FAQ: How did we design the YAFS5 survey?
The Young Adult Fertility and Sexuality Study (YAFS) is the biggest series of surveys on Filipino youth. In this year’s YAFS, we are collecting data from over 14,000 households in the country. This will yield a large data set, to be sure, but just as important as the size of the data is the science of its collection.
Field staff of the MIMAROPA team help identify the enumeration areas and plan out survey operations in one sample barangay in Marinduque. | Photo by Field Supervisor Marvin Plata
As in any probability survey, the concept of randomization is at the heart of YAFS5’s sampling design. It means that the sample is selected based only on chance, and we do not necessarily target specific groups such as indigenous peoples (IPs) or communities with high rates of teenage childbearing. In other words, our sampling design ensures that the data provide a representative picture of young adults in each region and the whole country.
Like its predecessor, the YAFS5 employs a two-stage sampling design. In the first stage, we randomly selected 974 sample barangays in the Philippines, regardless of which province and city/municipality they are from. This figure was computed based on a target of about 800 youth respondents per region and 15 youth respondents per sample barangay. The number of sample barangays in each region varies depending on the number of youth aged 15-24 in the region – that is, there are more sample barangays in regions with more youth. For instance, there are more sample barangays in NCR compared to BARMM because there are 2.3 million youth aged 15-24 in NCR while there are 958,000 youth in BARMM as of 2015. The sample barangays are spread out in almost every province and cover about 700 cities and municipalities in the country, as Table 1 shows.
In the second stage, the field team randomly chooses a target of 15 households in every sample barangay through systematic sampling. Starting from a randomly selected household, the next sample households are identified using a sampling interval that is computed based on the number of households in the sample barangay. In the case of a large barangay (a barangay with more than 400 households), the field team divides the barangay first into several enumeration areas (EA), as determined by natural (e.g., river), administrative (e.g., purok, sitio) or structural boundaries (e.g., bridge, road). Only one EA is randomly selected from which the sample households will be randomly chosen.
All sample households, regardless of whether there are young people or not, are enumerated using the household questionnaire. If a household is not eligible for interview, the interviewer moves on to the next household using the same sampling interval until the field team reaches the target of 15 completed household interviews. After each completed household interview, all members aged 15-24 in the household are eligible for interview using the individual questionnaire.