英语阅读理解分析

The V alidity Analysis on Reading Comprehension Abstract:The present testing study aims to reveal the correlation between validity and reliability in order to examine whether such tests can measure text—level processing ability. However, the central purpose of this paper is to evaluate the validity of reading comprehension. The study is carried out with a sample of English college—entrance examination in order to determine whether the reading part be designed properly. The results indicate that, the design of reading comprehension needs improving to reveal learners’ real reading ability.Key Words: validity, reading comprehension, college—entrance examination1.IntroductionJ.B.Heaton states that the validity of a test is the extent to which it measures what it is supposed to measure and nothing else. Every test, whether it be a short, informal classroom test or a public examination, should be as valid as the constructor can make it. It is concerned with the question “How much of an individual’s test performance is due to the language abilities we want to measure?” V alidity is an integrated evaluative judgement of the degree to which empirical adequacy and appropriateness of intepretation and actions based on modes of assessment. As with reliability, the quality of validity is more complex than initially appears. Many English textbooks distinguish among a number of types of validity. According to J.B.Heaton, validity is divided into face validity, content validity, construct validity and empirical validity. If a test item looks right to other testers, teachers, moderators, and testees, it can be described as having at least face validity. Today, most designers of communicative tests regard face validity as the most important of all types of test validity. Content validity depends on a careful analysis of the language being tested and of the particular course objectives. The construct validity is capable of measuring certain specific characteristics in accordance with a theory of language behaviour and learning. The fourth type of validity, empirical validity is obtained as a result of comparing the results of the test with the results of some criterion measure. This paper, however, mainly discusses the first type—face validity with the sample of the reading comprehension of Guangdong Province’s college—entrance examination in 2009.2. MethodologyThis section explains the methodology employed to collect the related data from the reading comprehension of Guangdong Province’s college—entrance examination in 2009. The purpose of the study is to provide hypothesis—generating data for future research on analizing the validity of the reading part.1)The style—choosing analysis in reading comprehensionTable 1style title types of items the number ofwords in a test Passage A narration rescue multiple choice 358 Passage B narration an art multiple choice 359competitionPassage C narration handwrittennotesmultiple choice 348Passage D exposition the cover ofchildrens’ storybook Picture andinformationmatching5032)The pionts covered in reading comprehensionTable 2PassagesTestingabilityPassage A Passage B Passage C Passage D TotalDetailedcomprehension41. 43. 44 46. 48. 49 51. 52. 53 56--60 14Word/sentencededuce42 47 56--60 7Passage structureParagraphsubstancePassagesubstance56--60 5Author’sintention45 50 54. 55 43.FindingsAccording to the English college—entrance examination outline in 2009 which requires learners manage to read newspapers, magazines, books, advertisementm, exposition and notes, and to not only read between the lines but also read beyond the lines, we can safely indicate that there exists at least four shortcomings in the design of reading comprehension.1)The style is repeated and unitary with only two styles: narration and exposition. According to the English college—entrance examination spaper, magazines, books, advertisement, exposition and notes should be included as possible. Passage A, B, C all belongs to narration, while Passage D is exposition. Generally speaking, the senoir school learners are more familiar with the style of narration than other styles such as advertisment, newspapers, and notes, which will directly leads to lower the face validity. Besides, both of the authors in Passage B and C adopt the first personal pronoun to state the evens that are related to learners’ daily life, which also lowers the face validity.2) As for the items in the paper, the surficial skills and the indepth skills are misproportional. According to the outline in 2009, it requires learners to master the following reading skills: to identify the main idea in a text; to understand the concrete information in a text; to deduce the meaning of new words by contextual clues; to make a general judgement and inference; to identify the basic structure of a text and to recognize the author’s intention, viewpiont and attitude towards an event. However, from table two, we can see without difficulty that the surficial skills are overwhelm the indepth skills. The first item, detailed comprehension, occupies 70% of the total items, only 4 items belongs to indepth items (understanding the author’s intention), which occupies 20% of the total. Furthermore, the design of this paper neglects the synthetical reading skills. That is, none of the items belongs to passage structure or paragraph substance.Therefore, in spite of the four reading materials, there are few comprehensive systematic programmes which have been constructured from a detailed analysis of the skills required for efficient reading.3)The items presented in the test are so monotonous that it is difficult to ensure the face validity and increase the chance of guessing factor. Broadly speaking, although multiple—choice items are sometimes the most suitable instruments for testing reading comprehension, they should not be overused. The reason is that multiple—choice itself has limitations. First, multiple—choice lacks the language authenticity. Second, multiple—choice cannot directly tests learners’ real communicative ability. Third, multiple—choice may tend to mislead some weak learners to guess the answers. Frequently, other item types are far more interesting and useful. The text itself should always determine the types of questions which are constructed. Indeed, the same text will demand at least two or three different types of items. However, from table 1 we can conclude that each passage owns only one type of items. The three former passages are all presented as the ways of multiple—choice items without the types of completion items or the open—ended items, while the fourth passage belongs to picture and information matching item, which violates the principles of testing design.4. DiscussionFrom analyzing the reading materials of the Guangdong Province’s college—entrance examination in 2009, we can deduce two reasons of the existing problemes of the reading text. First, the professional quality standars and supervising system needs improving. That is, it lacks both the professional quality standars and supervising system while they drafting the provincial assignment policy on testing. Second, the number of professional testors is insufficient. The monotonous reading materials presented in the texts can support the piont. Being time limited, some testors choose reading materials just on the basis of their subjective judgement so that it is hard to form a scientific texting standard. Take the style of narration for example, the three former texts belongs to the same style which lacks the variety of text. And the learners’real reading ability is hard to distinguish as a result of decreasing the face validity. Another apparent shortcoming is that the types of items are too monotonous, only with two items: multiple—choice item and matching item. Therefore, this paper puts forward five suggestions to improve face validity from the macro and micro levels.1) To set up professional quality standars and supervising system.Professional quality standars and supervising system should be set up timely while they formulate the provincial testing system as to ensure the quality of testing and the face validity. Meanwhile, the difficulty level shoud be kept at the general level among each province in order to avoid over—difficult or over—simple text to appear.2) To choose and train professional testors.It is known that the sufficent professional testors are basic not only to ensure the quality of testing, but also to improve the face validity. The testors’ professional level is directly related to the quality of the tests. The more professional they are, the higher valid tests will be made, and the fewer shortcomings can be avoided.3)To identify the exact goal of reading comprehension while framing the test.To guarantee the college—entrance examination’s validity of reading comprehension, it is wise for testors to identify the different reading levels in a test. That is, a reading test should include the levels from simple level to difficult one. And the goals of testing are divided into learning about the text structure and catching the main information. Furthermore, the reading test should acquire learners to master the indepth items such as the writer’s attitude, the implicated meaning and the reading skills such as reading quickly,skimming and scanning.4)To design scientific test content of reading.Testors should take the test content into consideration as to ensure the face validity while designing texts. Firstly, the source of a text should be clear for it is the most significant factor to guarantee the face validity. Testors should identify whether the text is from a section of a fiction or the newspapers or other else. Secondly, the style of text can not be neglected. The basic principle is that monotonous style should avoided such as the above three texts are narration in Guangdong Province’s college—entrance examination in 2009. Meanwhile, the style is closely related to the difficulty of the reading test. That is, testors should choose various styles which can reflect different levels when they design texts. Thirdly, the item choice also plays an important part in the text. To improve the face validity, multiple—choice is not the only choice. Besides, testors should consider other types of items such as the matching tests, true/false reading tests, completion items or open—ended and miscellaneous items. Last but not least, the length of a text is an unavoidable factor to consider while choosing the text. Generally speaking, over the number of 500 in a text is considered to be too long. For long text will tend to cost learners’ time and energy, which is not scientific enough to ensure the validity.5)To emphasize the backwash effects of the testA good test of language should emphsize the backwash effects. The effect of tests on teaching and learning is known as backwash. Ethical language testing practice, as it is felt, should work to ensure positive backwash from tests. Certain backwash of testing on the one hand, it is fair to point out, on the other hand, testing has been one of the greatest single beneficial forces in changing the direction of language teaching in many areas and in encouraging the more responsive teachers to examine only their own teaching methods but also the language they are teaching. However, to guarantee the face validity of the reading test, testors should learn about the learners responce and the evaluation of the test. Questionnaire and investigation are the effective ways to get the concrete and related information. Through the backwash, testors can find which part is far from learners’real reading level and which skills are neglected. Being the positive backwash, face validy of a text can be improved, and the whole reading test can maximize its effect to reveal learners’ real reading ability.5. ConclutionMost of the present study mainly discusses how the two concepts (reliability and validity) are embodied in a test. The results of this analysis indicate the fact that lacking professional quality standars, supervising system and professional testors will lead to unpolished test. V alidity, as the most important standard, plays a significant part in testing a reading test. Whlie designing the reading test of Guangdong Province’s college—entrance examination, testors tend to neglect the effect of the choosen style, the importance of testing overal reading skills and the types of items, which all can lead to lower the face validity. Although from analyzing the above tables we can easily find out the prombles need solving, and put forward the related solution to deal with them, there are still some practical conc erns that must be taken into consideration in deciding whether a test is valid. We just take one sample of the reading comprehension of Guangdong Province’s college—entrance examination in 2009 to test the reading valid, which may lack reliability. We can do a further research on the face validity with more samples (eg: taking the resent ten—year Guangdong Province’s college—entrance examination papers for comparing to test the reading validity ) to provide a more convinced evidence.References:. J.B. Heaton Writing English Language Tests; Foreign Language Teaching and Research Press . Tim McNamara, 2003, Language Testing上海外语教育出版社. Bernard Spolsky, 1995, Measured Words, 上海外语教育出版社.Bechman,L.F.&Palmer,nguage Testing in Practice [M].Oxford University Press,1996.2009年高等学校招生全国统一考试大纲.2009 年普通高等学校招生全国统一考试（广东卷）---英语试卷。