The hypothesis that destination choice by a heterogeneous population group may appear to be a random process because of the conflicting and interacting effects of many variables on the choice decision is examined. A formal model of random destination is specified, and its predictions are tested by using data for 3 destinations, 8 classes of destination and 10 population groups. The findings support the development of Markov, linear learning, multinominal logit, and multivariate models of destination choice.
Abstract