Colin Camerer - Robert Kirby Professor of Behavioural Economics at California Institute of Technology

Data mining in lab behavioural economics
'There is a huge amount of new data associated with economic choice and psychological factors,’ according to Colin Camerer. The internet, public data archives, but also the rich physiological and behavioural data we are able gather in experiments provides us with large sets of potential variables to consider. Using traditional methods, large data will easily lead to false discoveries and overfitting. Colin highlighted the usefulness of machine learning techniques to “data mine”. These methods guard against overfitting by clearly distinguishing between in-sample fit and out-of-sample prediction, and by accounting for the use of the same data to examine many different hypotheses. Colin went on to show two applications of these big data techniques in behavioural research. First, he presented a bargaining experiment where these methods were used to find the behavioural variables that best predict stalemates. Second, he showed the usefulness of big data methods in neuroscience, where brain data were used to infer peoples’ valuations of a public good.

Nick Chater - Professor of Behavioural Science at University of Warwick (Business School)

Virtual Bargaining: A micro-foundation for social decision making and interaction
This talk focussed on bargaining and communication. You can view the presentation slides or listen to a snippet of audio about behaviour that is 'hard to fake'

Paul Slovic - Professor of Psychology at University of Oregon 

(In)Actions Speak Louder than Words: Confronting the Collapse of Humanitarian Values in Foreign Policy Decisions
What are our true values when it comes to saving human lives? According to Paul, ‘Genocide is real, but we fail to feel that reality’. Paul’s inspiring talk shed light on psychological factors that can explain why we don’t intervene in mass murder and genocide. Among other reasons, Paul highlighted Psychic Numbing, according to which information fails to convey affect and emotion and therefore to motivate an appropriate action. Why? The value of human life decreases as the magnitude of the losses increases, because we exhibit diminishing sensitivity in the value we attach to lifesaving. People care less and less when losses of life become larger also due to feelings of inefficacy. Given the central role emotion and feelings play on how people respond to risk, when you become emotionless, you don’t care enough to react. Paul referred also on how the need to protect national security trumps saving foreign lives. In the threat of increasing risks to national security, the need to protect distant lives becomes less prominent and leads to inaction.   

Catherine Eckel – Sara & John Lindsey Professor at Texas A&M University

Sacrifice: Researching Terrorism in the Lab 
The paper analyses the behavioral determinants of terrorism as extreme intergroup punishment in a laboratory experiment. Participants are randomly assigned to groups and compete in a widely asymmetric tournament. One group has a large advantage on resources, and systematically takes half of the earnings of the disadvantaged group. The experiment studies the introduction of a very particular type of intergroup punishment: individuals in the losing group may unilaterally decide to attack the other group at a substantial personal cost: their whole earnings. We characterize the behavioral determinants of sacrifice by comparing different types of asymmetries between groups. Relative to a control treatment with no asymmetries, economic inequality alone has no significant effect on the likelihood of intergroup punishment. However, the results strongly support the link between political asymmetries (when the advantaged group decides the asymmetry size), and extreme intergroup punishment. Interestingly, more productive individuals are more likely to sacrifice themselves to harm the other group, and boost sacrifice intensity in the political inequality treatment.