Page 205 - Contributed Paper Session (CPS) - Volume 2

CPS1824 Sanggi L.
used are household size, type of living quarters, occupancy type, apartment
area, sex, age, etc.

2.2 Following rules
The panel sample has evolved over time with household structure changes.
Some individuals move out to form their own household, others move
overseas or die, and other individuals move in, or are born. All members of
sample household in 2015 are considered Permanent Sample Members(PSM).
All other members are Temporary Sample Members(TSM), and are considered
part of the sample for as long as they share a household with a PSM. It
corresponds to ‘initial population plus current cohabiters’ of type suggested
Lynn(2006).

Figure-1: Examples of the longitudinal development of a panel (Lynn, 2006)
Table-1 : Number of samples in the KPCLS
(unit : households, persons, %)
wave 1 wave 3
wave 2 (B-A)/A
(A) (B)
Cross- Household 382,218 411,215 427,839 11.9
sectional HH members 953,430 1,045,631 1,083,501 13.6
Longitudinal HH members 929,808 908,105

Population changes by immigration, emigration, births and deaths.
Emigrants, births and deaths can be captured by following rules, but it is
difficult to find immigrants after the wave 1 by following rules. Therefore as
time goes on, the sample will gradually differ from the target population,
which affects the reliability of the cross-sectional estimates. To solve this
problem, there are methods such as introducing rotational panels or top-up
samples.

194 | I S I W S C 2 0 1 9