poly: Are people roughly Bayesian?

>Date: Thu, 7 Jan 1999 12:18:20 -0600
>Reply-To: admin@SSRN.COM
>Sender: Microeconomics Working Paper Abstracts
> <MICRO-WPS@PUBLISHER.SSRN.COM>
>From: Economics Research Network <Admin@SSRN.COM>
>Subject: ERN Micro WPS Vol. 4, No. 1, 01/07/1999
>To: MICRO-WPS@PUBLISHER.SSRN.COM
>
>"Broadening the Tests of Learning Models"
>
> BY: STEPHEN KITZIS
> Fort Hays State University
> HUGH KELLEY
> University of California at Santa Cruz
> Department of Economics
> ERIC BERG
> University of California, Santa Cruz
> DANIEL FRIEDMAN
> University of California at Santa Cruz
> DOMINIC MASSARO
> University of California at Santa Cruz
>
>Paper ID: University of California at Santa Cruz, Department of
> Economics Working Paper #401
> Date: June 1998
>
> Contact: DANIEL FRIEDMAN
> Email: Mailto:dan@cash.ucsc.edu
> Postal: University of California at Santa Cruz
> Santa Cruz, CA 95064 USA
> Phone: (831)459-4981
> Fax: (831)459-5900
> Co-Auth: STEPHEN KITZIS
> Email: Mailto:PSSK@FHSUVM.FHSU.EDU
> Postal: Fort Hays State University
> Hays, KS 67601 USA
> Co-Auth: HUGH KELLEY
> Email: Mailto:hukelley@cats.ucsc.edu
> Postal: University of California at Santa Cruz
> Department of Economics
> Santa Cruz, CA 95064 USA
> Co-Auth: ERIC BERG
> Email: Mailto:eberg@cats.ucsc.edu
> Postal: University of California, Santa Cruz
> Santa Cruz, CA 95064 USA
> Co-Auth: DOMINIC MASSARO
> Email: Mailto:massaro@fuzzy.ucsc.edu
> Postal: University of California at Santa Cruz
> Santa Cruz, CA 95064 USA
>
>Paper Requests:
> Send a check for $6 payable to "UC Regents" to Ms. Marilyn
> Chapin, Department of Economics, 217 Social Sciences I,
> University of California, Santa Cruz, CA 95064.
>
>ABSTRACT:
> For many years psychological studies of the learning process
> have used a simulated medical diagnosis task in which symptom
> configurations are probabilistically related to diseases.
> Participants are given a set of symptoms and asked to indicate
> which disease is present, and feedback is given on each trial.
> We enrich this standard laboratory task in four different ways.
> First, the symptoms have four possible values (low, medium high,
> and high) rather than just two. Second, symptom configurations
> are generated from an expanded factorial design rather than a
> simple factorial design. Third, subject are asked to make a
> continuous judgment indicating their confidence in the
> diagnosis, rather than simply a binary judgment. Fourth,
> cumulated performance scores, payoffs, and the availability of a
> historical summary of the outcomes are varied in order to assess
> how these treatment modulate performance. These enrichments
> provide a broader data set and more challenging tests of
> models.
>
> Using 123 subjects each in 480 trials, we compare five
> existing learning models plus several variants, including the
> well-known Bayesian, Fuzzy Logic, connectionist, exemplar and
> ALCOVE models. We find that subjects do learn to distinguish the
> symptom configurations, that subjects are quite heterogeneous in
> their response to the task, and that only a small part of the
> variation across subjects arises from the differences in
> treatments. The most striking finding is that the model that
> best predicts subjects' behavior is a simple Bayesian model with
> a singe fitted parameter for prior precision to capture
> individual differences. We use recent rolling regression
> techniques to elucidate the behavior of this model over time and
> find some evidence of overresponse to current stimuli.
>
>
>JEL Classification: D83
>
>

Robin Hanson
hanson@econ.berkeley.edu http://hanson.berkeley.edu/
RWJF Health Policy Scholar FAX: 510-643-8614
140 Warren Hall, UC Berkeley, CA 94720-7360 510-643-1884
Received on Fri Jan 8 13:23:09 1999