| dc.description.abstract | This study is concerned with the investigation of the factors
that influence infant/or early childhood mortality at the macro level. The factors investigated include the following district level variables: per-cent women with five or more standards of education; percent cases of malaria; hospital beds per 1000 persons; total fertility; population to health facility ratio; per capita high potential agricultural land as a proxy for
rural income; kilometres of roads per 1000 square kilometres as a proxy for socio-econo[lic development; percent urban; and population densityo
The investigation is divided into six chapters. Chapter
1 gives the historical background of mortality in the developing countries, Africa and Kenya in particularo Chapter 11
discusses and presents all the data utilized in ttis research.
Chapter 111 describes the models used in this studyo Chapter IV is the analysis of infant/or child nortality differentials and their correlates in Kenya. Chapter V examines the relationship between infant/or child mortality and the educational levels of mothers. Chapter VI gives a sUInmary of
the findings, conclusions and recommendations for further studieso
This is followed by appendices and references/bibliographyo
A multivariate analysis was used to examine the 1979 mortality differentials in Kenya by districtc Data for all the variables by districts were gat-her-eadn.d.r-egr-sesdon wasvper-f'ormed
on these datao It was found that a significant correlation exists between infant/or child mortality and education and malaria variableso The reSults for the other
ex)
variables were rather inconclusive.
Only four out of nIne variables considered explained 52.5 percent of the variation present In the mortality index as compared to 59.4
percent explained by all the variables together. These variables were as• follows: women's education, malaria, agriculture and population
density.
The district data analysed in this study revealed that the variations in child mortality levels by district are partly a result of the existing educational differentials among the districts. Indeed,
inter-regional differentials in mortality appeared to persist even after
controlling for differences in Ivomen's educational levels. Thus vomens education plays a major role as a major factor in mortality decline only.
All the hypothesized relationships were confinned except for the
'hospital beds' variables whose positive relationship with child mortality as suggested by the regress ron results, was not anticipated | en |
