These tools will no longer be maintained as of December 31, 2024. Archived website can be found here. PubMed4Hh GitHub repository can be found here. Contact NLM Customer Service if you have questions.


Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

  • Title: [Some problems related to the collection of demographic data in the population censuses taken in Latin America during the 1980s].
    Author: Arretx C, Chackiel J.
    Journal: Notas Poblacion; 1986 Dec; 14(42):51-76. PubMed ID: 12314803.
    Abstract:
    Questions permitting indirect estimates of population movement have become more common in Latin America censuses because they are the most timely and reliable basis for estimates when good vital statistics are lacking. A question on the Survival of the last child or of children born in the last year (unlike the more common questions on total number of live births, survival of all children born, or date of the last live birth or the number born in the last year) has not yet provided good results and is not recommended by the Latin American Demographic Center (CELADE) except in experimental censuses and demographic surveys. All countries should include a question on the "year of arrival" for foreign-born persons so that migration trends over time can be assessed. Some countries with incomplete vital statistics have omitted a question on whether the mother of each person in the household is living or not; this question has been very useful in estimating adult mortality. Although there has been great progress in including questions allowing study of population dynamics, there are disquieting indications of decline in the quality of data in some of the Latin American censuses taken in the 1980s. The content of questions in the census form or the order or manner in which they are included may lead to errors. Data users should participate in formulation of questions to ensure that the desired information is sought, and persons intimately familiar with local usage should participate to ensure that questions will be understood. An example of a poorly framed question occurred in the fertility section of the Argentinian census, where the 8th and subsequent births were lumped together in a precoded question, greatly limiting the usefulness of the data. The user should not have to skip from 1 place to another on the form; the inconvenience may lead to error or omission. The best order for questions is still a matter of study. In general questions pertaining to the entire universe should be followed by questions directed to specific subgroups, with clear instructions provided. The number of questions should be limited. Most countries include about 20. Precoding should be carefully planned to avoid loss of needed detail. All aspects of the census form should be subjected to careful evaluation in pilot and experimental censuses. There are indications that omission rates increased in the most recent census for 5 of the 9 countries whose 1970 and 1980 decade censuses were evaluated by CELADE. When nonresponse rates for particular questions become too high, the temptation to fill in the information using special problems should be avoided because of the danger of biasing the results. The original data should be preserved if assignments are made, and the criteria for assignment should be clear to the reader. The major risk of having a sample of households complete a more detailed census form is that the sample will not be representative. CELADE does not recommend post-census surveys because the amount of information they can impart does not justify their significant diversion of economic and manpower resources.
    [Abstract] [Full Text] [Related] [New Search]