Files

Abstract

Revealed preference methods require survey data on past resource use, and numerous studies have found reported recreation frequency to be overestimated and concentrated on prototype (rounded and calendar-based) values. This paper develops an approach to treat extreme values and rounded responses in survey datasets and thereby improve model fit and resulting welfare estimates. We illustrate how, when modeling single-site trip data, model fit can be improved by transitioning from a discrete to a continuous distribution at a cut-point where response behavior begins to exhibit rounding. We feel this method will be useful for recreation demand research and may have broad applicability to the general analysis of count data.

Details

PDF

Statistics

from
to
Export
Download Full History