This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
In the modern society, reading comprehension is becoming more and more important. There are few aspects of our lives that are not linked in one way or another to reading. As far as language acquisition is concerned, reading is the most important one in the four basic skills for the learners who learn English as a foreign language. Reading comprehension appears in most standard tests such as NMET (National Matriculation English Test) and is regarded as an important part in them. For most Chinese students, reading is also regarded as the most efficient way to learn English.
Testing is an indispensable part to the making of teaching. A successful test can not only inform the teachers and students of the progress the students have made but also diagnose weakness in teaching and learning, which contributes in turn to the development of teaching. As a matter of fact, the importance of testing in teaching has long been recognized and many researchers have been working on this field. However, problems have been found in the reading test in NMET. In this paper, the illustration of the concepts of reading, reading ability, the purpose and function of reading test as well as the validity and reliability will be first given as the theoretical background of the whole paper. The problems will then be examined followed by the possible solutions and suggestions to perfect the Chinese reading test.
2 Literature Review
To study the validity on English reading test in high school, some of the theories should first be defined. In this part, the definition of reading and reading ability, the purpose and function of reading test as well as the validity and reliability of reading test will be explained as the following.
2.1 Reading and Reading Ability
2.1.1 The definition of reading
According to Widdowson (1979), reading is the process of combining textual information with the information a reader brings to the text. In this view, instead of extracting information from the text, reading activates a range of knowledge in the reader's mind which, in turn, may be refined and extended by the new information supplied by the text. Just as Goodman (1975) has stated, reading is a receptive language process with an essential interaction between language of the writer and thought of the reader. In this case, reading is likely to be different for the same reader on the same text at a different time or with a different purpose.
2.1.2 The definition of reading ability
Many teachers, text writers and the constructors of language tests have pointed out that reading is made up of different skill and strategy components, across which students may exhibit differences in level of proficiency. Barrett (1968) distinguishes five skills which concern reading ability, literal comprehension, recognitions of the ideas in the text, inferential ability, evaluation and appreciation. Reading ability thus can be regarded as the ability of the reader to adjust and use reading skills/strategies according to different reading purposes so as to efficiently recognize, comprehend and/or elaborate written texts.
2.2 The Purpose and Function of Reading Test
2.2.1 The purpose of reading test
One function of test is the evaluation of student performance for purposes of comparison or selection. However, it should not be the sole purpose or even the chief purpose of testing in schools. In terms of the purpose of reading test, it should aims at testing students' ability of handing information and understanding discourse as well as students' reading skills and reading rate. It can also reveal what knowledge the students have mastered and what they haven't.
2.2.2 The function of reading test
In this case, the function of reading test can be a guider to the reading teaching practice both for teachers and students. According to the reading test, teachers can know what language the students have achieved and what they haven't; what reading skills students have mastered and what they still need to be trained, they can know what need to be taught to the students.
2.3 Validity and Reliability of Reading Test
2.3.1 The validity of reading test
According Henning (1987), validity in general refers to the appropriateness of a given test or any of its component parts as a measure what it is purported to measure. Any test then may be valid for some purposes, but not for others. Henning's definition allows for degrees of validity: tests are more or less valid for their purposes; validity is not an all-or-nothing matter (Alderson et al, 1995). This important point implies that judgment will have to be exercised when one decides on the relative validity of a test on the basis of evidence.
To determine the validity of a test, it is necessary to relate it to a certain criterion, which represents what the tests supposed to measure. Different types of validity demand different kinds of criteria. Generally, there are construct validity, face validity, and content validity. Construct validity is a process through which language testing can be put on a more rational and scientific footing. This type of validity assumes the existence of certain learning theories or constructs underlying the acquisition of abilities and skills (Heaton, 2000). While, face validity is concerned with the test's "surface credibility or public acceptability" (Alderson et al, 1995) and it generally involves the layman's (typically, the test takers, teachers, administrators, and test users) judgment of the test. Content validity concerns with whether or not the content of the test is sufficiently representative and comprehensive for the test to be a valid measure of what it is supposed to measure (Henning, 2001). The test should be so constructed as to contain a representative sample of the course, the relationship between the test items and the course objectives always being apparent (Heaton, 2000).
2.3.2 The reliability of reading test
Reliability is a necessary characteristic of any good test. It is a measure of accuracy, consistency, dependability, or fairness of scores resulting from a particular examination (Henning, 2001). A reliable test is a test that is consistent and dependable and can give the same results when it is given on different occasions or used by different people.
Two kinds of reliability are often discussed by people, test reliability and scorer reliability. Test reliability reflects the actual level of agreement between the results of one test with itself or with another test. Such agreement, ideally, would be the same if there were no measurement error, which may arise from bias of selection, from bias due to time of testing or from examiner bias. Meanwhile, scorer reliability is the consistency of scoring provided by two or more scorers. It can be affected by either the intra-rater reliability or inter-rater reliability. The former refers to the inconsistency within one scorer and the latter refers to the inconsistency between the scorers.
2.3.3 The relationship between validity and reliability of reading test
As we have seen, test validity and reliability constitute the two chief criteria for evaluating any test, whatever the theoretical assumptions underlying the test. The fundamental problem, however, lies in the conflict between reliability and validity. The ideal test should, of course, be both reliable and valid. However, the greater the reliability of a test, the less validity it usually has. Consequently, it is essential for test writers to devise a valid and reliable test.
3 The Validity Problems of Reading Test in High School in China
After discussing the theories of reading and reading test, in this part, the problems of reading test in high school will be analyzed, followed by the analysis of the reasons behind them. As testing is close to teaching, it is very important to design a valid reading test, because a good test can have an active backwash effect on teaching.
3. 1 Validity Problems
Nowadays, different types of English language testing are thriving in China. However, the dominant one which has the biggest quantity of candidates is still the National Matriculation English Test (NMET). The reading test part of it focuses on testing students' deep understanding ability such as making inference/ prediction, getting the main idea, understanding the standpoint and purpose of the author and understanding the discourse structure. The purpose is clear and definite however, the types of the items in reading comprehension are so limited-mainly using the Multiple Choice Questions (MCQ), which has some drawbacks in testing students' ability really and effectively. MCQ of reading tests are the post reading questions, and thus, the testers test reading comprehension as a result not as a process. In this case, the test takers become passive information recipients rather than active article interpreters, which does not correspond to the theory that emphasis on the reader's interaction. Besides, the MCQ cannot test students' generalizing ability effectively.
The impact of the test can be called backwash, and some experts say the backwash is an indivisible part of the validity. Hu (1996) says MCQ often has a negative backwash effect, and some teachers and students may focus on cultivating their receptive skills but neglect cultivating the productive skills, unaware of writing and speaking in English. Owe to the negative backwash of NMET, in some classroom test, some teachers also use MCQ as the main item to test students' reading ability. In order to make students get high marks, especially for grade three students in high schools, teachers mainly teach them the language knowledge and let them practice sample papers of NMET, and teach them pragmatic inferences more than propositional inferences sometimes, which make the teaching process become a test-oriented teaching and students' language competence is hard to be improved. Therefore, from the discussion above, we know that NMET as well as some classroom tests could not avoid its low validity, due to overusing the MCQs.
3.2 Reasons for the Problems
Various reasons have been analyzed by a lot of educational experts. The major reason for the problem can be interpreted as follows.
First and foremost, in order to keep the NMET has a high reliability, especially scorer reliability, the NMET designers use the MCQ as the main type of items in reading test with the purpose that they reckon marking of MCQ is not affected by the personal judgment of the test marker, which can make the NMET has higher scorer reliability. Therefore, the marking process can be more objectively. Besides that, it tends to be more efficiently for marking MCQs.
Apart from that, in order to make students to get high marks in the NMET, the teachers in high schools focus on teaching students to master some tips for figuring out answers and practice NMET samples as much as possible. Due to the reason that some teachers are lacking of test theories, it is hard for them to design a valid test paper according to the nature of the reading and the reading ability.
4 Designing a Valid Reading Test
After discussing the validity problems of NMET and some classroom reading tests, possible solutions to the problems and suggestions on how to design the reading test will be discussed in the following part. An example of the proper design of the reading test is provided in the appendix.
4.1 The Selection of reading materials
Before preparing questions, a very important task is choosing the reading materials. If the choice of reading material is not appropriate, it will bring difficulty to the proposition. Besides, it will also make the test lose reliability and validity. The reading comprehension test in teaching should meet the purposes and requirements of teaching in the certain semester (stage).
Choices of reading materials should be able to meet the requirements of the testing to achieve their aims. In addition to the theme, language lucid, compact structure, attention should also be paid to the following aspects.
To begin with, the types of reading materials should be emphasized. From one point of view, the authentic materials should be advocated. Authentic tests must contain passages that test takers will read in real-life situations, because the artificiality of the language may affect the generalize ability of the test score and thus the construct validity of the test (Bachman & Palmer, 1996). Consequently, the styles of reading materials should be similar to which the students read in class. In addition, the types of materials that students may have read and they need to read should be used. The passages can be taken from newspapers, magazines, journals, non-literary books, leaflets, brochures, etc. They are generally authentic but some editing may take place for clarification. They may be informative or descriptive or contain opinion and comment.
What comes next is the difficulty of reading materials. The length of the passages must be suitable for the test takers, considering their present level of language proficiency. While the excessive length of the passage may put the test takers under stress and increase their anxiety level, a brief passage may encourage frequent regressions and word-by-word approach to reading rather than test their skills of searching for information rapidly and efficiently. Therefore, a suitable passage should be one that contains sufficient information and provides the test takers with ample opportunities to apply the various reading skills that they have acquired. The difficulty of language and vocabulary should be appropriate. The requirements of students' vocabulary and grammatical knowledge should not go beyond the language ability of the students, because the purpose of reading test is to test students' ability of getting information through reading the written materials.
The novelty of the content of reading materials is another aspect that should be highly considered. The materials used must have new information, that is to say, it is unknown information to the students, avoid using the articles that student has read or are very similar with the articles they have read or the answers can be known from general knowledge.
The last point I intend to mention here is the suitable subject matter, which means that the extracts of general topical content are better choices than those of specialized content. If the subject matter of the passage is familiar to some test takers but not to others, the test is thought to be biased. The knowledge of reading materials should be appropriate. The knowledge that reading material contained should comply with students' knowledge structure level, not too specialized. Requiring students to have the background knowledge is undoubtedly be inappropriate to raise the difficulty questions.
4.2 Items Selection
In the reading comprehension tests, mainly five formats are used: Multiple Choice Questions (MCQ), Short Answer Questions, True or False (T/F), Translation, Summary. These formats can be classified, according to how they are marked, into two types: one is objective, the other is subjective. MCQ and T/F belong to the former, the others to the latter. Many reading practitioners prefer the objective test formats for their high marking reliability, while many others argue that it is the high validity which subjective test formats enjoy that is the most essential.
Subjective tests are generally believed to be more valid tests. Most educators would accept that subjective tests would have more validity than objective tests. Teachers have access to some scores of subjective tests from their students, and they know what the scores really mean. If the students bring them the scores of objective tests and ask the teachers questions about what the scores mean or ask for advice about how they can improve their scores, perhaps teachers can do nothing. On the contrary, if students bring the scores of subjective tests and ask the same question, teachers at least know where they can get clues about how well he or she learns because the scores are sustained by most obvious evidence.
In order to improve senior middle school students' reading ability, subjective items should be increased in the reading test, because it can test students' actual reading ability effectively. Teachers can propose some questions after students reading the materials, but the answers expected given by the students should not be too long. In this way, reducing the words students need to write according to the students' actual level can make score to be more objective.
Although MCQ are sometimes the most suitable instruments for testing reading comprehension, they should not be over used. Frequently, other item types are more interesting and useful. Sometimes the same text will demand at least two or three different types of items. (Heaton, 2000)
4.3 Other attentions should be paid
Besides what have been mentioned above, there are some other aspects that teachers should take into consideration.
Teachers are supposed to understand the students' verbal ability, reading Intelligent and reading skills roundly, with fully consideration of what students know, what they do not know as well as what they should review. Teachers should also be clear about if the content of the test has beyond students' knowledge structure and the level of knowledge in this area. The more information the teachers have about their students, the easier it is for them to design the test. Then the questions will not be so vacuous that they cannot reflect the true level of students' reading ability objectively.
Second, deictic word should been written in detail before each item so as to inform students of what to do, how to do it clearly and accurately. Meanwhile, it is important for teachers to make sure that the correct answer to the test question is not the original sentence of the reading material. If so, it will be difficult to test the actual reading ability of the candidates and it will give the teaching practice a wrong guidance with decrease of the difficulty of questions.
Last but not least, interferential items should have some interference with the conformation to the rules of grammar, the length and the structure of the correct answers. In other words, these items should be maintained similar to the answers so as to avoid inadvertently divulge clues, otherwise, some options would lose effectiveness.
All the above are the suggested solutions to the problems in Chinese High School reading test. One possible version of the revised reading test in NMET is provided. (see Appendix)
This paper mainly discusses the concept of reading, reading ability, validity, reliability and the function of reading test. The problems in NMET reading test in China are analyzed. Because of the over-use of the MCQ in reading test which could hardly test the reading ability that senior middle school students should master required by high school English curriculum, the author points out that teachers should increase the subjective questions in the reading test, so as to achieve a high validity. Then the author puts forward her own ideas about how to design the reading test from the choice of reading materials, items selection and the attentions should be paid when design the test paper. Reviewing this paper, due to some objective restrictions, some limitations also explored, such as lacking of research method and examples, which would make the article being a little bit hard to convince readers. In the future research, the author was going to put more emphasis on balancing the validity and the reliability of reading test which is designed for a large number of candidates.