1. SAMPLE

The population of KLoSA includes, in principle, all adults aged 45 and over. (Many surveys of elderly people in other countries include those aged 50 and over. But KLoSA extends its population to those aged 45-49 because career changes in these middle ages have become an important social issue since the financial crisis in later 1990s, with many people in their late 40s having been laid off from their career jobs. This extended population also enables examination of the relationship between economic activity in the middle ages with that in older ages, and with later life in general.) However, Jeju island was excluded for interviewing convenience sake. The first wave included only those living in households. Institutionalised persons, who were excluded from the first wave, will be included in the second wave of the main survey.

1.1. Stratification of EDs

The sampling frame of KLoSA comprises enumeration districts (EDs), as identified by the National Statistical Office’s 2005 Census. Under this frame, Apartment EDs and Ordinary housing (non-apartment) EDs totalled 261,237, excluding island areas and institutions (social welfare facilities). For Wave One, 1,000 sample EDs were selected, with the aim of securing a maximum valid sample size of 10,000 persons, from six households per sample ED. This was decided on the basis of the fact that the average number of household members aged 45 and over was 1.67 in the 2000 Census.

Before selecting sample EDs, the population was stratified by type of area and by type of housing. The EDs, clustered in 15 metropolitan cities and provinces, were stratified first by type of area (urban/rural), then by type of housing (Apartment/Ordinary housing, i.e. non-apartment) EDs.

An effort was made to produce reliable data even from those cities and provinces with small populations. Of the total 1,000 sample EDs, 15 cities and provinces were allocated first, each with 15 sample EDs. The remaining 775 EDs were then allocated in proportion to the population of the 15 cities and provinces. Within each city or province, EDs were allocated subsequently by type of area, i.e. urban and rural areas; again, in proportion to the population of the areas concerned.

The final numbers of sample EDs by area and by type of housing are shown in Table 1. Of the 1,000 sample EDs, 409 EDs were allocated to Apartment EDs and the remaining 591 EDs were allocated to Ordinary EDs. In urban areas, 363 EDs were allocated to Apartment EDs and 440 to Ordinary Eds; while in rural areas, 46 EDs and 151 EDs were allocated respectively.