A Caltech Library Service

Uncovering Behavioral Strategies: Likelihood-Based Experimental Data Mining

El-Gamal, Mahmoud and Grether, David M. (1993) Uncovering Behavioral Strategies: Likelihood-Based Experimental Data Mining. Social Science Working Paper, 850. California Institute of Technology , Pasadena, CA. (Unpublished)

[img] PDF (sswp 850 - Jul. 23, 1993) - Submitted Version
See Usage Policy.


Use this Persistent URL to link to this item:


Economists and psychologists have recently been developing new theories of decision making under uncertainty that can accommodate the observed violations of standard statistical decision theoretic axioms by experimental subjects. We propose a procedure which finds a collection of decision rules that best explain the behavior of experimental subjects. The procedure is a combination of maximum likelihood estimation of the rules together with an implicit classification of subjects to the various rules, and a penalty for having too many rules. We prove that our procedure yields consistent estimates (as the number of tasks per subject and the number of subjects go to infinity) of the number of rules being used, the rules themselves, and the proportion of our population using each of the rules. We apply our procedure to data on probabilistic updating by subjects in four different universities. We get remarkably robust results which show that the most important rules used by the subjects are Bayes rule, representativeness rule (ignoring the prior), and conservatism (over-weighting the prior). In our procedure, the subjects are allowed to make errors, and our estimated error rate is typically 20%.

Item Type:Report or Paper (Working Paper)
Additional Information:The first coauthor acknowledges financial support from NSF grant #SES-9223701 to the California Institute of Technology. We thank the Jet Propulsion Laboratory for giving us access to their Cray YMP, on which all of our calculations were made.
Group:Social Science Working Papers
Funding AgencyGrant Number
Series Name:Social Science Working Paper
Issue or Number:850
Record Number:CaltechAUTHORS:20170824-160040708
Persistent URL:
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:80777
Deposited By: Jacquelyn Bussone
Deposited On:28 Aug 2017 23:48
Last Modified:03 Oct 2019 18:34

Repository Staff Only: item control page