Zubeda Khan

Fertility histories: with and without restrictions -- an analysis of PLM data.

Data were collected from the Pakistan Population Labor Force and Migration (PLM) Survey on reproductive histories of 9416 married women with 38,746 children relying on a sample of 11,000 households. Biased histories could produce biased estimates of the duration of breast feeding and contraceptive use prevalence, as information gathering was restricted to the last and open intervals only. 14% of these intervals were started 10 years before the survey and 9% of them 19 years before. Using a logistic model the 2-12 year period prior to the survey was selected to analyze birth history. Birth intervals initiated 2-6 years, 2-5 years, and 2-4 years preceding the interview was determined. The estimate based on intervals began 2-12 years before the interview and was compared with other sets of estimates. Birth intervals 2, 3, and 4-8 were examined separately. The longer the period prior to the survey in the unrestricted sample, the higher the amount of unbiased results. 85-96% of the results were unbiased. When the fertility history was restricted to the last closed and open interval only 25-57% of the results were within the confidence interval. Breast-feeding duration and contraceptive use were biased up to 75%.