Sample Size And Distribution

The Indices are estimated by surveying the respondents from a panel of around 1,68,165 randomly selected households. The 522,000 respondents of the panel of households have minimum age of 15 years or above.
A stratified multi-stage survey design is deployed to select the sample of households. The Primary Sampling Units (PSUs) are the villages and towns of the 2011 Census. The Ultimate Sampling Units (USUs) are the households from these PSUs.
The broadest level of strata for sampling purpose is the Homogeneous Region (HR), which is a set of neighboring districts that have similar agro-climatic conditions, urbanisation levels and female literacy. HRs are also approximately of the same size, except in the north east, where the entire set of north-eastern states is considered as one HR.
The sub-strata are the rural and urban regions within each strata, i.e. within each HR. The sampling frame for rural PSUs is the list of villages and for urban PSUs is the list of towns as of the 2011 Census. Interviews are well spread out over the entire country and over rural and urban regions. Data are collected over a specialized mobile phone application and validated in real-time at the central office. Appropriate weights are assigned to observations for ensuring accurate population estimations.
  • Urban Strata
    Given the high variance in the size of towns, towns within an HR is stratified further based on their size. All towns of an HR is stratified into four strata based on the number of households as follows:
    Sr.No. Category Households
    1 Very Large Town More than 200,000
    2 Large Town Between 60,000 and 200,000
    3 Medium Sized Between 20,000 and 60,000
    4 Small Towns less than 20,000
    Minimum one town is randomly selected from each of these sub strata. A total of 326 towns is selected through this process. From each of these towns, 21 Census Enumeration Blocks (CEB) are randomly selected. A CEB is a cluster of about 100-125 neighboring households. Then households are selected through a process of systematic random sampling from each of the CEBs. A total of around 7,442 CEBs and 1,06,761 urban households are selected through this process. The sample size for the urban PSU (a sample town) is around 320 and the CEB within the town is fixed at 8. The larger urban sample size reflects the greater diversity in town-size and the approach of the sampling methodology to capture this diversity adequately.
  • Rural Strata
    Like the urban sample, the rural sample selection is also a two-staged sampling process. The first step is to select villages and next step is to select the households. Villages are selected through simple random sampling process. Villages are the PSUs. Households are selected from each selected village through a process of systematic random sampling. Rural India has about 97 Homogeneous Regions (HRs). The rural sample comprises around 61,404 households from 3838 villages. For the rural PSU (a sample village), the sample size is fixed at 16. The sample households from some Census Enumeration Block (CEBs) and some villages are increased following a re-balancing of the sample in response to the 2011 Census results. The sample size from some CEBs are raised to 14 or 16. And, the sample size from some villages are raised to 24 households.