A Selection Bias Approach to Destination Choice

摘要: Research on consumer search behavior commonly envisages destination choice as a two-step process: (1) delineate the set, and (2) evaluate choices therein. However, much of empirical work in choice—including logit nested formulations—models only latter, not set delineation itself. In presence correlation between error terms selection, statistical estimators are biased, problem that Heckman others have called selection bias. this paper, an alternative two-stage method is proposed to estimate parameters models selection. Monte Carlo simulation used explore properties these estimators, show magnitude bias inherent traditional methods estimation.