2. Selection probabilities and sampling weights The survey aimed to represent households and population (aged 15 – 65 years) in all accessible areas of Myanmar, both urban and rural. Hence, two sets of selection probabilities were used – one for households and others for individuals aged 15 – 65 years old. These two sets of sampling probabilities yielded the sampling weights for households and individual aged 15 – 65 years old. 2.1 Selection probabilities for different sampling stages First Stage Unit (FSU or PSU), Township The sample township selection probability can be obtained as follows given the selection of townships using probability proportionate to size with replacement (PPSWR), No. of sample townships * Total households in sample township Probabilty (FSU) = Total household in entire strata Second Stage Unit (SSU), Ward or Village Tract Sample wards and village tracts were drawn similarly, by probability proportionate to size with replacement (PPSWR). Selection probability of SSU can be obtained as follows. No. of sample wards OR village tracks * Total households in sample ward OR village track Probabilty (SSU) = Total household in township urban (forwards) OR township rural (for village tracts) 98 ICTs in Myanmar: 2016