An analysis of economic preferences and decision making using paired comparison choice experiments

• Main Author: Kingsley, David C
• Format: Dissertation
• Language: English
• Published: ProQuest Dissertations & Theses 01.01.2007
• Subjects: Airlines; Bids; Conjoint analysis; Decision making; Economic models; Economics; Experiments; Festivals; Methods; Preferences; Public good; Random variables; Researchers; Standard deviation; Statistical methods; Studies; Valuation; Wildlife sanctuaries; Willingness to pay
• ISBN: 9781109945379; 110994537X

The first chapter of this dissertation develops a random utility model specifically designed for paired comparison data. Previous methods require removing the choices made between two items and thus remove valuable information. Incorporating this data provides a preference ordering quite different from previous methods. Future research must determine which preference ordering better represents actual preferences. The second chapter of this dissertation investigates the decision-making process within paired comparison data. Results suggest that respondents are uncertain. Experience expressing one's preferences appears to reduce this preference uncertainty, referred to as preference refinement within this dissertation. Respondents are increasingly able to discriminate their choices between pairs of items as they progress through the experiment. This research suggests that answers to common nonmarket valuation surveys likely contain substantial noise and that respondents have the ability to refine their preferences and reduce this noise. The third chapter of this dissertation tests the hypothesis that preference uncertainty contributes to the disparity between an individuals Willingness to Pay and Willingness to accept. This value disparity is often found to be significantly larger than neoclassical economic theory would suggest. It is shown that allowing respondents to refine their preferences, using a paired comparison choice experiment, prior to eliciting valuations substantially reduces but is unable to eliminate the value disparity.