Institute for Development Studies, University of Nairobi

Because the secondary school selection examination is so crucial to the students' life chances, what is taught in Kenya's primary schools is determined as much by the examination questions as by the formal curriculum. This paper investigates the examination's backwash effects as well as its efficiency in identifying pupils; who will make the best use of secondary school opportunities. At present the examination is mainly a selection instrument geared to identifying secondary school entrants. It largely ignores the interests of pupils for whom primary education is terminal. A high proportion of the items test academic, specialised knowledge and skills which are of little use to primary school leavers, most of whom must create economic opportunities for themselves, in agriculture or self-employment. But, paradoxically, this reduces rather than improves the efficiency of the examination as a selection instrument. Item analyses of the 1970 and 1971 mathematics papers demonstrated that items which test practical, everyday-mathematical skills are more efficient as selectors than items which test secondary-level skills. The main reason for this is that many teachers in low-cost schools (which make up 99% of all primary schools in Kenya) do not themselves have an adequate grasp of the more academic topics. The examination thus becomes as much a test of the teachers as of the pupils. For this reason, the examination is more efficient as a selection tool in high-cost, urban schools than in low-cost, mainly rural, schools. The intelligent pupil from a low-cost school is at a double disadvantage in competition with a similar pupil from a high-cost school. Because he has been less well educated, his total mark is likely to be lower, by at least one standard deviation. In addition, his chances of being identified as a pupil of high potential are much reduced because of the lower efficiency of the examination in low-cost schools. Some ways in which both the efficiency and the relevance of the examination might be improved are suggested.