Spatial Thinking in Planning Practice: An Introduction to GIS

42 CHAPTER 7: GEOGRAPHIC CONSIDERATIONS IN PLANNING PRACTICE "is chapter is composed of two sections, a book chapter by O’Sullivan and Unwin on the Pitfalls and Potential of Spatial Data, and a small compiled session on geo-coding. "e book chapter identi!es major problems in the analysis of geographic information and statistical analysis of spatial data, related to spatial autocorrelation, modi!able area units, the ecological fallacy, scale, and non-uniformity of space and edge e#ects. It also discusses relevant geographic concepts central to spatial analysis such as, of distance, adjacency, interaction (!rst law of ge- ography), and neighborhood. Finally, it discusses proximity polygons and shows how variogram clouds are used to analyze relationships between data attributes and their spatial location, through matrices. Spatial data require special analytic techniques thus standard statistic methods have signi!cant problems to analysis spatial distributions. "ere are !ve major problems. First, spatial data tend to violate assumption that samples are random because in geography phenomena do not vary randomly through space, leading to the given problem of spatial autocorrelation (data from locations near to one another in space are more likely to be similar than data from locations remote from one another), which introduces the problem of redundancy due to biased samples. Second, the modi!able areal unit problem (when aggregation units used are arbitrary with respect to the phenomena under investigation) tends to a#ect the ‘coe&cient of determination, R square. "ird, the ecological fallacy (when statistical relations observed at one level of aggregation are assumed to hold because the same relationship holds when we look at more detail level). In this case the thread is that statistical relations may change at di#erent levels of aggregation. "e fourth problem is related to scale, which might a#ect spatial analysis based on the geographic scale at which the phenomenon of interest is analyzed. Lastly, another problem distinguishing spatial analysis from conventional statistics is the non-uniformity of space . "is issue refers to the fact that analysis might !nd patterns –thus clusters-, simply as a result of where people live and work. An exam- ple is the “edge e#ect” (it emerges when arti!cial boundary is imposed on a study). Although geospatial referencing provides ways to look at data. "ere are four useful concepts to analyze the spatial distribution of associated entities and spatial relationships: (i) distance (it can be measure as the simple crow’s 'ight distance between the spatial entities of interest, though it can be measured in more complex ways). (ii) Adjacency (it is of the thought as the nominal, or binary, equivalent of distance. It is argued that two spatial entities are either adjacent or they are not: there is not a middle ground). (iii) Interaction (it is considered a com- bination of distance and adjacency and rests on the ideas that nearer things are more related than distant things: !rst law of geography). And (iv) neighborhood (there are many ways to conceptualize it (e.g., with respect to sets of adjacent entities, a region of space de!ned by distance from an associated entity, etc.). One way of pulling these four concepts together is to represent them in matrices. Lastly, the chapter discusses the proximity polygons, a tool used to specify the spatial properties of a set of objects through partitioning a study region into proximity polygons. "e proximity polygon of an entity is the closest region to the entity. "e variogram cloud is an exploratory tool (though di&cult to interpret) that o#ers a general picture of relationships between the spatial locations of objects and the other data attributes. It does it by plotting the di#erences in attribute values for pairs of entities against the di#erences in their location. Read the book chapter O’Sullivan, David, and David John Unwin. Geographic information analysis. John Wiley & Sons, 2003. "e Pit- falls and Potential of Spatial Data. Chapter 2 CREATING DATA THROUGH GEOCODING Geocoding is the process used to convert location codes, such as street addresses or postal codes, into geographic (or other) coordinates. "e terms “address geocoding” and “address mapping” refer to the same process. Geoc-