Errata for Anselin, Bongiovanni, and Lowenberg‐DeBoer, AJAE, 86(2004), p. 675–687
本文是对原论文的勘误，修正了模型诊断部分的表述错误，并将表3中的未加权总和改为按景观区面积加权的平均值，以正确反映物理和经济结果。
p. 682 – The second sentence in the section entitled “Coefficient Estimates and Model Diagnostics,” should be: “Because the model was estimated with dummy variables for all landscape areas and a restriction that the dummy variable coefficients sum to zero, the estimated intercept and nitrogen coefficients are mean responses and the estimates for the dummy variable coefficients are interpreted as the difference from the mean response.” p. 684 ‐ The published values in Table 3 are unweighted sums; they should be weighted averages, with weights being the area of the field in by landscape zone (e.g. Low East, Slope East, Hilltop, Slope West). This does not affect the text, but it does make it difficult to understand how the conclusions in the text were derived. The revised Table 3 below has the weighted averages. Physical and Economic Results by Model and Estimation Technique Neighborhood Criteria N Rate for Max. Yield (kg ha −1 ) Maximum Yield (kg ha −1 ) Optimal N Rate (kg ha −1 ) Optimal Yield (kg ha −1 ) Net Returns ($ ha −1 ) Breakeven VRT Fee ($ ha −1 ) Queen criterion (8) OLS 176.35 6,921.74 69.87 6,558.49 410.60 2.40 OLS w/interactions 165.08 6,885.24 63.52 6,538.73 412.22 3.28 SAR (GM‐iterated) 315.41 7,350.51 74.55 6,528.78 406.38 6.58 SAR (ML) 441.82 7,881.83 74.75 6,639.50 413.87 7.57 SAR (GM‐GHET) 319.78 7,376.27 74.04 6,537.89 407.24 6.58 SAR (GM‐GHET) 478.54 8,042.88 74.85 6,685.85 417.00 7.67 Rook criterion (4) OLS 176.35 6,921.74 69.87 6,558.49 410.60 2.40 OLS w/interactions 165.08 6,885.24 63.52 6,538.73 412.22 3.28 SAR (GM‐iterated) 319.47 7,429.79 80.61 6,614.89 409.44 6.72 SAR (ML) 362.93 7,609.61 75.35 6,633.90 413.20 6.93 SAR (GM‐GHET) 361.61 7,559.82 78.21 6,592.97 409.06 6.96 SAR (ML‐GHET) 392.64 7,717.31 75.80 6,653.79 414.36 7.11