Table 3: First level of sub–strata for quantitative research Type of sub-strata Population size of 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 – 10 Smaller townships Less than 0.25 million (< 0.25 million) Various stages of sampling Stage 1: Townships Townships were the primary sampling units (PSUs). A list of townships with the respective number of households (Source: Results of Population and Housing Census of Myanmar 2014, Dept of Population (pp 21-36)) was obtained and functioned as the sampling frame for the PSU. The sample townships were selected with probability proportionate to size with replacement (PPSWR) to enable simpler calculation of sampling weights. Stage 2: Wards in urban and village tracts in rural A list of wards/village tracts with respective number of households in the sample township was the sampling frame for this stage. In most sample cities/townships, 2 sample wards from urban areas and 3 village tracts from rural areas were selected with PPSWR. Stage 3: Segments in wards and village tracts A segment was defined as a street/block7 in urban areas and a village in rural areas The sampling frame for this third stage was a list of segments with the respective number of households within sample ward/village tract. This was constructed in the field for selection of third stage unit. Two clusters (segments) per sample ward and three per village tract was selected with PPS Systematic Sampling 7 Note: As per Myanmar’s administrative system ICTs in Myanmar: 2016 93