Table 6. A comparison of simulation results for single-stage cluster sampling.

(1) 1-Stage Cluster Method-1000 trials	(2) Average residences sampled per simulation trial	(3) Mean value of H-T estimator for 1000 trials	(4) Variance of H-T estimator for 1000 trials	(5) Standard Error of the Mean (SEM)	(6) Number of sections sampled
(A) Unstratified	396	26,270	84,413,782	256	4
(B) Stratify by number of buildings per section	443	25,935	16,983,502	115	4
(C) Stratify by number of persons per section	506	25,909	5,503,716	65	4

Comparison of the standard deviation σ, the variance σ ², and the SEM (Standard Error of the Mean) for the single-stage sampling protocol. The uncertainty of the Horvitz-Thompson population estimate decreases as a function of the protocol used to partition the 20 sections of Bo City into 4 strata: (A) unstratified single-stage sampling (B) stratification by number of buildings per section (C) stratification by the total number of persons per section. There are 1,979 residential structures in the 20 sections, and a measured population of 25,954 persons.