Review of Behavioral Economics > Vol 5 > Issue 2

A Flexible and Customizable Method for Assessing Cognitive Abilities

Andrea Civelli, University of Arkansas, USA, andrea.civelli@gmail.com , Cary Deck, University of Alabama, USA
 
Suggested Citation
Andrea Civelli and Cary Deck (2018), "A Flexible and Customizable Method for Assessing Cognitive Abilities", Review of Behavioral Economics: Vol. 5: No. 2, pp 123-147. http://dx.doi.org/10.1561/105.00000081

Publication Date: 13 Sep 2018
© 2018 A. Civelli and C. Deck
 
Subjects
Behavioral Economics
 
Keywords
JEL Codes: C9C90C91
Cognitive Abilities TestsRaven’s MatricesExperimental Economics Tools
 

Share

Download article
In this article:
1. Introduction 
2. The Matrix Puzzles 
3. Experimental Procedures 
4. Puzzle Performance 
5. Validity of the Matrices as a Substitute for the RPM 
6. Conclusions 
A. The Software to Generate the Matrices 
B. Experimental Designs and Details 
References 

Abstract

This paper describes the properties of a set of puzzles that are behaviorally similar to those of the common Raven Progressive Matrix test. Our puzzles consist of a three-by-three grid of images with the lower right element omitted. Each image is characterized by six characteristics that can vary along several patterns. Lab experiments demonstrate that the puzzles become more challenging as the number of characteristics that change increases. Further, the ability to correctly solve our puzzles is shown to be correlated with scores on the Raven Progressive Matrix test and with performance in a beauty contest game. Due to the manner in which our puzzles are constructed, there are a large number of unique puzzles that can be generated for use in economics experiments using software described in the paper. Thus our puzzles are well suited for use as an alternative method to assess the cognitive ability of respondents and for use as a real effort task with multiple levels of cognitive difficulty.

DOI:10.1561/105.00000081

Replication Data | 105.00000081_supp.zip (ZIP).

This file contains the Matlab code.

DOI: 10.1561/105.00000081_supp