences. We present here the results of a novel behavioural economics
experiment, originally designed for primates, but performed in this case on both adult and child
participants. This paradigm adapts the traditional ultimatum game by adding theft to the first
player’s decision options. Results show a highly significant difference in the response patterns of
adults and children, with the former demonstrating markedly greater pro-sociality in their
decision-making, and the latter’s decision making influenced by a significant effect of age. The
extent to which these results may have been influenced by the experimental setting is discussed,
as are implications for other-regarding preferences in general.
12.12.2012
Martin Neil
Queen Mary, University of London and Agena Ltd
Modelling legal arguments using Bayesian idioms
A Bayesian network (BN) is a graphical model of uncertainty that is especially well-suited to legal
arguments, since it helps visualise and model dependencies between different hypotheses and
pieces of evidence, as well as calculate the revised probability beliefs about all uncertain factors.
Although BNs have been widely discussed and recently used in the context of legal arguments
there is no systematic, repeatable method for modelling legal arguments as BNs. Hence, where
BNs have been used in the legal context, they are presented almost as a fait accompli, with no
justification into the reasoning and working that support their construction. This renders the
process of building BNs for legal arguments ad-hoc, with little possibility for learning and
improvement. This talk describes a method for building useful legal arguments in a consistent and
repeatable way. The method complements and extends recent work on objected-oriented BNs for
complex legal arguments and is based on the recognition that such arguments can be built up from
a small number of basic causal structures (referred to as idioms). During the talk I present a
number of examples that demonstrate the practicality and usefulness of the method.