Abstract

The first registers of formal use of stratification as a sample design were conceived in a time when sampling statistics was still called “the representative method.” In 1926, Arthur L. Bowley produced a report for the International Statistical Institute, entitled Measurement of the precision attained in sampling, where the first ideas of stratification with proportional allocation was deemed to be laid out. Later, in 1934, Neymann published On the Two Different Aspects of the Representative Method: The Method of Stratified Sampling and the Method of Purposive Selection, consolidating the fundamental theory behind stratified random sampling designs. Improving samples based on stratified designs depends on the availability of useful auxiliary data. In this presentation, some of the roles that remote sensing techniques can play in the digital era to support the stratification of geographic areas will be explored, considering their use in agricultural and environmental surveys.