Survey ID Number
MNG_2003_RHS_v01_M
Title
Reproductive and Health Survey 2003
Sampling Procedure
The survey was conducted using a two-stage sampling method, which gives an equal probability of selection of households. This means that the data are fully comparable with the RHS 1998. The sample frame comprised the listings of households prepared annually in bags and khoroo across the country. Activities designed for improving the quality of data of the sample frame were conducted in the fourth quarter of 2002 and the first quarter of 2003. The actual sampling was based on the 2003 first half-year data.
Distribution of the RHS Household Sampling by Aimag, Mongolia 2003
-- Aimag, Clusters, Number of Households
1) Arhangai, 1-13, 390
2) Bayan-Olgii, 14-23, 300
3) Bayanhongor, 24-34, 330
4) Bulgan, 35-42, 240
5) Gobi-Altai, 43-50, 240
6) Dornogobi, 51-56, 180
7) Dornod, 57-64, 240
8) Dundgobi, 65-70, 180
9) Zavhan, 71-80, 300
10) Ovorhangai, 81-94, 420
11) Omnogobi, 95-100, 180
12) Suhbaatar, 101-107, 210
13) Selenge, 108-118, 330
14) Tov, 119-129, 330
15) Uvs, 130-139, 300
16) Hovd, 140-149, 300
17) Hovsgol, 150-164, 450
18) Hentii, 165-173, 270
19) Darhan-Uul, 174-182, 270
20) Ulaanbaatar, 183-269, 2610
21) Orhon, 270-279, 300
22) Gobisumber, 280,30
Total number of households 8400
It was determined from the experience of RHS 1998, and of other countries which have conducted similar surveys, that 25-30 households per cluster would provide an optimum representation. Therefore, this time 30 households were selected for a cluster (The best cluster "take" depends upon the intra-cluster versus inter-cluster heterogeneity of the principal variables being measured; this can only be determined after carrying out a survey). For the survey it was planned to select 8,400 households, which is a 1.47 percent sample of all households in the country. This implied the selection of 280 clusters of households. Baghs and horoos were the primary sampling units (PSUs). All 1,674 PSUs were stratified implicitly by aimag and soum, and the selection of the 280 sample PSUs (or clusters) was done systematically with a random start, with probability proportional to the number of registered households. Households were then selected systematically with a random start within each PSU, using an interval directly proportional to the number of households in the PSU. The selected households were interviewed using the household schedule. All women between the ages of 15 and 49, inclusive, who slept in the household's dwelling the night prior to interview, were eligible to be interviewed using the women's interview schedule. Three husbands out of five married women interviewed in each PSU were interviewed using the husband's interview schedule.