Includes abstract. / Includes bibliographical references (leaves 63-61). / Anglewicz, Adams, Obare et al (2009) show that the mean parities for the women who were interviewed in the Malawi Diffusion and Ideational Change Project (MDICP) surveys of 1998 and 2004 are generally higher than the mean parities for the women who were interviewed in the Malawi Demographic and Health Surveys (MDHSs) of 2000 and 2004 respectively.

Death distribution methods, particularly the Generalized Growth Balance (GGB) and the Synthetic Extinct Generations (SEG) methods, have been observed to lead to the most accurate estimates when estimating mortality [1]. The more general version of the SEG method corrects for differential coverage of censuses directly by adding a constant (6) to the age-specific growth rates such that the correction leads to a horizontal series of age specific estimates of completeness. This research attempts to obtain the best variation of this version of the SEG method from a range of choices for an open interval age as well as well as methods of estimating life expectancy. completeness and 6. This task is accomplished by starting with a base population with known mortality then applying random errors in completeness. age misstatement and net migration to it to generate numerous datasets consisting of simulated census counts and simulated vital registration deaths by age. Variations of the SEG method are then applied to the simulated datasets to correct for the underestimation of mortality caused by the data errors. The best variations are found by statistical analysis of the difference between the true mortality and the estimated mortality for each variation and dataset generated. Using the Coale and Demeny model life tables to estimate life expectancy. selecting the a that results in a minimum variance in the age specific estimates of completeness. estimating completeness using the median value of the age specific. estimates Of completeness for ages 15 and older and using the 85+ age group for the open interval is observed to be the variation of the SEG method that leads to the most accurate estimates of mortality.

It is important for health planners to have timeous and accurate data on deaths. The Department of Home Affairs is responsible for the registration of deaths and the City of Cape Town has a well-established system of collating the death statistics based on vital registration, but the completeness of the death registration has not been assessed previously. The completeness was assessed for the City of Cape Town by comparing their statistics with an estimate based on data obtained from adult deaths reported in the 2001 census. A second approach assessed the trend in completeness between 1996 and 2004 by identifying three rates of mortality considered to be stable over time (non-lung and non-oesophageal cancers, the 10-14 age group and the 60+ age group) and inspecting to observe whether there was any trend apparent over time. Since deaths in most cases are under reported, and the under reporting usually differs in completeness between children and adults, child deaths from the ASSA model projection assuming that they are more complete were compared with the child deaths from the vital registration between 1996 and 2004 to check for completeness of the child vital registration data in Cape Town and its eight health districts The results show high levels of completeness in the adult deaths for Cape Town as a whole in 2001, around 95 per cent, but varying levels in the health districts. The completeness of reporting of male deaths in Cape Town declines with age, whilst completeness for females is fairly level with respect to age, with similar trends being observed in the health districts. Completeness of child (0 -4) death registration averaged around 60 per cent, about 35 per cent lower than the completeness of adult deaths in Cape Town. Cape Town as a whole and most of its health districts revealed two levels of completeness in the registration of deaths, 1996-1999 and 2001-2004 with 2000 sometimes consistent with the first and sometimes with the second period or different from either period in some of the health districts. In conclusion, the completeness estimates obtained are more rigorous from 2001 onwards suggesting that they can be reliably used to monitor trends in the levels of mortality in the city of Cape Town.

Includes bibliographical references (leaves 70-72). / It is found that birth intervals have increased from about 28 months in the 1960s to about 51 months by the year 2000, with the greater part of this increase occuring after 1985. Fertility fell from birth intervals are congruent with each other, in terms of timing and tempo. Of the two main proximate determinants identified; marital status and contraceptive use, the latter is founf to be the more dominant force behind changes in birth spacing. Differentials by marital status are not significant. The research adds to a growing body of studies on the nature of fertility tansitions in sub-Sahara Africa, and would particularly be useful in explaining observed differences in fertility transitions between countries in the region.

Includes bibliographical references. / This research focuses on estimating infant and under-five mortality in South Africa for the period 1998-2012, both to update previous estimates taking into account new data and to assess the reasonableness of all estimates. Data from the 2011 Census and the 2012 HSRC survey were used for this purpose. The 2011 Census provided data from deaths reported by households as well as the survival of the most recent births. The 2012 HSRC provided full birth history data for women aged 15-49 which were used for direct estimation of childhood mortality. Deaths reported by households together with census estimates of the number of children under the age of five are used to produce estimates of infant and under-five mortality using the synthetic cohort life table approach.

Includes bibliographical references (leaves 62-64). / The contraceptive prevalence rate in Zambia is high, while fertility decline is very slow. From 1992 to 2007, the contraceptive prevalence rate (CPR) increased from 8.9 per cent in 1992 to 32.7 per cent in 2007; while total fertility rate (TFR) fluctuated between 6.5 and 6.2. The study uses three Zambian Demographic and Health Survey (DHS) data sets for 1996, 2001-2 and 2007 and applies multivariate logistic regression techniques to identify factors affecting the use and choice of modern contraceptives by Zambian women. The study seeks to identify how contraceptive use in Zambia influences fertility and why both contraceptive use and fertility are high. In our results, place of residence (urban/rural), age, education, number of living children and formal employment were identified as factors that are significantly related to use of modern contraceptives.

Includes abstract. / Includes bibliographical references. / An understanding of the errors found in census questions is important in order to assess the level of confidence in the census data and to get an appreciation of the impact of using these data in estimating mortality derived from census data. While demographic methods are often used to determine the direction of bias in the data, direct evaluation studies are required to determine the nature and extent of biases with more accuracy. Equally important is an understanding of the characteristics of the respondents who produce better responses in censuses. This can be used to inform selection of better respondents in order to improve the quality of the collected data. This research uses census data from a survey, which is matched to the longitudinal Health Demographic Surveillance System site (HDSS) data from Agincourt, Limpopo Province, South Africa, in order to assess the biases found in data used in child and adult mortality estimation that uses indirect techniques.

Includes bibliographical references. / Subnational estimates of child mortality are difficult to produce and are rare in Sub- Saharan Africa. It is the overall aim of this research to derive estimates of child mortality rates for the municipalities and provinces of South Africa using the 2011 census data, and to assess the results in relation to the level of poverty and inequality. The estimation of child mortality rates is achieved through the use of direct synthetic cohort methods with Bayesian spatial smoothing. The Bayesian spatial smoothing process is used to generate municipal level estimates of child mortality rates. The model utilises information from neighbouring municipalities by controlling the effects of women’s education and HIV/AIDS.

Includes bibliographical references. / This dissertation investigates the robustness of recent births in estimating infant mortality rates from the proportion of deaths observed among births reported in a 24month period. The Blacker Brass technique is applied to all births reported in the 24month period and to most recent births in the 24 month period. The study uses birth history data from 76 Demographic and Health Surveys conducted in 16 countries across the developing world between 1986 and 2011. All births (and the deaths of those births) occurring in five 2-year periods before each survey were extracted to obtain five estimates of infant mortality using the Blacker-Brass and direct estimation methods from each dataset. This allows trends in infant mortality for the 10-year period before the survey to be compared and relative errors to be calculated. The results showed a decline in infant mortality in most datasets and are consistent with the United Nations and the World Health Organisation 2013 estimates. The relative errors did not indicate any systematic bias of the Blacker-Brass method applied to all births; however, further investigations showed that the method underestimated infant mortality in the period closest to the survey date in most datasets. Furthermore, the relative errors were positively correlated with the directly estimated level of infant mortality. There were, however, no significant differences in the relative errors across countries.

Includes bibliographical references (leaves 78-83).