Substantive Issues in Spatial Modeling
Spatial modeling can be a very powerful tool for understanding the way space impacts behavior and policy, and can be applied to a number of types of data and situations.
Example: Spatial Models: Limitations, Issues, and Emerging Developments
Like all statistical modeling approaches, spatial modeling has limitations. The usual set of assumptions that apply to standard linear modeling also apply here, and can be especially problematic. For example the assumption that the residual variances are randomly distributed is particularly problematic if there are large inter spatial unit variation in the observations. In pooled models, this is dealt with via the Random Eﬀects versus LSDV choice, but in the cross sectional case the assumption is that whatever heterogeneity there is has been explained by the inclusion of the right independent variables-another version of the omitted variable problem discussed above. If the heterogeneity is due to spatial eﬀects, in can be modeled; if not, it may undermine the stability of the results.