The overall objective of this dissertation is to develop and empirically assess an integrated and systematic theoretical framework to illuminate the effects of polygyny on infant mortality. Mosley and Chen's analytic model was integrated with that of Nam, Eberstein, Deeb and Terrie to provide a unique framework for guiding the application of multi-equation structural models to data from the Ghana Demographic and Health Surveys (GDHS). / The pertinent question examined in this study relates to the role marriage form plays in infant mortality. The key point is that if polygyny is beneficial as expected from the environmental theory, for example, then comparable infants should have a lower probability of dying in polygynous than in monogamous unions. This analysis tests the relative effects and the mechanisms through which polygyny affects infant survival. Another dimension studied is the mediating role which polygyny plays in the link between background factors and infant survival. / The study is carried out using a sample of infants born in the five years preceding the 1988 Ghana Demographic and Health Survey (GDHS). Two statistical models are used, a structural equation method and simulations. A causal model is constructed in which maternal education, maternal occupation, place of residence, ethnicity, polygyny, age at first marriage, preceding birth interval, prenatal care, breastfeeding, and type of toilet facility in the household are examined as determinants of infant mortality/survival. Socioeconomic variables are considered as exogenous with polygyny, age at first marriage, preceding birth interval, prenatal care, breastfeeding, and toilet facility as endogenous in the model. The model is tested by maximum likelihood method and logistic regression. / The findings indicate that maternal education, occupation, rural residence and ethnicity positively influence polygyny, which in turn affects infant survival. Polygyny is found to be an important mediating factor between background characteristics and the proximate determinants in affecting infant survival. Traditional care, unsupplemented breastfeeding, and the presence of sanitary toilet facilities are also found to affect infant survival. The results from both the logistic regression and maximum likelihood procedure indicate that the effect of the proximate determinants are in the expected direction. These findings highlight the relevance of family structure (polygyny) as an intervening factor in accounting for infant survival. / In conclusion there is evidence to suggest that maternal socioeconomic characteristics are associated with not only polygyny but also maternal behavior, environmental contamination and infant survival. Overall, polygyny appears to mediate the effect of socioeconomic variables and proximate determinants on infant survival. / In terms of policy implication it is suggested that mothers be given basic techniques regarding disease prevention. Above all, consideration must be given to family structure in planning and implementing programs with respect to birthspacing, prenatal care and breastfeeding. / Source: Dissertation Abstracts International, Volume: 54-07, Section: A, page: 2747. / Major Professor: Isaac W. Eberstein. / Thesis (Ph.D.)--The Florida State University, 1993.

Mortality in the terminal age intervals may increase as societal conditions improve, because increases in death rates for chronic conditions may offset declines in death rates for acute diseases. We examine mortality at ages fifty and above in female populations of thirty-eight countries. We control for variation in quality of the mortality data, and find that economic development, economic distributional inequality, and primary health care have independent cross-national effects on cause of death structures. These effects are not uniform across the age intervals of interest. As improvements occur in level of living and health care, age-specific death rates decline except at the oldest ages, at which point they may increase. These results are interpreted in terms of their relevance for mortality research, theory, and policy. / Source: Dissertation Abstracts International, Volume: 45-01, Section: A, page: 0309. / Thesis (Ph.D.)--The Florida State University, 1983.

This dissertation attempts to examine the relationships between socioeconomic and demographic variables and current use of contraception among Kenyan women aged 15-50. / The study utilizes both path analysis and multiple regression to determine the direct and indirect effects of selected predetermined variables on contraceptive behavior of currently married, fecund and non-pregnant women. / The results demonstrate that number of living children, ethnicity, woman's occupation and husband's occupation have significant direct effects on contraceptive use, while accessibility is shown to have the strongest direct effect on contraception. Education is shown to influence contraception behavior through other intervening variables such as women's occupation, living children and accessibility. It is also indicated that age, ethnicity and place of residence are strong determinants of education, while education was in turn positively related to woman's occupation. It is argued that the relationship between fertility preference and contraceptive use is significant particularly in the study of unwanted fertility and unmet needs for contraception. / The study concludes that policy makers ought to determine the demographic and contraceptive needs of different subpopulations in order to formulate better strategies relevant to these groups. / Source: Dissertation Abstracts International, Volume: 45-09, Section: A, page: 2999. / Thesis (Ph.D.)--The Florida State University, 1984.

This study attempts to identify some of the causes of migration disposition based on a field investigation conducted on a sample of the young rural population in Kenya in 1979. / The motivation for this study stemmed from the fact that migration is frequently seen as undesirable and the root cause of problems at both origin and destination. Such migration frequently results in a loss of the most talented young people from the rural areas. The migration of young adult and adult males leaves many rural households without male heads. Within urban areas, migration contributes substantially to the high rates of urban population growth and the associated problems, such as under- and unemployment, housing shortages and the shortage of basic infrastructure. Moreover, the lack of agreement about the causes of migration has greatly reduced the ability of policy makers to control this process. / Continuing migration from the rural areas indicates that the process is performing a role in these rural communities. To understand migration in this established form calls for an examination of the factors operating in the premigration stage that dispose some people to migrate and others not to migrate. / This study assumes that people's behaviour is motivated by their attitudes and attitudes are to a large extent socially determined. Migration dispositions is thus behaviour determined by conditions in the rural communities. New attitudes appear to have emerged in rural communities and migration appears to be the means through which they can be realized and it is sanctioned, and norms to induce conformity to the migration behaviour have emerged. In other words, migration appears to have become a socially accepted behaviour. The rural society is socializing the young to expect to migrate in future. Thus migration dispositions are examined within the socio-cultural milieu where attitudes to migrate are formed. / The log-linear modeling used to analyze the data confirms that factors in the objective environment, normative factors, and psychological factors determine migration dispositions. These findings suggest that young people will continue to be disposed to migrate from the rural areas as they continue to westernize. / Source: Dissertation Abstracts International, Volume: 43-08, Section: A, page: 2799. / Thesis (Ph.D.)--The Florida State University, 1982.

This dissertation is a work about the effects of education and religion on the current family size and the fertility preferences on the Kenyan women. / It utilizes the method of path analysis and multiple regression to test a series of models, both recursive and non-recursive, that purportedly explain the variation in the dependent variable of interest. / The findings indicate that western religion, ethnic affiliation and age are important determinants of educational attainment. Educational attainment is in turn an important determinant of age at first birth and does not affect significantly, and directly, the current fertility levels of the Kenya women. It, however, affects directly the number of additional children wanted. The study also demonstrates the strong negative effect of current fertility levels on fertility preferences. / It is however shown by the non-recursive model that educational attainment has a reciprocal relationship with age at first birth, and so has the number of living children with additional children wanted. The non-recursive model is shown to be a better predictor of both the current fertility and fertility preferences, than the recursive model. / The study concludes that relationships between fertility levels and the variables used here can be country specific, but also bear close resemblance to findings in the same area in western nations. Age at first birth, education and religion are seen as potentially practical variables for policy use, to influence both the current and expected future fertility levels in Kenya. They are, however, part of a rubric of factors relevant for changes in Kenya's fertility levels in whatever direction. / Source: Dissertation Abstracts International, Volume: 43-10, Section: A, page: 3425. / Thesis (Ph.D.)--The Florida State University, 1982.

This research acknowledges the importance of the labor market structure in determining the relationship between female employment and childbearing. The labor market structure used in this research is a modification of a framework presented in Edwards (1979). There are two dimensions in the framework; (1) occupations and (2) systems of control or industries. The industrial segments can be distinguished on the basis of schedule flexibility, the presence of fringe benefits, work hierarchies, and work activities. The differences across occupational segments are primarily based on level of expertise, ease of entry and exit, work commitment, and human capital investment. Data from the U.S. Survey of Income and Education conducted in 1976 are used to test the impact of the labor market structure on the female employment and childbearing relationship. All of the women studied are currently married with husbands present, 18-35 years of age, and employed in 1975. The dependent variable is a dichotomy with the presence of a child less than one year of age in the household coded 1. A probit estimation technique is used to determine the hypothesized effects. The impact of the labor market structure on the female employment and childbearing relationship is estimated using two analytical approaches. In the first approach, the location of women in each labor market segment is treated as a proxy for the female employment experience. In the second analytical approach, the labor market segment is treated as a contextual effect. Based on the results in both analytical approaches, the simple independent primary, the bureaucratic secondary, the technical secondary, the bureaucratic independent primary, and the technical subordinate primary segments appeared to accommodate childbearing and rearing to a greater extent than the other labor market segments. The flexibility of work / schedules as well as the generosity of the fringe benefit packages were believed to be contributors to the variation in effects across labor market segments. It was suggested that weeks worked is an appropriate employment feature to adjust if the accommodation of childbearing and rearing among employed females is a desirable outcome. This research indicates that female employment and childbearing and rearing can be compatible if the appropriate resources and services are present. / Source: Dissertation Abstracts International, Volume: 44-06, Section: A, page: 1935. / Thesis (Ph.D.)--The Florida State University, 1983.

A majority of elderly migration/residential mobility studies to date have been cross-sectional and atheoretical, and have not considered the effects of cohort or period, but only that of age. In this study the 1940-1980 U.S. Census public use sample micro data files are used in performing cohort and multiple-period analyses. Hypotheses are developed specifying the expected relationships of determinants of elderly long-distance migration and elderly local mobility derived from the literature and introduced by the author, and levels of interstate migration and local mobility, and the expected effect of more or less recent period-of-birth cohort membership on these relationships. The results of the tests of these hypotheses indicate that in addition to the effects of age, period, and cohort on migration/mobility levels per se, there are also marked differences in the effects of the independent variables when examined within different periods and within cohorts. The discussion points out the problems with generalizing from research using data from only one or two periods, and the need for research that takes into account factors underlying cohort and period differences in levels of different types of elderly mobility, as well as the importance of their interaction, if advances in theoretical development in this area are to be made. / Source: Dissertation Abstracts International, Volume: 47-07, Section: A, page: 2752. / Thesis (Ph.D.)--The Florida State University, 1986.

This study tests the female education threshold hypothesis which posits that there is a critical level of educational attainment beyond which fertility begins to decline from traditional high levels. Using data from a sample of currently married women of child bearing ages 15-49 in rural Sierra Leone, non linear regression analyses reveal a threshold value of six years of schooling for rural women, varying substantially within and between subgroups of the population. Women below and above the threshold value exhibited the expected positive and negative coefficients on fertility respectively, even after controlling for a variety of variables known to influence fertility such as maternal age, tribal and religious affiliations, household income and size of place of residence. / The relationship between proximate variables and threshold level was also examined. The findings show that the combination of proximate determinants through which female education operates to produce the expected negative effects on fertility for women above the threshold, also holds in varying combinations for different subgroups of the population. / Overall this study suggests that further advances in understanding the relationship between fertility and the female education threshold concept could be enhanced by a closer look at segmented threshold levels derived from subgroups of the population and the role played by proximate determinants in this relationship. / Source: Dissertation Abstracts International, Volume: 47-07, Section: A, page: 2752. / Thesis (Ph.D.)--The Florida State University, 1986.

The primary objective of this study was to explore the interrelationship between fertility and migration using a sample of rural Kenyans, ages 13 to 21. This was done by examining the differences in mean levels of desired family size for migrants and nonmigrants. No substantial or significant differences were found which suggests that migration is not "selective" of individuals with low levels of desired family size. / A second step in this analysis examined the mean levels of desired family size for individuals who had gone through a decision-making process in regard to migration and those who had not. It was found that those respondents who had been through the decision-making process had substantially and significantly lower levels of desired family size, even after controlling for a variety of variables shown to affect both migration and fertility. / A follow-up procedure allowed the differences in mean levels of desired family size to be attributed to either the act of migration or the decision-making process. This analysis showed that the differences due to the decision-making process were generally larger and more likely to be statistically significant. Thus, these findings indicate that the decision process is a more important factor then migration in influencing lower desired family size. / Overall this research suggests that further advances in understanding the migration-fertility relationship could be enhanced by closer examination of the correlates and consequences of the decision-making process. / Source: Dissertation Abstracts International, Volume: 46-01, Section: A, page: 0271. / Thesis (Ph.D.)--The Florida State University, 1984.

This study provides a comparative evaluation of the effectiveness of survival rate and time-series regression techniques used for estimating state and substate net migration across the years 1960-1980. Estimates from these techniques are compared to a theoretically exact benchmark series of actual net migration calculated from the vital statistics method. / Florida and New York and their state economic areas underwent substantial and divergent population change during the time period of this study. The state of Florida was one of the fastest growing large states in the nation, but some of the state economic areas grew very slowly during this time. The state of New York lost population; some state economic areas lost population the entire time, but other areas increased in population size. These divergent patterns provided a rigorous test of the two techniques compared. / Survival rate and time-series regression estimates of net migration are compared to the benchmark by these divergent areas of population change. There were no statistically significant differences between the survival rate estimates and the benchmark series for any state economic areas, with the exception of the areas in New York which underwent stable (<5%) population growth. Nonparametric tests showed differences in patterns of the signs between these series for the loss areas of New York as well. The New York state level survival rate estimates were also substantially different from the benchmark series. Inaccuracy in the survival rate estimates may be due to the impact of migration on the age structure of the areas. / Regression models were developed specifically for state economic areas, by population change areas. The final models reflected the economic and demographic characteristics of the areas. However, regression estimates of 1980 net migration compared to the benchmark fared poorly at all geographic levels, with the exception of the state of New York. / It is concluded that geographic level of analysis does play a role in research results, partially explaining past inconsistencies in the literature. The regression results are further discussed in terms of the properties of the models and the functional forms of both the regression equation and the independent variables. The survival rate results are also discussed in terms of the statistical properties of the model. / Source: Dissertation Abstracts International, Volume: 47-12, Section: A, page: 4516. / Thesis (Ph.D.)--The Florida State University, 1986.