Estimation of age distributions, census coverage and death registration completeness in Kenya

Abstract

This thesis has examined two aspects of demography t namely; the completeness of death registration and reIative coverage of censuses in Kenya at the national level and by sex. The data used was the 1969 and 1979 censuses and the death registration during the period between 1969 and 1979. Two techniques were used, namely; His version of the generalized growth balance technique and the Palloni - Kaminski foward - backward projection method. The results showed that the completeness of death registration was 26.18 percent for males and 18.25 percent for females using Hill's technique. The Palloni - Kominaki method gave relatively lower death registration completeness of 19.39 percent for males and 14.50 percent for females. The relative coverage of the the 1979 census with respect to the 1969 census was 98.79 percent for males and 97.11 percent for females by the Hill's technique. Using the Palloni - Kaminski method,however the coverage was 103.58 percent for males and 100.57 percent for females. The second aspect of the thesis was the detection and correction of errors due to age misreporting. For detection,the age ratio technique and the age specific growth rate technique were used. For correction, the Demeny - Shorter method and the Saxena - Gogte method were used. Having corrected the errors due to age misreporting individually and jointly, then the age specific growth rate technique was again applied to the corrected data for a check up. The effect of changes in census coverage mortality level and life expectancy at old age was also looked into.