Investigating Middle School English Teachers’ Use of Conversational Texts in M-C Written Tests

Abstract

This study investigates how middle school English teachers use conversational texts in multiple-choice written tests. For that aim the study analyzed 45 middle school English tests collected from 15 schools in a metropolitan city. The focus of analysis was on the test items’ content-dependency, the kinds of knowledge measured and the kinds of tasks required by them. For additional information the study also conducted a questionnaire survey with 84 middle school teachers, focusing on the teachers’ perceptions and practices as test constructors. In addition, the study also compared the use of conversational texts in school tests and in authorized tests (i.e., the National English Achievement Test, TOEFL, TOEIC and IELTS) in the hope to obtain information about what to do to improve the quality of school tests. According to the result, quite a few test items made with conversational texts showed weaknesses in the focused areas. Based on the result, the study argues that to ensure the quality of middle school English tests with more valid test items, an item construction model as well as a more appropriate teacher training with opportunities for enough practice for item construction must be available.