Hi all,
Here's our (untimely) abstract. Any comments would be much appreciated.
Thanks,
Dennis & Jeremy
How Healthy Are We? A disadvantage to using ranked data for health state
valuation
Health state valuations (HSVs) constitute a fundamental building block for
measuring health system performance. To obtain HSVs, investigators often ask
respondents to value levels of disability using the time trade-off (TTO)
method. This procedure is costly, time-consuming, and difficult to
implement. Salomon (2003) proposes a model that estimates HSVs using only
ranked data, tests it with the 1993 British General Public Survey, and finds
the results nearly identical to a TTO-based model. We extend Salomon (2003)
by allowing HSVs to vary according to individual characteristics like age
and marital status, but find that some quantities of interest cannot be
estimated with the rank data. Our results show one limitation to using
ordinal data to measure cardinal quantities.