Research Article
A Study on the Explanatory Power Analysis of Technical Decision Theory
Published: January 1995 · Vol. 24, No. 1 · pp. 29-52
Full Text
Abstract
This study attempted to analyze the explanatory power of descriptive decision theory using analytical techniques from normative decision theory. Since decision theories under uncertainty require both simplicity and high explanatory power to be practically useful, the analysis of explanatory power holds very significant meaning. The HILO structure developed by Chew and MacCrimmon was used as the analytical technique. Using this HILO structure, Chew and MacCrimmon analyzed that their weighted linear model could explain 6 out of 16 choice patterns. Meanwhile, Becker used the HILO structure to analyze the explanatory power of prospect theory, which was proposed to enhance the explanatory power of normative decision theories under uncertainty. Since prospect theory, as a descriptive theory, does not possess the mathematical model that normative theories have, Becker utilized the properties of the proposed weighting function to verify that this theory could explain 6 out of 16 choice patterns. In this study, we analyzed the explanatory power of venture theory, which was recently proposed as a model capable of explaining choice behavior under a broad range of uncertainty situations. In venture theory, the value function from prospect theory is used as is, while the variables affecting the decision weight coefficients are specified differently from prospect theory. This theory explains the cognitive and motivational processes by which people assign decision weight coefficients through anchoring and simulation. These decision weight coefficients are represented by graphs of the venture function expressed as a function of probability. While the HILO structure was adopted as the analytical method, since venture theory only represents decision weights through several venture curves, analysis could only be conducted based on these curves at present; therefore, a new graphical method was developed and used in this study. The analysis results verified that these venture curves could explain 12 out of 16 choice patterns in the HILO structure. Furthermore, a survey of 77 respondents revealed that 93% of the subjects selected the 12 choice patterns that venture curves could explain. This supports the explanatory power of venture theory.
