1
عمومی::
مجری
Four approaches to implementation have been tried: (1) using gridded PSU boundaries produced by GridSample, (2) manually segmenting gridded PSU using satellite imagery, (3) non-probability sampling (e.g. random-walk, "spin-the-pen"), and random sampling of households.
The UN's Multiple Indicator Cluster Surveys (MICS), the US Government's Demographic and Health Surveys (DHS), and the World Bank's Living Standard Measurement Surveys (LSMS) stratify samples by sub-national region, and sample roughly 10,000 households in a two-stage design that is widely used by survey implementers to maximize statistical power and feasibility while minimizing costs and potential biases [14, 15, 16].
In standard large-scale household surveys, implementers sample communities first (called clusters, or primary sampling units-PSUs) from recent census enumeration areas.
While the GridSample output shapefile includes values needed to calculate PSU selection probabilities, the survey implementer must track the number of households enumerated in the field in each PSU, and household response rates to correctly calculate sample weights.
However, we strongly recommend that implementers employ a method to anonymize buildings added to the crowd-sourced map such that interviewed PSUs cannot be identified.
واژگان شبکه مترجمین ایران