A Further Discussion of Optimal Comparable Selection and Weighting, and a Response to Green

Using a revised model framework that views expected adjusted prices of corn-parables as random variables, Green (1994) demonstrates that Vandell's (1992) minimum variance estimator is preferred under the classical ordinary least squares (OLS) assumptions. As a result, the minimum coefficient of variation estimator proposed by Gau, et al. (1993) is preferred only when the classical OLS assumptions are relaxed. We demonstrate that, even under Green's revised framework, there is not sufficient evidence in the paper to justify this claim. Copyright American Real Estate and Urban Economics Association.