Measuring bifurcation points in choice behavior: principles and illustration

This paper suggests a new stated adaptation response format to measure bifurcation points in choice behavior. The method assumes that individuals attach bifurcation points to critical attributes beyond which they change their behavior. Fibonacci numbers are used to identify these bifurcation points efficiently. The resulting data are analyzed using hazard models. The suggested new methodology is illustrated for a sample of 335 travellers who were about to travel by train from Hong Kong to Guangzhou. Respondents were asked to state whether they would retain the choice of train if the travel cost by train were increased by certain amounts, varied according to the Fibonacci principle. Experience with the implementation of the suggested methodology showed that most respondents found the experiment quite interesting. We were able to identify successfully the bifurcation point for all respondents who agreed to participate in the exercise. It took each respondent only a few minutes. The method was thus proved to be effective and efficient. By and large, the estimated coefficients seem consistent with findings elsewhere and in the anticipated direction. This suggests that the data collected by the proposed methodology have face validity.