Sample age group Sample size Margin of error 15-65 15-69 (15-65) 8400 7500 (7,204 for analysis) +/-2.5% +/-3% Despite the differences in sample size and margins and margins of errors, the results from the two nationally representative surveys are largely comparable. The 2016 follow up survey was covered the population aged 1569, but only the data from the population aged 15-65 was analyzed for the sake of comparability with the 2015 baseline survey. . The results can be disaggregated by 15 states and regions including Nay Pyi Taw territory, 6 geographic regions, and urban vs rural location as well as by gender and age groups as appropriate. The townships are set to be Primary Sampling Units (PSU). 32 of the 330 townships (from Kachin State, Kayah State, Kayin State, Chin State, Sagaing Region, Rakhine State and Yangon Region) are excluded from the PSU sampling frame due to inaccessibility and security concerns. The data is stratified as follows: -Main strata: 19 strata of 15 states and regions including Nay Pyi Taw Union territory which can aggregated to form 6 geographic areas – Northern hills, Eastern hills, Middle Dry Zone, Lower Valley, Ayeyarwady Delta and Long Coast. -First level Sub strata: This level of stratification depended on the population size in the cities or townships. Table 5: First level of sub-strata for quantative research Type of sub strata Population size in the cities or townships Big cities Greater or equal to 1 million (> =1 million) Other major cities Between 0.25 million and 1 million (0.25 – 1 million) Smaller townships Less than 0.25 million (< 0.25 million) -Second level Sub strata: Urban & Rural within selected sample cities/ townships for disaggregating the estimates between urban and rural. Samples are allocated two fifths and three fifths between urban and rural as appropriate respectively. Four stage sampling was carried out. The stages, briefly, are; Stage 1: Selection of township Stage 2: Selection of wards in urban areas and village tracts in rural areas in selected townships Stage 3: Selection of clusters in specified wards/village tracts Stage 4: Selection of households in selected cluster (segment) Once a household was selected, a Kish grid was used to select one respondent for the individual specific component of the questionnaire.