Labb och grupper
ASLAB
Laboratoriet för applikationssystem.

CSELAB
Cognitive Systems Engineering Laboratory.

GIS-gruppen
Geografiska informationssystem

MDA
Laboratoriet för människan, datorn och arbetslivet.

NLPLAB
Laboratoriet för databehandling av naturligt språk
.

HCS: intro Forskarutbildning Seminarier
Forskning Grundutbildning Personal & kontakt

HCS seminarieserie vt 2004

Do I know what you want?

Annika Wallin
Centre for Adaptive Behaviour and Cognition, Max-Planck-Institut für Bildungsforschung, Berlin, and Lund University Cognitive Science.

Torsdag den 27 maj kl 13.15 - 15.00
IDA seminarierum Alan Turing (Estraden)

Imagine that you are sitting in a restaurant in a foreign country, trying to decide what to eat. You can deduce the content of certain dishes, but not enough to really pinpoint what you want. One option is to pick a dish somewhat at random and hope that you like it. Another is to look at the guy at the nearest table. What is he eating? And does he look content? Which option serves you best depends, among other things, on how similar his preferences are to yours. Is he on a weird diet? Then maybe you don't want the same food. Is he a gourmet chef? Go ahead and order his meal ÿ it is likely to be memorable.
To use the behaviour of others', that is, to utilise social information, is a shortcut to successful choices and behaviours in a broad range of domains. We follow the recommendations of experts, we go to the most popular movies, and we eat the porridge Gunde Svan prefers. Given that the experts know what they are talking about, we are representative of the average moviegoer, and Gunde Svans' success partly was caused by what he had for breakfast, this is a good idea.
The extent to which our preferences and the person(s) we are learning from are sufficiently similar will partly determine whether it makes it makes sense to use social information in choice or not. Mill puts it wonderfully: "Customs are made for customary circumstances and customary characters". The question I will try to answer today is whether, and how well, we judge the preferences of other individuals. If we can make such judgments well, we have a useful tool for distinguishing between situations in which social learning is likely to be successful and situations in which it is less likely to be so. We thus judge whether our circumstances and character are sufficiently similar to the man at the next table for us to imitate him, rather than to try to guess on our own.
In the talk I will present a field study in which we measured (among other things) how good people are in guessing the preferences of their friends and acquaintances. The setting is a popular Berlin café, and the participants, people who come there to have some coffee together. How good are they in predicting what their friends choose of the menu, and how do they do it?
(The study that will be presented is joint work with Barbara Fasolo, MPI, and London School of Economics)




Välkomna!




Senast uppdaterad: 2004-02-11, av: Sture Hägglund Ansvarig utgivare: Lars Ahrenberg © 2000