Ž . Journal of Risk and Uncertainty, 20:2; 161176 2000  2000 Kluwer Academic Publishers. Manufactured in The Netherlands. Which Error Story is Best? ENRICA CARBONE carbone@unisannio.it ecarbone@unina.it JOHN D. HEY jdh1@york.ac.uk Department of Economics, Uni  ersity of York, Heslington, YORK YO1 5DD, UK. 00 44 1904 433786 Abstract Ž . Ž . Two recent papers, Harless and Camerer 1994 and Hey and Orme 1994 are both addressed to the same question: which is the ‘best’ theory of decision making under risk? As an essential part of their separate approaches to an answer to this question, both sets of authors had to make an assumption about the underlying stochastic nature of their data. In this context this implied an assumption about the ‘errors’ made by the subjects in the experiments generating the data under analysis. The two different sets of authors adopted different assumptions: the purpose of this current paper is to compare and contrast these two different error stories in an attempt to discover which of the two is ‘best’. Key words: errors, decision making under risk, experiments JEL Classification: D81, C91 Two recent papers concerned with the empirical investigation of theories of Ž . decision making under risk, Harless and Camerer 1994 and Hey and Orme Ž . 1994 , were both addressed to the same question: which is the ‘best’ theory of decision making under risk? As an essential part of their separate approaches to an answer to this question, both sets of authors had to make an assumption about the underlying stochastic nature of their data. In the context of deterministic theories of decision making under risk this involved an assumption about the ‘errors’ made by the subjects in the experiments generating the data under analysis. The two different sets of authors adopted different assumptions: the purpose of this current paper is to compare and contrast these two different error stories in an attempt to discover which of the two is ‘best’. A direct comparison of these two different error stories from the original two papers is not possible as they differed in a second crucial way in terms of the way that the data was fitted: Harless and Camerer fitted the data across all subjects whilst Hey and Orme fitted the data subject by subject. Since the Harless and Camerer data does not allow us to fit the Ž Hey and Orme error story because the number of observations per subject is not . sufficiently large we are obliged to use the Hey and Orme data and fit both error stories to that. This note reports on the results of so doing.