1. See, Hordijk and Paelinck 1978 175–9 Paelinck and Klaassen (1979 vii, pp. 5–11), Paelinck ( 1982, pp. 2–3 ).
2. For example, a recent book on Urban Econometrics,by Kau, Lee and Sirmans (1986) deals with the econometric analysis of urban economic models, but largely ignores spatial effects which may affect this analysis.
3. It is also very relevant in the physical sciences, such as plant ecology, geology, epidemiology, etc. However, since these fields are not central to applied regional science, they will not be further considered. Interested readers are referred to Cliff and Ord (1981), Ripley (1981), Gaile and Wilmott (1984), and Upton and Fingleton (1985) for extensive references.
4. In the literature the terms spatial structure, spatial pattern and spatial process are often ill—defined and used interchangeably. Here, I follow the distinction made by Haining (1986a, p. 59–60) between spatial flow, spatial pattern and spatial structure. Spatial flow is taken to be the “physical transfer of commodities, people, information,” etc. Spatial structure refers to the “background geography” and is rather fixed. Spatial pattern “relates to more volatile or changeable levels of spatial regularity that may be imposed on the more permanent structure.” Spatial processes are phenomena which relate the three elements. See also Bennett, Haining and Wilson (1985), for an overview.
5. For a recent overview, see Hsiao (1986). An exception here is the work by Scott and Holt (1982), and King and Evans (1985, 1986) on cross—sectional dependencies resulting from block structures in survey data. However, a spatial interpretation of these dependencies is not given.