Introduction
In addition to census data, the ONS Longitudinal Study (LS) includes information on a number of registered events occurring to LS members. Three types of events are collected for the LS:
- Events causing the entry of new members to the study (birth on LS dates or immigration)
- Events causing exits from the study (death of LS members or emigration)
-
Events occurring to LS members while alive and active in the study (birth of children, cancer registration, widow(er)hood)
The quality of events data within the LS depends on two components: firstly, on the quality of the original event data collected for the England and Wales (E&W) population; secondly, on the quality of the sampling of these events and of their linkage into the LS.
Events sampling and linkage quality are measured by sampling fractions and linkage rates. By definition, entry events cannot be linked to existing LS members, so an entry rate is used for them rather than a linkage rate.
- The sampling fraction is calculated as shown in FIGURE 1
- The linkage rate is calculated as shown in FIGURE 2
-
The entry rate for events is calculated as shown in FIGURE 3
Because the LS is a sample of around 1.1 per cent of the population of England and Wales, sampling fractions for each event should be close to this figure. If they are larger than 1.1 per cent, then the LS is "oversampling" that event; if they are smaller, there is "undersampling". Ideally, linkage and entry rates should be as close as possible to 100.
Sampling fractions and linkage/entry rates are normally calculated for each calendar year. The exception to this is during census years, when estimates are derived separately for two periods: from the start of the year up to the day before census night (pre-census), and from the day after census night until the end of the year (post-census). The formulae (FIGURE’s 1, 2 & 3) are then applied using the number of events occurring or expected to occur in these two separate parts of the census year. Details of how census part-year England and Wales figures have been derived can be found from the Census Years Apportionment page.
The expected number of events is calculated in different ways depending on the type of event and the probability of that event occurring in the LS population. Details of each method of calculating expected events are given in the relevant event page.
Users should note that the methods to calculate the expected number of events and the census year apportionment factors have been revised from previous publications. All linkage and entry rates published here have been derived using the new methodology and as such may differ from previously published figures.
