The Relative Performance of Various Estimators of Recreation Participation Equations

Publisher Website
Google Scholar
Abstract
This paper evaluates the predictive performance of four estimators (ordinary least squares, generalized least squares, logit, and probit) for recreation participation equations using subsamples of the 1972 National Recreation Survey. The objective of the paper is to offer some evidence on the performance of estimating techniques which are argued to be preferable, on theoretical grounds, to ordinary least squares with the examples of dichotomous dependent variables model drawn from the outdoor recreation literature. While our results are specific to the samples and models considered, they are consistent with evidence from several other applications and indicate that there is very little difference between these four methods in their ability to identify key determinants of behavior and in their predictive performance.

This publication has 1 reference indexed in Scilit: