Survey ID Number
LAO_2017_MICS-DHS_v01_M
Title
Social Indicator Survey II 2017
Sampling Procedure
The major features of the sample design are described in this appendix. Sample design features include defining the sampling frame, target sample size, sample allocation, listing in sample clusters, choice of domains, sampling stages, stratification, and the calculation of sample weights.
The primary objective of the sample design for the 2017 Lao Social Indicator Survey (LSIS 2017) was to produce statistically reliable estimates of most indicators, at the national level, for urban and rural areas, and for the 18 provinces of the country.
A multi-stage, stratified cluster sampling approach was used for the selection of the survey sample. The primary sampling units (PSUs) selected at the first stage were villages (PSU and Village are used interchangeably in this Chapter). A listing of households was conducted in each sample village, and a sample of households was selected at the second stage.
SAMPLING FRAME AND STRATIFICATION
The sampling frame for this survey consisted of a list of all villages in the country, arranged by province, with appropriate size estimates (number of households) and other relevant information about each village. The village register is maintained by Lao Statistics Bureau (LSB). It is updated in December each year. The version used as sampling frame was the village register of December 2015.
The 18 provinces were defined as the sampling strata. Within provinces a further, implicit, stratification - on village category - was achieved by systematic sampling from a list of villages ordered by village category.
SAMPLE SIZE AND SAMPLE ALLOCATION
The overall sample size for the 2017 Lao Social Indicator Survey was calculated as 23,400 households. For the calculation of the sample size, the key indicator used was the underweight prevalence among children age 0-4 years. Since the survey results are tabulated at the provincial level, it was necessary to determine the minimum sample size for each province.
The number of households selected per cluster for the survey was determined as 20 households, based on a number of considerations, including the design effect, the budget available, and the time that would be needed per team to complete one cluster. Dividing the total number of households by the number of sample households per cluster, it was calculated that 1,170 sample clusters would need to be selected for the survey.
The sample allocation over provinces was determined by a procedure where the sample at first was allocated proportionally to the square root of the number of households in each province. This allocation was further adjusted so that provinces getting less than 1,100 households in the preliminary allocation were given additional households up to 1,100. These additional households were taken from the three provinces that had the largest samples according to the preliminary allocation. The sample sizes for provinces vary between 1,100 and 1,680 households. The justification for using different sample sizes is that the standard errors for national estimates will be lower than the standard errors that would have been achieved with equal sample sizes over the provinces.
Within province the sample was allocated over implicit strata defined by village category. This was achieved by systematic sampling from a list of villages ordered by village category. This way of sampling resulted in approximately proportional allocation of the province sample over the implicit strata urban villages, rural villages with road and rural villages without road.
SELECTION OF VILLAGES (CLUSTERS)
Villages were selected from each of the sampling strata (provinces) by using systematic probability proportional to size (PPS) sampling procedures. The measure of size was the number of households in the village; the number was obtained from the LBS village register. Altogether 32 villages were so large in size so they had the probability equal to one to be selected to the sample. These large villages were thus selected to the sample with certainty.
LISTING ACTIVITIES
A new listing of households was conducted in all the sample villages prior to the selection of households. For this purpose, listing teams were trained to visit all the sampled villages and list all households in the village. The listing operation took place from December 2016 to February 2017 with 70 listing team members. In each Province, there were two teams each consisting of a lister and a mapper, except in Champasack, where three teams were assigned.
Listing could not be done in four villages. In two of the villages the area had been completely cleared of dwellings due to preparations for dam construction. One village was not accessible by car or motorcycle due to poor roads and one village could not be properly identified due to village mergers.
Large villages, where the number of households exceeded 300 households, were divided into two or more segments, and one segment was picked randomly before listing. Segmentation was done in 216 villages.
SELECTION OF HOUSEHOLDS
Lists of households were prepared by the listing teams in the field for each village. The households were then sequentially numbered from 1 to Mhi (the total number of households in each village or segment) at the Lao Bureau of Statistics, where the selection of 20 households in each village was carried out using random systematic selection procedures. The MICS6 spreadsheet template for systematic random selection of households was adapted for this purpose.
The survey also included a questionnaire for individual men that was to be administered in half of the sample of households. The MICS household selection template includes an option to specify the proportion of households to be selected for administering the individual questionnaire for men, and the spreadsheet automatically selected the corresponding subsample of households. All men age 15 to 49 years in the selected households were eligible for interview.
LSIS 2017 also included water quality testing for a subsample of households within each sample cluster. A subsample of 3 of the 20 selected households was selected in each sample cluster using random systematic sampling for conducting water quality testing, for both water in the household and at the source. The MICS household selection template includes an option to specify the number of households to be selected for the water quality testing, and the spreadsheet automatically selected the corresponding subsample of households.