Hello,
I work with two demographic surveys carried out at two point in time (1987
& 1992), in which retrospective data (marriage, births,...) have been
collected on all married women aged 15-49 at the time of the survey.
I would like to pool the two data sets and analyse the retrospective data
(age at marriage) with event history models.
My question is the following :
The period of time before 1987 will be "overrepresented" for the cohorts
aged less than 45 if I analyse the data together, since these data come
from two surveys. There is also the fact that the sampling fraction is
different in the two surveys.
I suppose I should use some kind of weighting to take this into account,
but I do not see clearly how to do it.
Does anyone has experienced a similar problem or could anyone give me some
references on the subject.
Thank you,
Bruno SCHOUMAKER
Institut de demographie
Departement des sciences de la population et du developppement
Universite catholique de Louvain
1/17 Place Montesquieu
B-1348 Louvain-la-Neuve, BELGIUM
Ph. ++ 32 (0)10 47 44 61
Fx. ++ 32 (0)10 47 29 52
[log in to unmask]
http://www.demo.ucl.ac.be
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|