ㆍ발행기관 : 글로벌영어교육학회(구 호남영어교육학회)ㆍ수록지정보 : STUDIES IN ENGLISH EDUCATION / 21권 / 2호 ㆍ저자명 : 장소영 ( Jang So Young ) , 김태영 ( Kim Taeyoung )
The main purpose of this study is to analyze the effects of the nationwide standardized listening tests for the middle school students in Korea. In this study, the validity of the test is evaluated by adopting the Rasch measurement model. Responses of 60 students on 20 listening test items were collected and analyzed. The results reported empirical evidence step by step on the basis of the validation procedures Wolfe and Smith suggested. The findings are as follows. First, Cronbach alpha coefficient is of .81 showing that this test is reliable, although some misfit test items were found. Second, judging from the item-person map, the test does not have difficult items enough to properly evaluate high level students. Finally, suggestions for teaching and learning are discussed. The Rasch analysis provides a simple and useful map which can be used as a way of giving proper feedback to the students. It shows the test items distributed in six dimensions, indicating that the correct or incorrect items are either above, within, or below the ability level of a student as well as the overall evaluation of test validity. It is expected to make it possible for a language teacher to carry out the measurement-driven instruction Cheng suggested by linking test results and classroom practices.