The Use of Random Geographic Cluster Sampling to Survey Pastoralists

Himelein, Kristen ; Eckman, Stephanie ; Murray, Siobhan

World Bank, Washington, DC

关键词: ACTIVE VOLCANOES;    AGRICULTURAL DEVELOPMENT;    AGRICULTURAL ECONOMICS;    AGRICULTURAL PRODUCTION;    AGRICULTURE;   

DOI  :  10.1596/1813-9450-6589
RP-ID  :  WPS6589

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

Livestock are an important component ofrural livelihoods in developing countries, but data aboutthis source of income and wealth are difficult to collectbecause of the nomadic and semi-nomadic nature of manypastoralist populations. Most household surveys excludethose without permanent dwellings, leading to undercoverage.This study explores the use of a random geographic clustersample as an alternative to the household-based sample. Inthis design, points are randomly selected and all eligiblerespondents found inside circles drawn around the selectedpoints are interviewed. This approach should eliminateundercoverage of mobile populations. The results of a randomgeographic cluster sample survey are presented with a totalsample size of 784 households to measure livestock ownershipin the Afar region of Ethiopia in 2012. The paper exploresthe data quality of the random geographic cluster samplerelative to a recent household survey and discusses theimplementation challenges.