Sampling
Sampling Procedure
The HSES is representative of the nation, region, urban/rural areas, residential location and amigas. It was enumerated in 320 soums and 9 districts of Mongolia. In total 11,232 households were randomly selected for the HSES 2017.
Sampling frame
The list of all households of Mongolia or the list of households registered in the population and households' database at NSO was used for HSES as a sampling frame.
Sample size
For the HSES 2017, the sample size is 11,232 households, of which 936 households were surveyed each month.
Samling design
The HSES's sampling strata or geographic domains of estimation were decided to the level of four residential location zones: Ulaanbaatar, Aimag center, Soum center, and Countryside.
Households were selected using two stages of probability sampling. In the first stage, Primary Sampling Unit (PSU) was selected, using probability proportional to estimated size, and then in the second stage, households are selected in the selected PSUs using simple random sampling.
In the HSES, Primary sampling units (PSUs) are kheseg in Ulaanbaatar, and bag in aimag, soum centers and countryside.
In Ulaanbaatar city:
- 360 PSUs were selected out of total khesegs of UB city
- 10 households were selected from each selected khesegs.
In aimag centers:
- 12 PSUs were selected from each aimag center. Taking into account the number of households, 24 PSUs from Darkhan-Uul aimag and Orkhon aimag.
- 10 households were selected from each PSUs.
In soum centers and countryside:
- 24 bags were selected from each aimag as PSUs. For Khentii, Selenge, Bayankhongor, Zawkhan, Bayarn-Ulgii and Uvs aimags, 36 bags were selected. For Uvurkhangai, Tuv and Arkhangai aimags, 48 bags were selected. For Khuvsgul and Darkhan aimags, 60 and 12 bags were selected, respectively.
- 8 households were selected from each PSUs.
Sampling distribution
The sample size is distributed using proportional probability sampling techniques. Sampling distribution is made to be proportional to each domain, taking into account the number of households of aimags.
The sampling weight applied for each sample household is equal to the inverse of its probability of selection.
In the 2017 HSES, 99.7 percent of the selected hosueholds, or 11193 households have participated in the survey.
