The chapter introduces the idea that the relationships between natural conditions and the outcome of an observation may be deterministic, random, strategic or chaotic, and that numerical ecology addresses the second type of data; it describes the role of numerical ecology among the various phases of an ecological research. The chapter includes discussion of the following topics: spatial structure, spatial dependence, and spatial correlation (independent observations, independent descriptors, linear independence, independent variable of a model, independent samples, origin of spatial structures, tests of significance in the presence of spatial correlation, and classical sampling and spatial structure), statistical testing by permutation (classical tests of significance, permutation tests, alternative types of permutation tests), computer programs and packages, ecological descriptors (i.e. variables: mathematical types of descriptors, and intensive, extensive, additive, and non-additive descriptors), descriptor coding (linear transformation, nonlinear transformations, combining descriptors, ranging and standardization, implicit transformation in association coefficients, normalization, dummy variable coding, and treatment of missing data (delete rows or columns, accommodate algorithms to missing data, estimate missing values). The chapter ends on a description of relevant software implemented in the R language.