Multiattribute Risky Choice Behavior: The Editing of Complex Prospects.

Abstract

This investigation draws upon concepts from prospect theory and multiattribute utility theory in an examination of the multiattribute risky choice behavior of 128 managers. The question of how managers code multiattribute prospects, and how coding relates to various independence assumptions, was explored. Results indicate that managers violate attribute independence in its general form, and in the form of the marginality assumption. The most common form of behavior was multiattribute risk aversion for prospects involving only gains and multiattribute risk seeking for prospects involving only losses. This result reinforces the importance of a target, reference point, or aspiration level that has been found in studies of single attribute risky choice. Furthermore, the result casts doubt on such commonly used multiattribute utility functions as the additive, multiplicative, and multilinear forms. Event independence, necessary for expectation models and a consequence of the cancellation of common components of prospects, was found to hold when the common values and probabilities were relatively small. When the common event had relatively large values and probabilities, there was some evidence that such events may influence choice.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1982
Accession Number
ADA111656

Entities

People

  • Dan J. Laughhunn
  • John W. Payne
  • Roy Crum

Organizations

  • Duke University

Tags

Communities of Interest

  • Biomedical
  • Human Systems

DTIC Thesaurus Topics

  • Air Force
  • Applied Psychology
  • Behavioral Sciences
  • Biological Sciences
  • Biomedical Research
  • Business Administration
  • Engineering
  • Human Factors Engineering
  • Information Science
  • Military Research
  • Operations Research
  • Probability
  • Probability Distributions
  • Psychology
  • Social Sciences
  • Students
  • Systems Engineering

Readers

  • Educational Psychology
  • Regression Analysis.
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.