NOTE: This item is not available outside the Texas A&M University network. Texas A&M affiliated users who are off campus can access the item through NetID and password authentication or by using TAMU VPN. Non-affiliated individuals should request a copy through their local library's interlibrary loan service.
Experimental studies of individual choice behavior in uncertain contexts
Abstract
The series of risky choice experiments that will be presented in this dissertation represent a major new initiative in the study of individual choice behavior in uncertain contexts. Three principal issues will be addressed. First, is individual choice behavior between relatively simple prospects consistent with the single attribute, cardinal utility approach of Von Neumann and Morgenstern's expected utility theory? Second, under what conditions do we observe risk averse, risk neutral or risk loving choice behavior? Finally, assuming that we find consistent and transitive preferences among subjects, the next phase of research would involve implementing conditions found to reliably result in intransitive choices among human subjects in verbal protocols. The experiments involve individual subjects (rats) who choose between real and highly valued alternatives with regular feedback regarding the effects of choice. Operant choice experiments in laboratories using non-human subjects are an often ignored and potentially rich source of empirical tests of the EU model and its variants. In marked contrast to human subject experiments, where payoffs are hypothetical or of little real value, these experiments consider lotteries where payoffs are real and of significant survival value to the animal. To the extent that one argues that "rational" behavior is at least in part learned behavior, generated as a consequence of interacting with one's environment, feedback is important. Three experiments are reported here. In the first experiment rats choose over prospects with equal expected values, which can be rank ordered according to a second degree stochastic dominance criteria. We test for risk aversion, transitivity of choices based on pairwise comparisons across several prospects, and, using a strength of preference measure, if the preference ordering matches the ordering in terms of the second degree stochastic dominance criteria. In the second experiment risk behavior at varying income levels is examined, where income is measured in terms of totel daily food intake. Income varies by a factor of 200%, with low income levels insufficient to cover longrun net energy requirements, and with high income levels approaching satiation. The third experiment is designed to induce violations of the independence axiom through mimicing conditions reliably resulting in violations with human subjects choosing over hypothetical outcomes. These involve tests of the "common ratio" or "certainty equivalent" effects.
Description
Typescript (photocopy).Subject
Entscheidung bei UnsicherheitEconomics
1984 Dissertation M135
Choice (Psychology)
Utility theory
Economics
Mathematical models
Collections
Citation
MacDonald, Don Neil (1984). Experimental studies of individual choice behavior in uncertain contexts. Texas A&M University. Texas A&M University. Libraries. Available electronically from https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -434992.
Request Open Access
This item and its contents are restricted. If this is your thesis or dissertation, you can make it open-access. This will allow all visitors to view the contents of the thesis.