Odanga, Francis2012-05-142012-05-142012-05-14http://ir-library.ku.ac.ke/handle/123456789/4581Department of Educational Communication and Technology, 59p The QA 14.O3 1992The purpose of this study was to analyse arithmetic computation errors commited by primary school pupils of standards four and six. It further sought to find out which of the types of the errors are more prevalent than the others and whether there are any other differences in the magnitudes of the errors made. The sample of the study consisted of primary school pupils of the two classes from Rauor Educational Zone in Wiham Division, Kisumu district. Six schools were randomly selected with sixty pupils from each school giving a total population of three hundred and sixty (175 boys and 185 girls). The data was collected using an arithmetic computation test developed by Professor Patel of the department of Educational Communication and Technology, Kenyatta University. The instrument consisting of eighty arithmetic computation sums involving the four basic operations of addition, substraction, multiplication and division, was administered to the pupils. The instrument was scored manually for errors and then analysed in order to categorise the errors according to Englhardt (1977). The results were then analysed using descriptive statistics and the chi-square test was used to test whether there existed any significant gender differences. From the results it was concluded that pupils in primary schools do make arithmetic computation errors. The magnitude of the errors made reduces as one moves from the lower classes to the upper ones. The most prevalent types of errors were the Basic fact, grouping and Defective Algorithm. It was also concluded that significant differences existed in some types of errors. Therefore, when teaching mathematics, there is need for teachers to put more attention on these errors through remedial teaching.enArithmetic--Study and teaching (primary) Winam Division, Kisumu District, Kenya.Thesis