| National Matriculation Test (NMT) is one of the most influential tests in our country,the scientific nature of which has been widely concerned. Since 2004 Jiangsu province has autonomously designed its own provincial matriculation tests while more and more provinces have adopted the national version. With the reform of new curriculum English is always at the forefront of NMT reform and thus the quality of the Jiangsu matriculation test (JSMT) has become the focus of people. Validity is one of the most important criteria to measure a test. Therefore, this thesis will analyze and compare the validity of JSMT and NMET(I) in order to offer suggestions for designing JSMET and pedagogic implications for the English teachers in senior high schools.The study selects the same items in 2014-2016 Jiangsu matriculation English test(JSMET) and NMET(I) including reading comprehension tests, cloze tests and writing tests, and uses the qualitative and quantitative methods to compare the validity of the two MET versions. First, the thesis analyzes and compares the face validity and content validity of 2014-2016 JSMET and NMET(I) based on Bachman and Palmer's task characteristics framework, the test specifications of the two MET versions and English Curriculum Standards for Senior High Schools. Second, the thesis analyzes and compares the construct validity of the two MET versions based on Bachman's communicative testing theory.Finally, an interview survey of 4 experienced English teachers in senior 3 from different schools is conducted for the suggestions of English teaching and the strategies for JSMET.The main research questions are: (1) How is the face validity of 2014-2016 JSMET and NMET(I)? What are the differences between them? (2) How is the content validity? What are the differences between them? (3) How is the construct validity? What are the differences between them? (4) What are the washback effects of JSMET? What teaching strategies can be proposed for English teaching based on the comparative study and interview?On the basis of comparative analysis, the findings are listed as follows: (1) The face validity of 2014-2016 JSMET is lower than that of 2014-2016 NMET(I) as to items design,and the face validity of reading comprehension tests and cloze tests (objective items) is lower than that of writing tests (subjective items) in both two MET versions. (2) The content validity of JSMET and NMET(I) presents their own characteristics. On the whole,the content validity of the two versions is rather high. Compared with NMET(I), JSMET reading comprehension tests have characteristics of longer length of passages,higher demand on students'reading speed and relatively lower readability of passages,more diverse genres and topics,more balanced distribution and a wider range of tested reading skills; JSMET cloze tests have features of longer length of passages, higher demand on students' reading speed and unbalanced distribution of vocabulary type; JSMET writing tests have a higher demand on students' writing speed, and they offer more appropriate and various context and information. (3) The construct validity of the two versions is high for grammar competence, textual competence, pragmatic competence and strategic competence are all checked. However, the demand on students' communicative language competence in JSMET is higher than that in NMET(I). (4) The difficulty of JSMET will increase the burden of teachers and students, and also lead to more difficult simulated tests.However, some items in JSMET can lead students to be better English learners and users.In addition,English teaching in senior high schools will focus more on integrated teaching strategies and language use in communicative contexts. |
PDF Full Text Request