Most lab-experiments study risky behaviour through "lotteries" whose properties (ie the full list of outcomes and their associated likelihoods) are explicitly described. A common finding in this setting is that people tend to overweight the likelihood of rare events. A criticism of this approach is that in everyday life, such "descriptions" are not often available. To explore more naturalistic situations, cognitive psychologists have recently come up with an alternative experimental paradigm that asks subjects to search information about risky prospects by themselves. Unlike the standard "decisions from description" finding, behaviour in “decisions from experience” is consistent with underweighting the likelihood of rare events.
Orestis revisited this 'Description–Experience gap' by conducting a lab-experiment where subjects received information about risky prospects in one of the following three conditions:
- from description
- from experience with a 'history table' recording previous events
- from experience without a 'history table'
The history table proved influential during information search, drawing subjects' attention away from elements such as a lottery's variance or the time-period and steering it towards the table's properties, such as its max capacity of recorded events.
The analysis of risky choices across treatments revealed a significant 'Description-Experience gap' which is mitigated by the presence of the history table. In this version of the "gap", subjects overweight the likelihood of rare events when making decisions from experience but less so compared to when making choices from descriptions.
The paper received financial support from the ESRC Network for Integrated Behavioural Science (NIBS1) via its small grant scheme. Professor Chris Starmer, Director of NIBS says; "At his career point to have a solo authored paper in a good field journal is a tremendous achievement for Orestis. Furthermore I am delighted to announce, after a competitive selection process, Orestis will be joining our NIBS team from September 2018 as a Research Fellow/Assistant for a period of two years".
A copy of the paper is available online and can be found in Theory and Decision, May 2018, Volume 84, Issue 3, pp 311–339.
View all featured publications
Posted on Thursday 3rd May 2018