The sampling design of the Survey on Overseas Filipinos (SOF) adopts that of the Integrated Survey of Household (ISH) which uses the new master sample design starting July 1996.
The multi-stage sampling design of the master sample consists of 3,416 PSUs in the expanded sample for provincial level estimates with a sub-sample of 2,247 PSUs designated as the core master sample for regional level estimates. The October 1999 Labor Force Survey was based on the core master sample.
The domain for the new master sample is similar to that of the previous ISH design with an addition of 23 newly created domains. The urban and rural areas of cities and municipalities with a population of 150,000 or more are considered as separate domains. The other urban and rural areas in each of the 77 provinces are likewise treated as separate domains. In view of the creation of ARMM and the separation of Marawi City and Cotabato City from Lanao del Sur and Maguindanao, respectively, the urban and rural areas of the two cities also form separate domains.
The frame for the first and second stages of sample selection were based mainly on the results of the 1995 POPCEN. The 1995 POPCEN list of barangays with the household and population counts is used in the first stage of sample selection. The stratification of barangays included in the frame, however, are based on the 1990 CPH and other administrative reports from the field offices of NSO.
The enumeration areas which constitute the secondary stage sampling units are those that were formed during the 1995 POPCEN.
Isolated barangays and/or barangays that are difficult and expensive to reach are excluded from the sampling frame. However, critical areas or barangays with peace and order problem which is generally temporary in nature are included in the frame.
The frame for the third stage of sample selection is the list of household from the 1995 POPCEN.
Barangays in each domain were explicitly stratified by urbanity. Within the urban/rural barangay stratum in each of the provincial domains, the barangays were implicitly stratified by municipal district, serpentine ordering of the municipalities, and grouping of the barangay, based on accessibility. The implicit stratification ensured geographic coverage.
The multi-stage sampling design of the master sample involves the selection of the sample barangays for the first stage, selection of sample enumeration areas or EAs for the second stage, and the selection of sample households for the third stage in each stratum for every domain.
The sample barangays were selected systematically with probability proportional to size from the list of barangays that were implicitly stratified. The preliminary count of households based on the results of the 1995 POPCEN was used.
The selection of sample EAs for the second stage was also done systematically with probability proportional to size. The EAs are those that were formed during the 1995 POPCEN. An EA is a physical delineated portion of the barangay. For barangays that were not divided into EAs, the barangay was treated as an EA.
The selection of sample households for the third stage was done systematically from the 1995 POPCEN List of Households.
The new master sample consists of an expanded sample of 3,416 sample barangays/EAs (2045 urban and 1371 rural) to improve the precision of provincial-level estimates from the household surveys. The October 1997 and October 1998 LFS used the expanded sample which has a national sample of about 41,000 households.