Output details

Can scale and coefficient heterogeneity be separated in random coefficients models?

This paper was co-authored with Professor John Rose at the University of Sydney. It shows that a recent trend in choice modelling, namely that of aiming to capture scale heterogeneity across agents, has been based on misunderstanding of underlying theory and misguided interpretation of model estimates. In an open review of the paper, Professor Kenneth Train – an internationally-leading choice modeller – applauded the authors for ‘putting some sanity back into the topic’.