| Under the guidance of the new curriculum, cities in Jiangsu Province made some adjustments in the content, format, and item types in Senior Entrance English test in 2007. There are two biggest and common ones in those adjustments. First, they set the learning outcomes or objectives of the 5th level curriculum standards as the measurement objective in the English exam. Second, they deleted some kinds of questions that examine the micro-skills and increased the number of questions that have real applying situations, such as task-based reading (including short answer questions after reading and form completing questions after reading) and blank filling with words from text (blank filling according to first letter). Since the applying of this new test mode, the monitoring and assessing of its quality have been mainly based on the conclusions of proposition, moderation and other test jobs by each city and the self-assessment of papers. They have to admit that they still lack effective monitoring and systematic research of the new test mode's scientificity,rationality standard ability and guidance quality.This essay mainly contains validation study of the new types of reading questions in senior entrance English test after adjusting, by assessing the impact of different testing methods and compositions on reading results under the unified validation framework, taking the questions of the 2009's Senior Entrance English Test in Yancheng as an example. Three general research questions are generated:1) whether the reading abilities of senior entrance English test can be divided into three levels, namely literal comprehension skills, inference skills and assessing skills; 2) whether the construct of reading in the test is well tagged and those questions can reflect our testing objectives; 3) whether the newly added types of reading questions can better reflect the real reading abilities of different students.My research on the components of reading in the senior entrance English test is mainly based on principal component analysis and principal axis factoring. The logical analysis and content collecting of the construct of reading is mainly based on expert judgement, followed by the research of exam takers'mental process using the introspective method to show the steps and methods exam takers take to get right answers. Finally, I use Pearson method to calculate the relationship of test scores of different groups that have different language skills on different types of questions in order to validate the effects of the new types of questions.The preliminary result of my research is that reading abilities of senior entrance English test can be divided into at least three levels:literal comprehension skills, inference skills and assessing skills. The esearcher prefers using three factors because they represent basic data structure and can better reflect the substantive characteristics of certain information. By expert judgement, introspective method and relationship analysis, I conclude that the construct of reading in the test is well tagged and it can show the latest development of reading theory. However, the strength of association of the construction is not strong enough. The new types of reading questions don't have a high representativeness. Task-based reading has differences with real reading process. Testing method has impact on testing result, which makes test scores unable to assess the reading comprehension abilities of test takers scientificly.In conclusion, no matter what kind of method we take, quantitative or qualitative, we can draw a conclusion from these collected evidences that new reading questions in Yancheng's Senior Entrance English Test do not have an obvious effect. Simply changing the test item types may not necessarily bring about the positive washback to meet the high expectations for the test reform. Instead, we must enhance the scientific proposition, rigor and professionalism in writing a test in order to achieve the intended validity of a reading test. |
PDF Full Text Request