Online Program

Evaluating Application of Adaptive Sampling Approaches to Survey Informal Firms: Simulation Results Based on (Real) Population Data (308025)

Keywords: Adaptive sampling, Informal enterprises

Utilizing a unique dataset from eSwatini (Swaziland) that provides a full census of all businesses operating in the country, including informal businesses, to conduct a series of simulations to evaluate efficiency gains from using Adaptive Sampling approaches compared to traditional SRS. This dataset is unique in that it includes GPS information, and so provides a real-life demonstration of the clustering of informal-sector businesses, which is used to estimate the effects of various sampling parameters. The implication of changing expansion thresholds and PSU (defined square areas in urban areas) is presented, as are the implication of using stratification. Efficiency of various approaches is discussed with respect to fieldwork activities including level of effort and sample control. Simulations also reveal useful insights to differences in statistical efficiency of these approaches with respect to this clustered population. Synthetic population data is used to test these findings on populations with different characteristics.