POPLINE Document Number: 202873
Author(s):
Abul-ata MF
Source citation:
Chapel Hill, North Carolina, University of North Carolina at Chapel Hill, The Institute of Statistics, 1987. xii, 231 p. (Institute of Statistics Mimeo Series No. 1841T.)
Abstract:
3 stochastic models for the i-th order birth interval are developed corresponding to each of the following ways for considering birth interval data: 1) all closed birth intervals (retrospective and prospective studies), 2) last closed birth interval, 3) prospective next (straddling) birth interval, 4) prospective interior birth intervals, and 5) open birth intervals. The method produces estimates of fertility hazards that are recovered (purged) from truncation bias for all closed, last closed, and open birth intervals. These hazard estimates are further utilized in driving 3 important length-unbiased cohort fertility quantities, viz., mean length of interlive birth interval, parity progression ratio, and birth interval survival function. The models were applied to data from the Egyptian Fertility Survey (1980) for a marriage cohort with 5-9 years of intact marital life. While the recovered birth order-specific hazard estimates are generally of the same magnitude irrespective of the ascertainment method of considering birth interval data, they are largely below the corresponding estimates derived from straddling birth intervals. Comparison between model-based and life-table-estimated survival probabilities of the 2nd birth interval indicates that the former is more likely to be better adjusted for truncation bias. (author's modified)
Keywords:
EgyptIndex page
Mathematical Model
Evaluation
Models, Theoretical
Research Methodology
Birth Intervals
Fertility Measurements
Measurement
Fertility
Comparative Studies
Bias
Life Table Method
Estimation Technics
Cohort Analysis
Fertility Surveys
Demographic Surveys
Africa, Northern
Africa
Developing Countries
Population Dynamics
Demographic Factors
Population
Studies
Error Sources
Demographic Analysis