Abstract

ARGOTE, Milenka L.. Sensitivity analysis of births in respect to Global Fertility Rate. Pap. poblac [online]. 2012, vol.18, n.72, pp.85-112. ISSN 2448-7147.

In the present article we perform a sensitivity analysis of births as for small changes in the change of global fertility rate (GFR) to determine whether there is butterfly effect. In relation to the method, we take as a data source the sample of the national Survey on Demography and Health (ENDSA) of Bolivia in 2003, which is a probabilistic sample at national level, stratified and two-phased; separately, the algorithms of estimation of GFR, of resampling and generation of statistical distributions of the births are created by quinquennial age groups from the sample. An index was built to assess the normality and homogeneity of distributions and a sensibility analysis was carried out by objective of the births based on the distributions. The results, according to the index the distribution of the 20-25, 25-30, 30-35 age groups are more similar to the normal and homogeneous; for the 15-20, 35-40, 40-45 age groups the distributions are similar to the normal and heterogenous; and the 45-49 group is different the normal and the most heterogeneous. According to the sensitivity analysis of births at variations of half, one and two percent over and under the mean GFR there was a nonlinear heterogeneous behavior in the different quinquennial age groups, and for the last group the variation percentage is higher than 30%, so there is an effect that amplifies the error in births; therefore, GFR is a sensitive variable, which according to these results is a factor to be accounted for in demographic projections.

Keywords : sensitivity analysis; resampling; statistical distribution of births; butterfly effect; error amplification.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )