In this section, we will find that applications of spatial statistics to trade and location become more realistic, not least because the methods used and the underlying dependency structures appear more like econometrics in the time series domain. While the methods discussed above are related to those for more typical social science lattice data, they are perhaps more similar to the application of time series methods in engineering or the physical sciences -- the kinds of processes economists and human geographers are involved in studying are only seldom events of the point pattern kind, or surfaces analysed in geostatistics. The understanding, however, of stationarity and isotropy that they bring with them does however carry over into studies of lattice data, including attempts to detect the spatial range at which neighbourhood effects, spillovers, make themselves felt. Two recent surveys covering the area of exploratory spatial data analysis explicitly are by Bivand (forthcoming) and D. Unwin (1996).