2017/11/01 by Claudio Fronterrè, Fronterrè, Claudio, Emanuele Giorgi +3
Economics, Econometrics and Finance · Environmental Science · #Applications (stat.AP) #FOS: Computer and information sciences #Methodology (stat.ME) #Soil Geostatistics and Mapping #Spatial and Panel Data Analysis #Sustainable Agricultural Systems Analysis
paper · pdf · doi:10.48550/arxiv.1711.00437
openalex publication_date 2017/11/01 · openalex created_date 2025/11/01 · openalex updated_date 2026/07/28
In almost any geostatistical analysis, one of the underlying, often implicit,\nmodelling assump- tions is that the spatial locations, where measurements are\ntaken, are recorded without error. In this study we develop geostatistical\ninference when this assumption is not valid. This is often the case when, for\nexample, individual address information is randomly altered to provide pri-\nvacy protection or imprecisions are induced by geocoding processes and\nmeasurement devices. Our objective is to develop a method of inference based on\nthe composite likelihood that over- comes the inherent computational limits of\nthe full likelihood method as set out in Fanshawe and Diggle (2011). Through a\nsimulation study, we then compare the performance of our proposed approach with\nan N-weighted least squares estimation procedure, based on a corrected version\nof the empirical variogram. Our results indicate that the composite-likelihood\napproach outper- forms the latter, leading to smaller root-mean-square-errors\nin the parameter estimates. Finally, we illustrate an application of our method\nto analyse data on malnutrition from a Demographic and Health Survey conducted\nin Senegal in 2011, where locations were randomly perturbed to protect the\nprivacy of respondents.\n