Institute for Development Studies, University of Nairobi

A group of verbal reasoning questions, of the type often used in intelligence tests, were included in the English paper of the Kenya secondary school selection examination (C.P.E. - Certificate of Primary Education) from 1971, and from 1974 scientific reasoning questions were included in the general paper. In early 1974 a full-scale item analysis of the 1973 selection paper was made. This paper presents some results from the analysis carried out for the years 1973 to 1976 of the English paper and the science section of the general paper. The scores of pupils from Nairobi high-cost primary schools and from rural low-cost schools are compared. Three interesting results of this analysis are presented and some possible explanations offered. For one thing, pupils from the Nairobi high-cost schools perform up to 70 per cent better on English items testing knowledge of specialised words, expressions and idioms, whereas in the science paper, descriptive items testing knowledge of specialised and technical terms produce a mean difference of less than 12 per cent. Secondly, the verbal reasoning items in the English paper give the Nairobi high-cost pupils a smaller advantage than the achievement items; whereas in science, the reasoning items give a bigger advantage than the descriptive and explanatory items. Finally, the observation items in the science paper produce an especially large performance gap between the two types of schools. These and other results suggest that the huge performance advantage enjoyed by Nairobi high-cost pupils in the English paper can be ascribed entirely to two sources: first, their greater familiarity with the language, and second, the superior quality of the teaching they receive. In both papers, questions which test higher-level intellectual skills, such as the ability to reason, are particularly sensitive to the effects of teacher quality, rather than reflecting the innate capacity of the pupils.