Alternative experimental design paradigms in choice experiments and their effects on consumer demand estimates for beef from endangered local cattle breeds: An empirical test

In this paper we empirically validate the hypothesis of context independent preferences in choice experiments under different experimental design approaches. Two alternative paradigms for constructing designs are analysed in this study under a simple Multinomial Logit (MNL) framework: optimal orthogonal in the differences (OOD) and D-efficient (EFD) designs. Once having taken into account potential differences in scale (inversely proportional to the variance of the random term in random utility models), preference estimates for beef attributes are found to be sensitive to the design generation process.