Significance Assessment in the Application of Spatial Analytics

The breadth of spatial analytics relied on in geography is remarkable. These methods are used to gain insight, typically in the context of planning, management, decision making, and policymaking. The challenge is invariably establishing meaning in obtained findings, providing insights and knowledge. This is done, however, in very different ways depending on the spatial analytic approach, with statistical notions of significance a prevailing tool in assessment. This article reviews significance assessment approaches in the application of spatial analytics. Noteworthy in this review is that many approaches can be considered through the lens of sampling. In some cases, the underlying sample is a few or as small as one. There are methods, however, that are based on sampling that is comprehensive, involving an implicit or explicit enumeration of all possible outcomes. This suggests that significance assessment using certain methods accounts for the entire range of possibilities, whereas other methods draw inference from scant sampling. In addition to the sampling perspective, extrapolation as well as indirect accuracy and anecdotal measures of assessment are not uncommon. The comparative review of spatial analytic methods suggests that significance is assessed in many different ways, making meaning interpretation challenging.

Automated summary

This article explores the diversity of significance assessment methods in geographic spatial analysis, including considering different sampling perspectives. Some methods are based on comprehensive sampling, while others rely on fewer samples, making interpreting meaning challenging.