Sampling
Sampling Procedure
Sampling frame
The TUS is a household based survey. Therefore, households are defined as sampling unit. The total sample size determined as 4,000 households or 0.5 percent of total households, representing the entire households in Mongolia. Since the survey covered 4 quarters, sample size was divided into four equal parts for four quarters.
Excluded are the Mongolian citizens residing outside the country or in military camps, orphanages, care centers, hospitals, prisons and other correctional institutions for more than six months.
All 8 provinces and 8 districts from Ulaanbaatar representing the five regions of Western, Khangai, Central, Eastern and Ulaanbaatar, that was surveyed.
Sampling method
The sampling design used a two-stage sampling method.
1. Stage 1: The proportionate number of households for each of the 5 regions of Mongolia was determined using probability sampling method. Following this method, it was established that there should be 640 HHs from the Western region, 800 HHs from Khangai region, 800 HHs from Central region, 560 HHs from Eastern region, and 1200 HHs from Ulaanbaatar. The lowest administrative units, kheseg of khoroo for Ulaanbaatar, bagh of soum for aimags (provinces), were defined as primary sampling units (PSUs).
In total, 200 PSUs were selected with probability proportional to size.
2. Stage 2: After household listing of the selected PSUs was carried out by the statistical departments or divisions, 20 households were selected using systematic sampling from each PSU.
Sample household
Primary data collection unit of the survey is the household because the TUS is a household based survey. 13626 people from 3990 households' selected 14173 people from 4000 households were participated in the study.
Age over 12 years, 6914 people from selected 10091 people was fully participated in data collection of the survey and was answered in individual questionnaire.
