CeLSIUS logo

Census logo

Study subjects

Since we are following individuals from one census to the next, our study subjects must be present in both 1971 and 1981 in order to be useful. This longitudinal requirement means that the sample sizes we obtain will be smaller than they were for the cross-sectional study, since certain members will enter or leave the study between censuses due to birth, death and migration.

The initial study cohort is taken from women aged 15-19 and living with their parents at 1971. However the actual study sample is all those from that group who are still present in 1981.

To illustrate this, the table below shows the numbers of women aged 15-19 and living with their parents at the 1971 census, and the numbers still present at the 1981 and 1991 censuses.

Since we need to know whether the subjects are living with their parents at the end of each study period, we need household data for their usual address at this point. Therefore, women who are enumerated as visitors or in communal establishments will not be included. (Note that the exclusions for 1971 have already been made in effect because we are looking at women living with their parents, and this can only be determined where valid household data is available.)

Therefore, the table also shows the numbers without these excluded groups.

Group Present 1971 only Present
1971 and 1981
1971 and 1991
Females aged 15-19, living with parents in 1971 13753 12288 12082
Females aged 15-19, living with parents in 1971, minus exclusions at 1981 or 1991 13753 11961 11929
Proportion excluded - 2.7% 1.3%

Data extracted from LS 05/05/03.

Notice how the sample size is slightly reduced both by the longitudinal requirement of each study, and by the exclusion of subjects for whom the appropriate household data are not available.