The Strategies In Responding To The Reading Comprehension Subtest Of The TEM 4

The Test for English Majors, Band 4 (TEM 4) has significant impact on stakeholders. Therefore, the validation of the TTEM 4 is a widespread concern in academic circles. With the development of relevant research, validation of the TTEM 4 in terms of test-taking strategies began to be unfolding since the turn of the 21st century. It became urgent since the reform of TEM 4 in 2004. Responding to the need, this thesis focuses specifically on the response strategies or test-taking strategies of the Reading Comprehension Subtest (RCS) of the TEM 4.The interest to test-taking strategies began since the late 1970s for test validation purpose. In the recent 20 years, three aspects of this line of study have been emphasized: test-taking strategies research for test validation purposes, strategy instruction for performance on high-stakes standardized tests and language proficiency related to test-taking strategies. In the international academic circle, this line of research is moderately increasing. In the Chinese academic circle, however, the empirical research is just in the beginning.The thesis, based on the previous domestic and international research, covers all the three above-mentioned aspects of test-taking strategies. The study attempts to address the following questions:1. Does the RCS of TEM 4 test what it intends to test in terms of response validity?2. What are the perceived frequencies of the use of the test-taking strategies in question?3. Does test-taking strategies instruction have a significantly beneficial effect on reading performance?4. Do high-achievers and low-achievers benefit from response strategy instruction equally?The first question involves the first aspect of this line of research - test validation. The second question is the extension of the first question. The third question is in reference to the second aspect - strategy instruction for performance on standardized tests. The fourth question is relevant to the third aspect - language proficiency related to test-taking strategies. The author accumulated the teaching experiences of response strategies of TEM 4 in advance to increase the validity of the experiment. A pilot study with six volunteers was carried out before the main study in order to ensure the reliability of the two response strategy questionnaires, which were piloted and refined. The effective sample of the main study contains 53 participants. Among them, there are 30 participants in an experimental group and 23 participants in a comparison group.The thesis adopts an organic combination of research methods: a survey study and an experimental study. The specific instruments are:1) Test battery. A pre-test and a post-test were administrated to all the subjects between the treatments for the teaching experiment. The pre-test is the RCS of TEM 4 in the year of 2008. The post-test is the sampled RCS of TEM 4 from 1998 to 2000. They equivalent forms that have the same rubrics and all meet the revised version of the Specification of RCS of TEM 4.2) Questionnaires, including Strategy Inventory of Language Learning (SILL), Style Analysis Survey (SAS), Questionnaire of Reading Strategy in the Testing Situation (QRSTS) and Questionnaire of Test-taking Strategy (QTTS). SILL and SAS were administrated to all the respondents in order to rule out learning strategy and learning styles as the intervening variables. The QRSTS and the QTTS were both administrated to the experimental group, but only the former response strategy questionnaire to the comparison group.3) Experimental instruction. The QRSTS and the QTTS are also instructional materials for the instruction of reading strategy in the testing situation and the instruction of test-taking strategy respectively. The two instructions were all administrated to the experimental group, but only the former instruction to the comparison group.4) Compositions. Each respondent are required to write an composition about their attitudes towards TEM 4, including test-taking anxiety, test-taking anxiety motivation and effort for preparing the test, in the exit test of the course. The author read, analyzed and compared the three affective variables in the compositions of the experimental group and the comparison group.Data were acquired from the above instruments. The quantitative data were analyzed through SPSS 16.0. The results show that learning strategy, learning style, anxiety, motivation and effort are ruled out as intervening variables. Thus, the teaching experiment is reasonably valid. Four conclusions can be drawn from the results in order of research questions:Firstly, the RCS of TEM 4 relatively successfully tests what it intends to test, though the frequencies of the reading strategies in the testing situations vary considerably.Secondly, all the test-taking strategies in the QTTS have 38%～73% probabilities, which are considered as regularly or frequently used strategies. The most popular test-taking strategies are going back to the text for clarification, making an educated guess, going back to the question for clarification, etc. These results conform to previous studies of test-taking strategies.Thirdly, it seems that test-taking strategy instruction does not make a statistically significant difference on the progress of reading performance.Fourthly, other things being equal, low-achievers tend to benefit more from the strategy instruction than high-achievers. The difference is statistically significant. The result, however, contradicts previous finding that the instruction of response strategies may not help low-achievers because of their limited proficiency of L2.The major theoretical contributions of the thesis are confirming that all the test-taking strategies that were selected from the previous studies are indeed frequently used; it also proposes that low-achievers tend to benefit more from the strategy instruction than high-achievers, which contradicts a previous finding that the instruction of response strategies may not help low-achievers because of their limited proficiency of L2. Its major methodological contribution is the attempt to develop composition as an instrument to measure the respondents' affective variables (anxiety, motivation and effort) and use Cluster Analysis of strategy frequencies through SPSS to distinguish levels of the frequencies. Its practical contribution is the validation the RCS of TEM 4 in term of response validity. The thesis has valuable reference to the research of test-taking strategies and response validity. It, however, has its own limitations. There is a large room for improvement in it. The author proposed suggestions for future studies.