Journal of Marketing Behavior > Vol 3 > Issue 2

Mental Accounting for Percentages Revisited: The Interplay of a Computational Error and Constituent Outcome Categorization

Haipeng (Allan) Chen, Gatton College of Business and Economics, USA, allanchen@uky.edu Haoying Sun, Gatton College of Business and Economics, USA, haoying.sun@uky.edu
 
Suggested Citation
Haipeng (Allan) Chen and Haoying Sun (2018), "Mental Accounting for Percentages Revisited: The Interplay of a Computational Error and Constituent Outcome Categorization", Journal of Marketing Behavior: Vol. 3: No. 2, pp 153-165. http://dx.doi.org/10.1561/107.00000049

Published: 13 Nov 2018
© 2018 H. A. Chen and H. Sun
 
Subjects
Behavioral Decision Making,  Individual Decision Making,  Consumer Behavior
 

Article Help

Share

Download article
In this article:
Prospect Theory and Mental Accounting
Computational Error and Categorization of Constituent Outcomes
Study
Appendix: Measures Used in the Study
Measure of the computational error:
Measure of ha ppiness:
Purchase intention:
Manipulation check:
References

Abstract

Drawing upon Kahneman and Tversky's (1979) prospect theory, Thaler (1985) proposed the mental accounting principles, which Heath et al. (1995) applied to outcomes expressed in the percentage format. Following the emerging literature on how consumers process percentages (Chen et al. 2012; Chen and Rao 2007), we identify the existence of a computational error and make predictions on how the error interacts with the categorization of a mixed percentage gain to affect hedonic framing. Our empirical evidence replicates the mental accounting principle among mathematically unsophisticated individuals who are misled by the computational error to perceive economically different but nominally equivalent options as the same when the constituent outcomes of a mixed percentage gain are categorized in different mental accounts. When they are categorized in the same mental account, however, unsophisticated individuals are indifferent between economically different but nominally equivalent options. We conclude by discussing how the computational error interacts with categorization of constituent outcomes to produce framing effects.

DOI:10.1561/107.00000049