Authors: Ricardo Truffello Robledo*, Pontificia Universidad Católica De Chile - Instituto de Estudios Urbanos, Monica Flores, Pontificia Universidad Católica de Chile - Observatorio de Ciudades UC, Matías Garreton, Universidad Adolfo Ibáñez - Desing Lab, Gonzalo Ruz, Universidad Adolfo Ibáñez - Escuela de Ingeniería
Topics: Geographic Information Science and Systems, Spatial Analysis & Modeling, Urban Geography
Keywords: Rationalization Algorithm, Spatial Sampling, Constrained clustering
Session Type: Paper
Start / End Time: 8:00 AM / 9:40 AM
Room: Council Room, Omni, West
Presentation File: No File Uploaded
Increasing omissions in recent versions of the Chilean Census coupled with the lack of sufficient spatial representativeness have negatively impacted the design and application of public policies that aim to tackle urban territories from a local scale. Moreover, oftentimes surveys concentrate in few areas while leaving entire neighborhoods unattended; a problem partially rooted in the belief that sampling should be absolutely aleatory, without paying attention to the phenomena identified by Henri Lefevbre as “the production of space” (Lefebvre, 1974) or in quantitative approach spatial autocorrelation paradigm (Anselin, 1995) . The convergence of these three difficult aspects has shaped the last years of spatial statistics applied research in Chile, pushing us to think and design spatial sampling models that are cost-effective while spatially distributed to represent the majority of the population in cities. In this presentation, I explore the theoretical aspects underpinning “homogeneous zoning,” a valid tool for identifying spatial autocorrelation and improving cost-effective sampling methodologies. The methodology proposed here seeks a spatial sampling consistent with the Chilean population Census, however, it can work with other data sets. Secondarily, this research explores the suitability of specific variables for generating statistical stratification in cities with social economical variables. Finally, I discuss the ways in which this methodology can be attuned with coherent sampling areas in a cartographic and demographic sense (size and number of people, respectively) in an optimization function with meta-heuristic approach (Duque, Anselin, & Rey, 2012) for use of INE (National Statistic Institute) like minimal sampling units .