Average vs. Individual: A New Perspective on Geographic Analysis

Keywords: Third Law of Geography, Geographic Similarity, Geographic Analysis, Averaging, Individual representation.
Abstract Type: Paper Abstract

Abstract

Current methods of geographic analysis are largely based on the concept of statistical averaging, which often creates the general conditions from an entire sample set. While this approach does offer a way to capture the general trend and patterns captured in the sample set, the uniqueness of geographic conditions over space as captured by individual samples are often smoothed out by this “averaging”. This paper presents a new thinking on geographic analysis based on the Third Law of Geography. Under the Third Law of Geography, geographic analysis can be conducted on the basis of geographic similarity between a sample and a point of interest. This allows the unique representation of a single sample to be used in geographic assessment, instead of the average conditions from an entire sample set. Research examples from spatial prediction have shown that geographic analysis based on the Third Law does not require samples to be of specific size nor a particular spatial distribution to achieve a high-quality assessment. The uncertainty associated with each prediction based on the Third Law is more indicative to the quality of the assessment, thus showing the spatial distribution of assessment accuracy, which can be used to allocate additional samples more effectively in further improving the quality of the assessed products. These properties make geographic analysis based on the Third Law of Geography more adapted in making effective use of volunteered geographical information which is becoming increasingly available in the spatial big data era.