Mortality Levels and Trends in Jordan Estimated from the Results of the 1976 Fertility Survey

This report discusses the possibility of estimating levels and trends in fertility and childhood and adult mortality from responses to simple retrospective questions included in a single census or survey. The Jordan household survey contains a large sample size of over 14000 households. For the analysis the data used were those on the standard recode file produced by the Jordan Department of Statistics and distributed by the World Fertility Survey. 2 types of estimation were used 1) direct which means that the numerators and the denominators of a rate or a probability are taken directly from the survey data; and 2) indirect which involves using model life tables and model fertility schedules in combination with results from surveys. Regarding child mortality it is shown that the reports from young women are untypical of the entire population when they report on children ever born and surviving but otherwise estimates are reliable. The overall impression is that the mortality rates of children aged 1-4 fallen rapidly by any standards on the East Bank catching up with the fast rates of change on the West Bank. In the indirect estimation of adult mortality the assumption has generally been made that time trends in mortality can be ignored but this was frequently untrue. This report demonstrates that the data collected by most World Fertility Surveys and from the Jordan survey in particular have considerable value as a source of mortality information in addition to the information on fertility. The absence of a firmly established reference is a handicap in assessing the accuracy of the adult mortality data and results but several features stand out: 1) the widowhood data from the household survey are of poor quality mostly because of inaccurate reporting of marriage order especially by female respondents; 2) the marriage histories from the individual interview produce reasonable reports on widowhood but the reports themselves are of limited value because the questions on the survival of the 1st spouse were addressed only to women aged 15-49; and 3) the data on household deaths in the preceding 24 months have proved to be of limited value for measuring adult mortality principally because of small numbers

Fertility · Geography · Population · Sociology · Statistics · Demography · Global Maternal and Child Health · Health and Conflict Studies · Insurance, Mortality, Demography, Risk Management · Mathematics