| Reading is a major way of acquiring knowledge for FL learners. In China as well abroad, reading has always been the focus of language learning and teaching, so the assessment of a reader's reading ability is then one of the most important criteria to measure one's integrative language proficiency.Nowadays, in our country, in all kinds of English proficiency tests, the testing of reading comprehension is usually designed around such factors as examinees, testing purposes, testing materials, difficulty of the reading materials, reading speed and testing methods. The testing patterns has been following the same one used to test the reading competence of native speakers of English: the tests are always composed of several short passages of different length, difficulty, topics and styles, followed by sets of multiple-choice items aiming at the measuring of both the examinees vocabulary and reading comprehension. With just 3-5 short passages, the traditional reading tests often fail to measure the examinees' actual competence in understanding different reading materials and it may lead to test or item bias, that is to say, the context of the material may be weighted to favor some examinees over others, which may actually cause the problems of instability and unreliability of the testing results. la addition, the requirements of the traditional reading tests to the examinees' reading competence differ greatly from their real-life reading behaviors. Over the past years, the traditional reading tests are criticized in may aspects, such as lack of construct validity, failure to provide a variety of reading materials, lack of item independence, etc. Since the majority of reading and testing experts have not yet reached an agreement on reading assumptions and taxonomy of affective and cognitive skills, then the precise number and type of skills that are tested have been dependent on the test designer's choice of a model of the reading process.hi order to avoid the defects of the traditional reading tests, this paper validates an alternative testing approach梩he short-context technique,which is quite novel in testing field and up to now not much researches have been done about it.Through the use of brief stimulus materials, the short-context technique can be used to test the examinees' overall reading ability rather than the understanding of the specific details as do the traditional reading test. The short-context technique aims to test the examinees' control of basic reading skills and strategies instead of their intelligence of reference and reasoning. In order to validate the short-context technique, an experimental exam was given to the students of non-English majors who are of different English levels. By analyzing the data obtained in the test, conclusions are drawn in the following aspects: short-context technique can differentiate the reading proficiency of English learners of different levels, it is especially suitable to test the reading proficiency of intermediate and advanced English learners.The whole essay is divided into five parts:Part I. The importance of reading test is introduced, and the general idea of the essay is presented.Part II. The nature of reading is revealed; the three reading process models, i.e. bottom-up model, top-down-model and interactive model are analyzed. The core of the process model "schemata theory" is mainly discussed and its applications in constructing reading tests are put forward; the nature of reading ability is put forward too.Part III. On the basis of reading model theory, some theoretical issues in constructing tests of reading proficiency are presented, such as background knowledge, difficulty, testing objectives and testing methods, etc.Part IV. The dilemma of traditional reading tests is discussed and an alternative testing approach 梥hort-context technique is presented and its validation such as construct validity, content validity, concurrent validity and face validity are analyzed respectively. I do not mean to replace the traditional rea...
|
PDF Full Text Request