Testing the External Validity of Contingent Valuation: A Field Experiment Comparing Hypothetical and Real Payments
Posted: 13 Jun 2007
Abstract
Background: Whilst the dichotomous choice and payment card contingent valuation (CV) approaches are both being increasingly used in health economics, very few studies have tested their external validity. This study developed a field experiment to directly test external validity of both techniques. Consideration was also given to methods of calibration of hypothetical and real behaviour.
Methods: The experiment compared hypothetical WTP with actual purchase decisions for a prepaid dental care plan among 298 study subjects in Thailand. Subjects initially received either a dichotomous choice or payment card CV questionnaire. Four bids, 200, 250, 300 and 400 Baht, were included in both formats. Responses to the bid amount included both a qualitative scale (ranging from 'yes, absolutely' pay the amount to 'No absolutely not') and a 0-10 certainty scale (where 0 represented 'absolutely uncertain' whether they would pay the specified amount and 10 'absolutely certain'). Following the hypothetical WTP questionnaire subjects were given the option of paying their stated WTP and purchasing the dental plan. The McNemar test was used to test for differences in proportions between hypothetical and actual yes responses across both CV formats.
Results: The percentage of hypothetical yes-responses significantly exceeded that of actual yes-responses at the 5% level of significance for both WTP formats. There was no significant difference in performance across the two formats. However, the proportions of hypothetical and actual WTP were not significantly different after the hypothetical WTP data were calibrated using both calibration methods. Comparison of the two calibration methods suggests that the 'certainty score scale' method is preferred to the 'qualitative' method.
Discussion: The findings from this study have important implications for the future application of WTP in health economics. They suggest that the development of calibration methods may strengthen the external validity of the CV technique.
Keywords: contingent valuation, external validity, calibration
JEL Classification: B41
Suggested Citation: Suggested Citation