| To Register SMDM Homepage | |
|
|
|
||||
Purpose: Using a traditional, open-ended willingness to pay (WTP) method can result in a highly skewed distribution with difficult to interpret outliers (e.g., “$10,000,000,000.00”), and can be confounded with respondents’ financial resources. In addition, in health contexts the traditional WTP measure can be relatively insensitive to objective differences in health conditions. We conducted two studies testing a newly devised open-ended format where participants were asked to express WTP in terms of a percentage of their available income. Study 1 directly compared this new measure to the traditional open-ended dollar format. Study 2 added a manipulation to determine if the advantage of the percentage format would persist when changed to a per month format.
Methods: We distributed written, anonymous surveys to members of the general public in a hospital cafeteria (study 1 n = 230, study 2 n = 208). The participants were randomly assigned to either a traditional open-ended WTP question or to the new measure using percentage of available income. Participants read descriptions of quality of life with a below-the-knee amputation and a spinal cord injury and were asked their WTP for a hypothetical treatment that could restore functioning for each condition. Order of presentation of the two conditions was counterbalanced.
Results: Study 1, which directly compared the two WTP measures, indicated that the newly devised ‘percentage’ WTP measure showed fewer outliers (p < .01), had better overall distributional properties, and was not correlated with income (r = .01; the traditional WTP was strongly correlated with income, r = .40, p < .001). This method proved to be more sensitive to the difference between the amputation condition and the more severe spinal cord injury condition than the traditional dollar amount method (p < .05). The difference in sensitivity persisted even after attempts to transform the traditional measures to address the distributional properties. Study 2 demonstrated that the advantage in sensitivity to severity persisted when approached as a per month format (p < .01).
Conclusions: An alternative to the traditional open-ended WTP format in which respondents indicated their willingness to pay a percentage of their available income showed better distributional properties, greater sensitivity to severity of health condition and was not correlated with income. This new method could be useful in many research applications.
See more of Poster Session - Public Health; Methodological Advances
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)