POPULATION SEGMENTATION IN URBAN RECREATION CHOICES.

The paper describes an investigation of various segmentation bases for capturing the behavioral differences in urban recreation demand. The analysis and evaluation of the segmentation bases were achieved through the calibration of disaggregate quantal choice models (by using the multinomial logit technique) for each population segment and statistical comparison of these models and their estimated coefficients. The segment models were then compared with the pooled model both in terms of the overall goodness of fit and in terms of the differences in their coefficient estimates.