Methodological comparison of alternative standard gamble chaining procedures
McNamee, P. and Scotland, G. (HERU) and Bhattacharya, S. (Applied Clinical Sciences, University of Aberdeen)
Empirical evidence has shown the standard gamble to be internally inconsistent – i.e. there is a tendency for values elicited indirectly, by two-staged gambles chained through the failure outcome, to exceed those elicited directly through standard reference gambles. It has been postulated that this phenomenon is due to ‘loss aversion’, and diminishing marginal sensitivity. The objective of this project was to assess whether internal consistency varies depending on the procedure adopted in the chaining process.
Utility values for adverse birth outcomes were elicited from women waiting to undergo IVF treatment using standard reference gambles and different chaining procedures. Analysis suggests that gambles chained through the failure outcome are systematically higher than standard reference gambles. Furthermore, values chained through the next best health state appear to be systematically higher than gambles chained through the worst ranked health state.
Scotland, G.S. and McNamee, P. A methodological comparison of alternative standard gamble chaining procedures. HESG, York. July 2006.