Robustness of Spatial Autocorrelation Specifications: Some Monte Carlo Evidence
通过蒙特卡洛实验，让不同空间自相关模型轮流生成数据，再用所有模型估计这些数据，主要依据预测能力评估各模型的稳健性。
Abstract This paper examines the robustness of various models of spatial autocorrelation through a series of Monte Carlo experiments in which each model takes a turn at the data generator. The generated data are then used to estimate all of the models. The estimated models are evaluated primarily on their predictive power.