An Evaluation of English Language Tests: A Case Study of Nyamira District

Thumbnail Image
Mecha, Olpher
Journal Title
Journal ISSN
Volume Title
Kenyatta University
The present study addressed the issue of whether teachers give all the types of tests. It also sought to establish the reliability of the end term tests that were given in first term and second term in the year 2002 in Nyamira district. The study also did a comparison between the teacher made end term tests and the KCSE national examination paper 2 at the content level with a view to establish whether all the examined areas are tested. A descriptive research design was used for this study whereby data was drawn from 12 randomly selected schools in Nyamira district. The 200I KCSE exam paper was used for comparison. Data was analysed using a conceptual theoretical framework drawn from the psychometric structuralist theory and the communicative theory. It was then analysed and yielded the following results. The study reveals that the teachers do not give all the types of tests that exist. While this is not absolutely necessary, failure to give these tests would lead to low proficiency in English. The data that is analysed also shows that the tests that are given are reliable, as only 2 of all the observed tests were unreliable. Finally the study reveals that most of the end term teacher made tests did not test all the areas that are tested in the KCSE national examination English paper 2. The study recommends that teachers should try and give all the types of tests in the classroom to enhance language proficiency. The teachers should also test all the areas that are examined in the national examination to enable students acquire the necessary skills to tackle this paper. More research is however needed in the area of reliability for concrete conclusions. Teacher training institutions should offer in-service courses to update the teachers on new developments in language testing
Department of English & Linguistics, 165pg. September 2005, PE 1068.K4 O4