Chapter 1 of the thesis is divided into 5 sections. Section A begins by defining a Discrete Choice Experiment (DCE), and outlines the key stages involved in conducting a DCE. Sections B and C outline theories underpinning DCE analysis. Section B outlines the characteristics theory of demand, whilst section C, explains random utility theory (RUT), compensating variation (CV), marginal willingness to pay (MWTP), and willingness to pay (WTP) analysis. Section D of the thesis provides a review of the DCE literature. Section E outlines the research questions addressed in the thesis including calculating WTP and hypothetical bias; the description of the cost attribute; preference heterogeneity; and altruism. Chapter 2 shows how DCEs can be used to calculate WTP, using a DCE relating to Deep Vein Thrombosis (DVT). Chapter 3 uses data from a DCE applied to Menstrual disorder and Gynaecology patients. It evaluates an experimental method I developed to establish whether respondents might fail to factor in the monetary attribute into their DCE decision making, leading to hypothetical bias. Chapter 4 applies essentially the same DCE design but only analyses data from Gynaecology patients. Chapters 4-8 all use data obtained from a DCE relating to preferences for different allocation criteria for allocating kidneys for transplantation. Chapters 5 and 6 look at preference heterogeneity which is observable using interaction dummy variables (the issue of unobserved preference heterogeneity is considered in chapter 7). Chapter 5 establishes how marginal rates of substitution (MRS) differ between different respondent groups including renal patients, healthcare professionals, live donors / relatives of deceased donors, carers, and ethnic minority versus non-ethnic minority patients. Chapter 6 establishes how MRS differs between non-white ethnic minority patients versus other patients; South Asian patients versus other patients; and according to respondent gender. Chapter 7 of the thesis compares results from models which do not cater for unobserved preference heterogeneity, with results from models which do. Initially 2 basic models which do not cater for preference heterogeneity at all (because they do not include dummy variables) are applied including random effects logit and conditional logit. Then models catering for unobserved preference heterogeneity including Mixed Logit and a Latent Class Model (LCM) are used. Finally there is an analysis involving the application of conditional logit with interaction dummy variables. Chapter 8 of the thesis explores how preferences might differ according to how altruistic respondents are. It establishes how respondent preferences differ according to respondent self-disclosed perspective when answering DCEs. In other words whether they claimed to answer the DCE in terms of what would be best for me; what would be best for me and others; or what is best for others. Finally chapter 9 involves a discussion of the findings emerging from the thesis, and draws conclusions about the merits of material contained in it.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:714858 |
| Date | January 2013 |
| Creators | Clark, Michael D. |
| Publisher | University of Warwick |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://wrap.warwick.ac.uk/88792/ |
Page generated in 0.0109 seconds