摘要：The implications of ignoring potential spatial dependence in county-level yield data are discussed. Spatial dependence in a county-level yield data set is identified and methods for correcting the dependence via spatial weighting matrices and generalized least squares regression are performed. The paper also examines how the spatial dependence declines as the distance between observations increases.
