| The Test for English Majors Grade 8 (TEM-8) is a large-scale nationwide criterion-referenced test for English majors in China. Since its inception, TEM-8 has undergone several modifications. The 2005 version of TEM-8 is the first of its kind designed in line with The 2000 Teaching Syllabus for TEM-8 and The 2004 Test Syllabus for TEM-8. For English majors, the ability to read efficiently is of vital importance. Therefore, this thesis was intended to explore the content validity of reading comprehension (RC) of TEM-8 from 2005 to 2007 in terms of the new framework of task characteristics revised from the Bachman and Palmer's framework (1999: 49).There are two research objectives in this research. The first is to explore the content validity of RC of TEM-8 from 2005 to 2007 in terms of the characteristics of input mainly based on The 2000 Teaching Syllabus for TEM-8 and The 2004 Test Syllabus for TEM-8. The characteristics of input include total length, difficulty variables, topics and genres. The second is to explore the content validity of RC of TEM-8 from 2005-2007 in terms of the characteristic of expected response based on these two syllabuses. The expected response involves the coverage of reading skills. The data collected in the study include TEM-8 papers from 2005 to 2007, The 2000 Teaching Syllabus for TEM-8, The 2004 Test Syllabus for TEM-8 and the nationwide means scores of RC of TEM-8 from 2005 to 2007. Data analysis involves three steps: counting, calculation and comparison.The major findings were summarized as follows:In general, on the basis of The 2000 Teaching Syllabus for TEM-8 and The 2004 Test Syllabus for TEM-8, the RC of TEM-8 from 2005 to 2007 is a valid reading test by the content analysis.(1) The characteristic of input of RC of TEM-8 concerning total length in each paper is consistent with requirements set by the syllabuses for TEM-8. And the total word number of passages and questions tends to become larger and larger within a reasonable range.(2) The difficulty variables as the characteristic of input of RC of TEM-8 can well reflect the characteristics of TEM-8 and the requirements set by syllabuses for TEM-8. Firstly, the layout of RC concerning average sentence length in these three years is under a comparatively balanced control, which can ensure that testees won't find it abrupt to process the passage sentence by sentence. Secondly, distinctive as each paper may appear, a thorough look suggests that three papers turn out to be very consistent with regard to the distribution of new information. The proportion of content words to the total words doesn't vary much from year to year and ranges from 50% to 60%, neither too compact nor too diffuse. Thirdly, the difficult level of RC in the three papers is proved to set appropriately in general, consistent with the requirement by the syllabuses and the readability of most passages is within the scale from 60 to 30, which corresponds to"fairly difficult"and"difficult". Fourthly, the sentence types and the subordinate clauses have moderate consistency in the three years, which can be inferred that the difficulty of the syntax is under a reasonable control.(3) Topics, as the characteristic of input of RC of TEM-8 are distributed equally in three papers and cover different fields. And variety of topics will guide students to read widely and extensively in the reading training.(4) The characteristic of input of RC of TEM-8 concerning genres is basically consistent with the requirements set by syllabuses. However, the genres such as advertisements, directions, and charts and graphs absent in these three papers need to be considered for the future test design.(5) These five question types as a way to manifest the coverage of reading skills as the characteristic of expected response of RC of TEM-8 are all distributed at different weight in each paper.Three implications are generated for the improvement of the test design and the language instruction. Firstly, the constructed response as well as the selected response should be adopted in RC of TEM-8 in the future with an attempt to fulfill the requirements of The 2000 Teaching Syllabus. Secondly, the genres such as advertisements, instruction books, and charts and graphs absent in these three papers need to be considered for the test design. Thirdly, apart from the textbooks, students should be encouraged to read more editorials and book reviews in newspapers and magazines of America or Britain, and historical biographies and literary works to a degree of certain difficulty published by English speaking countries, which can activate their background knowledge. In the process of reading, they should pay attention to their reading speed and finish some corresponding tasks.However, there are three major limitations in this research. Firstly, the lack of testees'performance scores results in the impossibility of calculating F.V. of each question. Secondly, this newly revised framework of task characteristics has some deficiencies. Some characteristics of input such as cohesion devices, figurative speeches and the key words in the passages for full comprehension were not investigated since they posed much difficulty on the measurement. Thirdly, only the content validity of RC of TEM-8 is involved in this research, which results in an incomplete validation study of TEM-8. |
PDF Full Text Request