Contingent valuation in health care: does it matter how the 'good' is described?

A general population sample of 104 Australian respondents completed an interviewer-administered contingent valuation (CV) survey that asked them to value five scenarios representing the same core improvement in health status. These scenarios varied only in the degree of narrative used to describe the condition causing the health problem being valued and labeling of this health problem. Results indicate no significant difference in willingness to pay (WTP) between expressing symptoms as a brief or moderate narrative, but a significantly lower WTP value when expressed in an extensive narrative. WTP also differed significantly according to condition 'labels'. Possible implications for CV research are outlined.