A Content Validity Study On Reading Comprehension Of NMET (2008～2013) Of Sichuan

NMET, as a large-scale, high-stakes, norm-referenced and standardized selective test, exerts great influence on the candidates and other stakeholders, making its quality a major issue to be concerned about. The quality of NMET is influenced by various factors such as validity, reliability, authenticity, backwash, practicality etc., among which validity is the most fundamental and central issue to be investigated.Content validity, as one facet of the unitary notion of the validity cake, is an inseparable part of the on-going validation process of any test. This thesis intends to have a content validity study on reading comprehension of NMET (2008-2013) of Sichuan to demonstrate the content features of the reading comprehension and explore the variations between them and discuss the reasons for the variations so as to provide some implications for test development in the process of professionalization and localization of language testing.This study assumes a dynamic and diachronic view on the content validation of reading comprehension (2008-2013) of Sichuan by the methodologies of text analysis, and statistic analysis etc. The framework of this study is a modified reading comprehension content validation framework based on the study of Construction Integration Model, Bachman and Palmer’s (1999) task characteristics and Weir’s (2010) socio-cognitive validation framework. In the process of the study, taking the English Teaching Syllabus (2000), English Curriculum Standards (2003), and Test Specification (2008) as the reference point, this paper gives an overview of the reading comprehension task characteristics of NMET (2008-2013) of Sichuan and attempts to explore the input features such as topic, genre, text length, reading speed, and readability etc., and the expected response of reading skills as well as the level of engagement to justify the content validity of the test.The research reveals that, theoretically, a dynamic view on content validity, i.e., taking context into account, is beneficial for us to better understand and account for the content validity variations over the years, while, practically, the reading comprehension of NMET (2008-2013) of Sichuan is in accordance with the basic requirements of the English Curriculum Standards (2003), Teaching Syllabus (2000), and Test Specification (2008), with fairly good overall content validity. The variations over the years largely reflect the changes of language test situation and the influence of the concept of the new curriculum reform.Nevertheless, there exist some variables that may contaminate the content validity. First, the coverage of topics is not wide enough to demonstrate its wide coverage especially in 2010. Secondly, the proportion of genre distribution doesn’t truly reflect the relative importance of each genre in senior high school or the intended language situation in that argumentation is belittled while narration may be overemphasized. Third, the range of readability scores covers such a wide range that it may go beyond the comprehensibility of the potential candidates, which may lower its actual content validity. And the sequence of passages is not always arranged in the order of increasing difficulty. In addition, the expected response of reading skills focuses too much on local understanding skills.