Analysis of KCSE performance in Nakuru County: A generalized estimating equations Approach.

Abstract

In the Kenyan education system, progression in tertiary education is dependent on a standardized national examination administered by the Kenya National Examinations Council (KNEC). The ministry of education guidelines stipulates that the pass mark for the university entry examination is C plus and above. A student who scores C+ or higher is eligible for direct admittance to university program. Publicly available data on Kenya Certificate of Secondary Education (KCSE) performance in Kenya for the years 2006-2010 was analyzed. Differences between the different school types (boys only, girls only, or mixed schools) as well as differences in performance between boys and girls were assessed. A generalized estimating equations marginal model was applied in order to account for association between scores within a school in the five year period using the SAS procedure PROC GENMOD. Flexibility in the trend was captured by additional quadratic and cubic time effects. GEE goodness of fit statistics, the quasi-likelihood under independence model criterion (QIC) was used to select best mean model as well as best working correlation structure for the study. Finally contrasts of interest were performed. A model with school, gender specific intercepts and common slopes was selected with exchangeable correlation structure. Results indicated that there was a significant difference between the different school types in their candidates probability of attaining the stipulated minimum university entry grade. In particular, boys only schools had the highest probability, followed by girls only schools and finally mixed schools. Moreover contrasts indicated that boys in boys only schools had a higher success rate than boys in mixed schools. Girls in girls only schools had a higher success rate than girls in mixed schools while boys in mixed schools performed better than girls in mixed schools. The success rate in KCSE however did not depend on the year under review as was evident in the linear, quadratic and cubic slope parameters which were not statistically significant. Keywords: Exchangeable correlation,Generalized estimating equations, KCSE, QIC