How do we conceptualise geographical linkages in statistical analyses? Identifying explicit and implicit understandings of spatial weights in planning

点击翻译按钮获取中文摘要

The submitted proposal outlines a project to gain a comprehensive conceptual understanding of spatial weights. Spatial weights are used in planning and related spatial disciplines to formalise and operationalise spatial associations in statistical analyses. As such, they play an important role in methodological procedures of evidence-based planning and spatial observation and have a strong influence on expected analytical results. The latter eventually feed into planning processes, demonstrating the importance of a solid conceptual understanding of spatial weights. While many technical properties of spatial weights are well understood, a comprehensive understanding of the researchers' intended meanings of formalised spatial relationships is still lacking. Do weights represent the potential for spatial interaction? Are they models for spatial proximity? Do they reflect the propensity of similarities in space given passively by external circumstances? These levels of meaning are not well understood to date and often remain implicit in published research practice. Such understanding is essential to comprehensively apprehend analysis results and to feed them into planning processes in a meaningful and informed way. This is where the proposed project comes in. The proposed project pursues three goals: (1) to identify co-existing conceptual understandings of spatial weights, (2) to understand how researchers create and use spatial weights, and (3) to gain a comprehensive picture of the historical lineages of conceptual understandings of spatial weights. Objective 1 is pursued by extracting published understandings of spatial weights from scholarly articles through a systematic review combined with a qualitative analysis strategy. To achieve objective 2, selected authors will be interviewed in person. The transcribed interviews will be qualitatively analysed using narrative analysis to better understand the underlying decision-making process for selecting spatial weights. Regarding objective 3, a quantitative and qualitative analysis of the citation network of the literature corpus identified in the review will be conducted, including exploiting the temporal dimension. All work packages and objectives are thus interlinked. The expected results have the potential to alleviate methodological specification problems, make spatial weights easier to understand and more defensible in empirical applications, and contribute to a better understanding of how we as a planning community represent space in our quantitative methods.