Testing the assumptions of rationality, continuity and symmetry when applying discrete choice experiments in health care
In the absence of revealed preference behaviour, economists use experimental techniques to estimate welfare changes. Such an approach relies on a number of assumptions concerning individual behaviour. This paper uses a discrete choice experiment to test three of these assumptions - rationality, continuity and symmetry. The experiment was carried out with users of a rheumatology clinic in the Grampian area of Scotland and was concerned with preferences for a specialist-nurse-led clinic. Two tests of 'rationality' were included in the experiment. Tests were carried out to see if respondents always chose on the basis of their preferred staffing, suggesting discontinuities in the utility function. The axiom of symmetry was tested using a split sample design, with respondents divided into two groups. Each group received a different questionnaire that varied with respect to the order of the choices. Over 30% of respondents provided at least one 'irrational' response. Such respondents did not differ significantly in their characteristics from 'rational' responses, suggesting that utility estimates would not be biased if this group were excluded from the analysis. Seventeen per cent of respondents showed signs of having non-compensatory utility functions. Evidence was found to support the axiom of symmetry. Future work should explore the axioms of rationality and continuity in more detail.
Volume (Year): 8 (2001)
Issue (Month): 1 ()
|Contact details of provider:|| Web page: http://www.tandfonline.com/RAEL20|
|Order Information:||Web: http://www.tandfonline.com/pricing/journal/RAEL20|
When requesting a correction, please mention this item's handle: RePEc:taf:apeclt:v:8:y:2001:i:1:p:59-63. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Michael McNulty)
If references are entirely missing, you can add them using this form.