ABSTRACT
In traditional univariate modeling approaches, a single dependent variable is modeled as a function of independent variables. However, there are instances where two or more dependent variables depend on each other or share commonly shared unobserved characteristics. If this is the case, bivariate and multivariate dependent variable models are appropriate. This chapter presents a number of these models including the bivariate ordered probit, the bivariate binary probit, the multivariate binary probit, and simultaneous estimation of discrete and continuous variables. Extensive discussion is presented relating to approaches used to account for cross-equation correlation in bivariate and multivariate models, and numerous examples are provided to support the discussions relating to the models. The examples and discussions in this chapter show that bivariate and multivariate dependent variable models have many applications in the analysis of transportation data and beyond.
