Accessibility statement

The Explanatory and Predictive Power of Non Two-Stage-Probability Models of Decision Making Under Ambiguity

John D Hey and Noemi Pace

Journal of Risk and Uncertainty 2014

Representing ambiguity in the laboratory using a Bingo Blower (which is transparent and not manipulable) and asking the subjects a series of allocation questions (which are more efficient than pairwise choice questions), we obtain data from which we can estimate by maximum likelihood methods (with explicit assumptions about the errors made by the subjects) a significant subset of the empirically relevant models of behaviour under ambiguity, and compare their relative explanatory and predictive abilities. Our results suggest that not all recent models of behaviour represent a major improvement in explanatory and predictive power, particularly the more theoretically sophisticated ones.

Keywords: Alpha Expected Utility, Ambiguity, Bingo Blower, Choquet Expected Utility, Contraction Model,  Subjective Expected Utility, Vector Expected Utility

Bingo Blower

You can find videos of the Bingo Blower on the linked page.

Visual Basic Program for the experiment

This is almost the one that was actually used in the experiment. You need to put the program in some folder and the files English Words, Italian Words and Questions in a sub-directory of that directory which must be called data. You can run the program in English or in Italian. The output will be printed to a file out.001 in that sub-directory. We should note that in the experiment itself, each subject got the questions in a randomly chosen order. In this demonstration version there is just one subject (number 1) and hence just one ordering.

When you execute the program you will find that the program stops when it writes "Please call over an experimenter". To continue it after this point you should do Control C.

VB program ( 116kb download)

English words ( 2kb download)

Italian Words ( 2kb download)

Questions ( 2kb download)

Gauss Program for the Analysis of the Data

You can fine the Gauss estimation program in the file hey and pace version for web.est below. A description of it can be found in the Word document Description of GAUSS program. It reads input files imp.01allz.out and ordz.out (available below) from the sub-directory c:\hey and pace\input data\ which you should rename if your directory structure is different. it writes the output to a file results.out in a sub-directory c:\hey and pace\output data\ which you should rename if your directory structure is different. We apologise if our code is messy.

hey and pace version for web.est ( 69kb download)

Description of GAUSS program (MS Word , 18kb)

imp.01 ( 1kb download)

allz.out ( 67kb download)

ordz.out ( 28kb download)


You can find the English and Italian instructions in the following Word Documents, alongside the English and Italian versions of the PowerPoint.

English Instructions (MS Word , 157kb)

English Powerpoint (MS PowerPoint , 442kb)

Italian Instructions (MS Word , 33kb)

Italian Powerpoint (MS PowerPoint , 481kb)

Background Mathematical Material

We have put all the background mathematical material in a document which you can download.

Background Mathematical Material ( 228kb download)