Abstract
This study presents a comprehensive review on discrete-continuous models. We classify discretecontinuous models into (1) structural type models derived from microeconomic theory and (2) reducedform models which mainly focus on detailed behavior descriptions in terms of data fitting, and outline the features of these models. Although the model selection may depend on what kind of phenomena we are dealing with, we generally argue that (1) Kuhn-Tucker approach might be preferable when the first priority of the analysis is maintaining the consistency with microeconomic theory, (2) reduced-form models might be preferable when we try to develop higher accurate behavioral models or behavioral models with the data potentially having observation bias, and (3) compromise discrete-continuous models between structural and reduced-form models can also be developed. We also review the empirical applications of discrete-continuous models in travel behavior research.
