Tomas Knapen, Department of Experimental & Applied Psychology, Free University Amsterdam, is this week’s visitor in the SWE colloquia series

Mapping the dark side: Visual selectivity of default network deactivations

The brain’s default network (DN) deactivates when participants focus externally to perform a task, and activates for internally referenced mental states such as mind wandering and autobiographical memory. Processing in the DN is thought to represent the highest levels of information integration, and changes to their responses are implicated in many psychological disorders.  Recent findings indicate that signals in the DN carry visual memory information, but the functional role of DN deactivations in particular remains unclear. Here we show that BOLD signal decreases in the DN are tuned to the spatial location of visual stimuli. The visual selectivity of these deactivations was similar to that of concurrent activations in the frontal and parietal regions of the multiple demand network. Furthermore, visually selective deactivations allowed us to decode the location of a visual stimulus from DN nodes, demonstrating that the DN contains functional representations of the visual field. Our results indicate that responses in the DN are pinioned to responses in the visual system, providing a candidate organization for the mnemonic functionality of the DN. Our results suggest that the DN may utilize sensory reference frames for higher-level cognition such as autobiographical memory and social thought.

On 25 October, Ruben Arslan visits from the Center for Adaptive Rationality, Max Planck Institute for Human Development, to give a talk as part of the SWE colloquia.

What can we learn from one simple question? Reconciling cognitive scientists with self reports by treating them as a task

If we are to take self-reports of risk preferences seriously, we should understand how people understand our questions and which risks they think about when we ask a general question. People can give us a consistent answer that predicts actual behaviour, but how do they actually do it? I show preliminary data based on coding texts that people wrote about the risks they took. I hope this can help us derive a fuller idea of what the lay conception of risks look like and teach us to ask better questions and design better tasks.

Professor JOHN LIST, PhD, University of Chicago, USA, will this week deliver the 8th Bernoulli Lecture for the Behavioral Sciences

Using field experiments to make the world a better place

Thursday, October 18, 2018, 18:15
Bernoullianum, Grosser Hörsaal, Bernoullistrasse 30, Basel

John List uses field experiments to learn about the economics of life and to improve people’s outcomes. He has generated data that has provided insights into many issues, including pricing behavior, discrimination in the marketplace, the valuation of non-marketed goods and services, public good provisioning, behavioral anomalies, charitable giving, auction theory, and the role of the market in the development of rationality. In this lecture, he will focus on how to use field experiments to make the world a better place through a wide range of issues, including charitable giving, education, lowering inequality and helping organizations as varied as Uber and the UK government.

John List serves as Kenneth C Griffin Distinguished Service Professor of Economics at the University of Chicago, where he is also the Chairman of the Department of Economics. He received his PhD from the University of Wyoming in 1996 and hold positions at the University of Central Florida, the University of Arizona and the University of Maryland, College Park, before moving to Chicago. He served as Senior Economist in the President’s Council of Economic Advisers and was elected Fellow of the American Academy of Arts and Sciences in 2011 and the Econometric Society in 2015.

The Bernoulli Lectures for the Behavioral Sciences honor researchers who have contributed significantly to the development of the behavioral sciences, particularly in the fields of psychology and economics. The Bernoulli lectures are organized yearly by the Bernoulli Network for the Behavioral Sciences, a joint initiative of the Department of Psychology and the Faculty of Business and Economics of the University of Basel, with the aim of fostering interdisciplinary dialogue in the behavioral sciences.

Carlos Alós-Ferrer, Department of Economics, University of Zurich, is this week’s guest speaker at the SWE Colloquium.

Strength of preference, dual processes, and economic decision making

In a series of experiments, we investigate the sources of errors in economic decisions under risk. The first source is a “strength of preference” effect. Error rates increase and response times become longer as choice difficulty increases, in analogy with well-known psychophysical effects in perceptual discrimination tasks. However, in economic tasks the scale determining choice difficulty is not always obvious. When there is an objectively correct answer independently of attitudes toward risk, the effects (for error rates and response times) are explained by the distance between expected values (or winning probabilities). Differences in payoff-independent numerical magnitudes influence these gradual effects, but, contrary to pure discrimination tasks, they play a minor role. When correct answers depend on subjective attitudes toward risk, the gradual effects on error rates and response times are explained by “strength of preference,” as evidenced by cardinal differences in independently estimated, subjective utilities. The second source of errors reflects the presence of mental shortcuts capturing heuristics or behavioral rules, which interact with and modify the gradual effects arising from the strength of preference. The interaction of qualitatively different decision processes delivers testable hypotheses on errors and response times. Specifically, whenever decision processes are in conflict (prescribe different responses), error rates increase and errors are faster than correct responses. If different processes are in alignment, however, errors are slower than correct responses. These predictions become a test for the presence of multiple decision processes in economic decision making and complement the views derived from strength of preference.