India - Housing Condition Survey, July - December 2002, NSS 58th Round

Hamlet-groups / sub-blocks were formed by more or less equalising population. For large urban blocks, the sub-block (sb) having slum dwellers, if any, was selected with probability 1 and was termed as segment 1. However, if there were more than one sb having slum dwellers, the sb having maximum number of slum dwellers was selected as segment 1. After selection of sb for segment 1, one more sb was selected by simple random sampling (SRS) from the remaining sb's of the block and was termed as segment 2. For large blocks (having no slum areas) two sub-blocks were selected by simple random sampling without replacement (SRSWOR) and were combined to form segment 2. For urban blocks without sub-block formation, segment number was 1 or 2 depending on whether the block was having a slum or not. For large villages two hamlet-groups were selected by SRSWOR and were combined to form segment 2. For villages without hamlet-group formation, segment number was also 2. The segments were considered separately for listing and selection of the ultimate-stage units.