Sampling design and analysis using geo-referenced data

This chapter focuses on geo-referenced administrative data, such as, in particular, population and business registers in Italy. It describes the attention on problems related to completeness and locational accuracy of Italian geo-referenced registers and on consequences in sampling methods applied on them. The chapter presents the traditional problems in geo-coding procedures and the deriving errors in registers. It discusses a detailed explanation of spatial sampling methods for geo-referenced registers. The chapter illustrates several possible applications of spatial sampling in presence of coverage errors, as a consequence of failed completeness of archives, and sampling by imprecise data, as a consequence of locational errors. The idea behind sampling designs is to exclude contiguous units to be selected in the same sample. Spatially Correlated Poisson Sampling is a technique that selects spatially balanced samples, such as with locations of units that are well distributed over the population