groupes a individuelles 2 bis

The methods that permit the simultaneous analysis of a number of demographic phenomena, introducing different individual characteristics, had been put into place and permit a very detailed analysis of individual event histories. Numerous demographic or economic applications have been undertaken using these methods by Nong Zhu in China, Ewa Fratczak in Poland, Philippe Bocquier in Senegal, Fatima Juarez in Mexico. Nico Keilman gives a good synthetic approach of these methods.

Memory problems Life history data can be collected prospectively or retrospectively. The long delays associated with a prospective survey before analysis can be attempted, and the important losses of surveyed individuals through time, remove the interest of this approach. On the other hand, retrospective surveys avoid these difficulties and are preferred.However they may lead to important memory errors. It is therefore necessary to test the impact of these errors and see their effects on event history analysis of such data. A survey had been undertaken in Belgium, a country which keeps population registers, in order to verify the quality of retrospectively collected data. This survey showed that, if the errors are insignificant for family events, they became very important for migration. However a number of event history analyses on four separate data sets, the responses of husband and wives interviewed individually, then jointly as a couple and finally the data from the register, show that in spite of these errors, they have little impact on the analysis. Even when errors in the dating of past events are frequent, their sequence is correctly memorised, and the errors only form a kind of background noise, which does not prevent coherent information from being drawn from all the sources. Thus memory seems to be reliable when the analysis needs it to be. Incomplete life histories The search for more exhaustive data sources for life history analysis leads to interval-censored data that is information available only at certain times and not continuously through time. The methods to undertake such an analysis may be tested by complete retrospective data from the ‘Family, occupational and migration’ (3B) survey. INSEE’s Demographic Panel Study gives complete family histories but only incomplete migration histories, from the 1968 census. The analysis of such incomplete data shows that the quality of the information may need very complex and expensive treatment in order to resituate a more complete migration history. The survey on social, geographical, and property mobility in France during the nineteenth and twentieth centuries, undertaken by Dupâquier follows a sample with interval censored migration histories. It permitted an analysis of the moves in the past. However the lack of census data in this sample, which may have been collected using the individual schedules, makes such estimation less reliable.

From a first outline to the finished works
Oils Paintings
Watercolour
Murals
Drawings
Covers of book
Personal greting cards
Stage sets
Artistic Biography
Introduction of space into demography
From the study of groups to the study of individual event histories
A contextual and multilevel analysis
CV Scientific
National acknowledgement
International acknowledgement
Propagating new methods
Introduction
Paradigms and axioms
Methodology
Validity of hypotheses
causality
Other themes
Books
Contributions to edited volumes
Refereed articles
Reports
Papers for Conferences, Seminars
Video