ABSTRACT
In this chapter, the authors describe the concepts and methods to reflect the fact that where and when they observe a phenomenon of interest, both matters for the analysis of the data. They argue that selection between competing models should be guided by the principle of parsimony whereby preference among models with the same explanatory power should be given to those with the least number of parameters. The authors focus on prevalence data from repeated cross-sectional surveys as these are one of the most commonly used study designs for disease surveillance, especially in low-resource settings where disease registries are non-existent or geographically incomplete. They also describe the different stages of a spatio-temporal geostatistical analysis and provide tools that directly address the issue of specifying a spatio-temporal covariance structure that is compatible with the data. The authors present a general framework for the analysis of data from spatio-temporally referenced prevalence surveys.
