Cont Ent Validity Study On Reading Comprehension Tests Of NMET Of Chongqing

Being the biggest test in China, the annual National College Entrance Examination (NCEE) has a huge social influence. It is made up of five large-scale, high-stakes tests, among which there is National Matriculation English Test (NMET). As a pioneer of all standardized tests in China, NMET is always promoting the scientization of NCEE. It has been 6 years since Chongqing began the work of test design of NMET in 2004. In recent years, NMET of Chongqing attracts more and more attention. However, none scientific evaluation has ever made concerning the test and no systematic and comprehensive research has ever conducted to guide the test design.Reading, in particular, is one of the most important abilities for foreign language learners. It is an important means not only for knowledge acquisition, but also for enjoyment. Reading has always been the focus of English classrooms in Chinese middle schools. Accordingly, reading comprehension test is deemed as the most important part in NMET. Also, the discrimination of NMET depends on the validity of reading comprehension test to a large extent. We, therefore, can conclude that a reading comprehension test of good quality will ensure the success of NMET.We will study the reading comprehension tests of NMET of Chongqing from 2004 to 2009. The aim of our study is to assess these reading comprehension tests scientifically, and to encourage their improvement. By adopting a combination of qualitative and quantitative approaches, we evaluate the content validity of all the reading comprehension tests of NMET of Chongqing. Thus, some practicable suggestions for designing reading test are made hereinafter.To start with, we establish a framework of test characteristics for studying reading comprehension test. It is based on both the Bachman's framework of language task characteristics and the reading comprehension test of NMET of Chongqing. Then, combined with the nature of reading, an investigation is made to explore how the content of the reading comprehension section reflects the corresponding requirements laid down in the testing syllabus and teaching syllabus. The content validity of reading comprehension test of NMET of Chongqing is inspected and verified from five aspects:the test setting, the test rubrics, the input, the expected response, as well as the relationship between input and response. Finally, we propose some principles and methods in order to regulate the content validity of reading comprehension test. They will also help us optimize the design of reading comprehension test and improve its quality.The conclusions reached in the aforementioned study are as follows:First, the reading comprehension tests of NMET of Chongqing in the recent years have met the requirements in both testing syllabus and teaching syllabus. These tests are well designed. They cover all the knowledge and reading skills which must be tested. That is to say, the reading comprehension tests have reached a high content validity. Besides, from 2004 to 2009, the control of the discourse variables in a test is better and better. These tests are stable and continuous; their qualities are improved year by year.Second, through factor analysis and validity verification, we find a series of variables that affects test validity. Then, we identify their characteristics to ensure a good content validity of reading comprehension test. These variables are widely distributed in the test setting, the test rubrics, the input, the expected response, as well as the relationship between input and response. Because they all have an effect on the content validity of a reading comprehension test, it is reasonable and feasible for us to acquire good validity by regulating these variables. Meanwhile, the principles and methods of controlling the content validity can be applied not only to the reading comprehension test of NMET of Chongqing, but also to other reading tests of the similar level or type.Third, the study has also provided us with an inspiration for improving the teaching and testing of English reading. For example, the question format in NMET of Chongqing is limited, multiple-choice questions has been the only test form for six years; some weakness revealed in test entails better test design; the criteria—syllabi for test design should be updated; the backwash effects of the test on teaching should be valued; test-takers should increase their reading amount, and teachers should enhance the teaching of analytical reading; In addition, the gist of new curriculum should be further integrated into reading teaching and testing.