Utility theory and attitudes towards risk in management decision making

Dickson, Gordon C. A (1979) Utility theory and attitudes towards risk in management decision making. MLitt(R) thesis, University of Glasgow.

Abstract

A recognition of the fundamental and pervasive nature of decision making has resulted in the decision making process assuming an importance that is quite distinct from the detail of any particular problem. A logical and systematic approach developed over many years that shifted the emphasis away from the terminal act of selecting one course of action to an examination of the structure of the whole decision. This decomposition of decisions highlighted problems associated with various components. One such difficulty was the manner by which the consequences of selecting a course of action were to be represented. With monetary outcomes the use of the real value of money does not always represent a person's true preference pattern and Utility Theory has been put forward as a substitute by which direct preferences and attitudes towards risk can be codified and transformed into a numerical scale. In this study Utility Theory was viewed as a normative decision making aid and two sets of experiments, involving twenty-nine business managers, were carried out, firstly to create personal utility functions and secondly to gauge to what extent they were workable. A self administered utility test was specially developed to obtain pound-utile co-ordinates and a least squares regression approach adopted to fit a quartic function to them. This method avoided many problems experienced by previous researchers, particularly the length of time demanded of each subject and the mathematical description of the pound - utile relationship. As a test of the workable nature of the derived functions a second set of experiments were carried out where each manager was presented with twenty-four hypothetical business decision problems, contained in narrative form. Testing the effectiveness of the utility functions could not be done simply by comparing actual and prescribed behaviour alone as the functions were being looked upon as normative and not necessarily predictive of what each subject would do. The managers were therefore assigned to one of three groups. Group one were given details of their utility prescriptions prior to tackling the decision problems. Group two were given their prescriptions after they had provided answers to the decisions and were then given the opportunity to make any alterations and group three were given no information about their prescriptions, A significant variation was detected between the three groups with those having knowledge of their prescriptions displaying the smallest differences between actual and prescribed behaviour. While the timing of this knowledge did not seem to be important the managers, all mature executives, made considerable use of the information they were given. A wide monetary range was covered by the decision problems but for subjects, with knowledge of their prescriptions, the mean difference between actual and prescribed behaviour, over the twenty-four decisions, was as low as

• Item Type: Thesis (MLitt(R))
• Qualification Level: Masters
• Additional Information: Adviser: Myles Dryden
• Keywords: Management
• Date of Award: 1979
• Depositing User: Enlighten Team
• Unique ID: glathesis:1979-72259
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 24 May 2019 15:12
• Last Modified: 24 May 2019 15:12
• URI: https://theses.gla.ac.uk/id/eprint/72259