Survey on Childcare Needs, Time Use and Child Well-being in Indonesia, 2022

The survey adopted a stratified four-stage probability sampling design to produce nationally representative estimates of childcare needs, time use, and children’s wellbeing, with planned disaggregation by region (Java vs. Outside Java) and by urban–rural residence. Sampling was conducted using probability-based methods at every stage. Districts were selected using probability proportional to estimated size (PPES), villages and hamlets/RW/dusun were selected using probability proportional to size (PPS), and eligible households were selected through systematic random sampling within sampled hamlets.

At the first stage, districts and cities served as primary sampling units and were selected using probability proportional to estimated size (PPES). Districts were stratified by region (Java and Outside Java) and administrative type (regency [kabupaten] vs. city [kota]), and their selection probabilities were based on an estimated measure of size reflecting the number of eligible households -- defined as households with at least one child aged 0–6 years and at least one woman aged 15 years or older. This measure of size was constructed using auxiliary data to account for variation in both child presence and adult female population across districts. A total of 40 districts were selected nationwide, and these districts are shown in the sampling document provided under Downloads.

At the second stage, five villages were selected within each sampled district using probability proportional to size (PPS), where size was defined as the total number of households (taken from PODES). Villages were stratified by urban and rural classification prior to selection to ensure balanced representation across settlement types.

At the third stage, one hamlet/RW/dusun was selected within each sampled village using PPS based on the number of households, with hamlet-level household counts obtained from village officials during field listing.

At the fourth stage, 15 eligible households were selected within each sampled hamlet using systematic random sampling from the eligible-household list compiled in the field. In each selected household, one eligible child aged 0–6 years was selected using a Kish method. Before applying the Kish table, eligible children were listed by sex and age: female children first, followed by male children, with each group ordered from oldest to youngest.

For more details on the sampling process, users can consult the accompanying sampling documentation.