The Variations in the Levels of Performance of Learners among the Various English Language Speaking Skills Tested in both Written and Oral Approaches: Evidence from Secondary Schools in Eldoret Municipality

Abstract

It has been observed in Kenya that the mode and the scope of testing influence to a great extent the exposure of the students to a particular field of study. The use of written tests by KNEC in the KCSE English language examination of the speaking skill has influenced the teachers of English attitude towards the training and valid testing of the speaking skill in the classroom. Since the KIE English language syllabus recommends an oral evaluation at the end of secondary school English language course, the study sought to investigate the relationship between oral and written test scores in secondary schools in Eldoret Municipality, Kenya. Based on the study, this paper discusses the variations in the students' performance in the various speaking sub-skills tested in either mode of testing. The study was formulated and interpreted with reference to Communicative Language Teaching (CLT) approach. The study used correlation research design, which enabled the researcher to assess the degree of relationship between the scores attained from a written and an oral test of five sub-skills of the speaking skill. This assisted in establishing the adequacy of the written tests currently used in the assessment of oral as well as aural skills. Purposive, stratified and simple random sampling procedures were used to select the twelve schools and 360 Form Four students who participated in the study. The research instruments used were two achievement tests – one written and the other oral. Descriptive statistics such as mean and mode were used to compare the variables under study. Inferential statistics were also used. Pearson's product moment correlation was used to analyze the degree of relationship between the variables, t-test to establish significance of the difference between the mean scores and analysis of variance to determine between as well as within group variance.  The independent variable was speaking competence while the dependent variables were the speaking sub skills tested. These were rendition of an oral narrative, word stress, intonation, contrastive stress and conversation. The results of the study showed that variability in the students' performance in writing sub-skills as well as in the oral sub skills. This is an important finding for it implies that as many sub-skills as possible should be included in the tests and more particularly in the oral mode. It was, therefore, recommended that oral English test should be allocated a full paper on its own. Such a move would definitely put oral English in its rightful place and also provide for the necessary wash back effects.