| dc.description.abstract | v
ABSTRACT
Fertility and geographical segregation are some of the major factors influencing human population
growth. The fertility rate of Kisumu East Sub County is rated at 4.8 per woman exceeding the
average for the county, national and the global which is 4.2, 3.4 and 2.3 respectively per woman.
It has been established that the segregation may influence fertility, however, it was not well
understood how various aspects of segregation influence fertility. The purpose of this study was to
assess the influence of geographical segregation on fertility of women. Specific objectives were
to: establish the influence of exposure on the number of children; determine the influence of
concentration on the number of children; determine the influence of centralization on the number
of children; and establish the influence of clustering on the number of children born per woman.
This study was guided by Becker's economics theory of fertility and Hägerstrand theory of Spatial
diffusion. A minimum sample size of 384 women respondents aged between 18-49 years were
obtained as determined by the Fisher’s formula and selected using stratified random sampling,
cluster and snowball sampling techniques. Purposive sampling was employed to identify key
informants. Primary data were gathered by questionnaires, KI interviews and FGDs. Secondary
data were collected from DHS reports, census reports, and un-published media. Quantitative data
were analysed using descriptive statistics and inferential statistics: gamma statistics, spearman’s
rank correlation coefficients, multiple logistic regressions, multiple correlation coefficient and
multinomial logistic regression. Qualitative data were analysed by coding, creating categories,
themes and patterns then evaluating the usefulness of the information in answering the research
questions. The results showed that sharing common centres and mean number of children born per
woman, had a strong significant positive relationship (r=0.675). Daily and weekly social
interactions showed a strong positive significant linear correlation with fertility (r =0.732, p =
0.03). Cultural norm was a major factor likely to influence the first and the last childbirth.
Moreover, there is a strong positive and statistically significant linear correlation (r =0.50, p =
0.04) between social interactions and number of children born. Gamma statistic coefficient of
0.493 indicated moderately strong positive association between levels of geographical
concentrations and number of children born. Number of children born correlated negatively (r =-
0.612, p=0.02) with low geographical concentration. The results showed that the correlations
between centralization and number of children born was strong and positive (r=0.625, p=0.026).
Multiple linear regression analysis showed that 81.5% of variation of number of children born per
woman could be predicted from the combined influence of length of stay in the neighbourhood
and mean monthly income. The study recommended that productive social interactions like
seminars, conferences and group meetings should be highly encouraged among women through
setting up public arenas for social gathering where women can meet for exchange of vital
reproductive information. Emphasis should be put on women empowerment and reproductive
health. | en_US |
