《A Study on the Quality of Achievement Test Paper of Junior English 英语专业毕业论文.doc》由会员分享，可在线阅读，更多相关《A Study on the Quality of Achievement Test Paper of Junior English 英语专业毕业论文.doc（9页珍藏版）》请在三一文库上搜索。
1、A Study on the Quality of Achievement Test Paper of Junior English 内容： Abstract: Based on the theory of Bachman Multiple Choice Question, which occupies 15% of the total score; Cloze, 10%; Reading Comprehension, 30%; Task-based Reading, 10%; and the last part, writing, which occupies 15% of the tota
2、l score. 2.2. Reliability Estimate Reliability is the extent to which test scores are consistent: a quality assessing whether candidates can get the same results on different dates (assuming there is no change in their ability). Reliability is a property of the test as a measuring instrument, but is
3、 also relative to the candidates taking the test: a test may be reliable with one population, but not with another. And next the author will make estimation about this test paper utilizing the instrument-SPSS. The following is the descriptive statistics of the results of the terminal junior test pap
4、er. Graph1 统计量 LC MCQ Cloze RC TBR Writing TS 有效 40 40 40 40 40 40 40N 缺失 0 0 0 0 0 0 0 均值 13.90 9.95 4.30 25.00 4.50 5.88 63.50 中值 14.00 9.50 4.00 28.00 5.00 6.00 66.50 众数 17 9 3 28 6 8 67 标准差 3.477 2.864 2.289 6.421 3.226 3.618 14.915 方差 12.092 8.203 5.241 41.231 10.410 13.087 222.462 偏度 -.605 -.7
5、12 .164 -1.840 -.005 .045 -.916 偏度的标准误 .374 .374 .374 .374 .374 .374 .374 峰度 -.532 .302 -1.105 2.720 -1.203 -.903 .728 峰度的标准误 .733 .733 .733 .733 .733 .733 .733 极小值 6 3 1 6 0 0 24 极大值 19 15 8 30 10 13 87 Where: LC=Listening Comprehension, MCQ=Multiple Choice Question, RC=Reading Comprehension, TBR=T
6、ask-based Reading, TS=Total Score. And the following one is the normal distribution of the students total scores: Graph2 From the value of skewness and kurtosis, and the normal distribution of the total score, it can be discovered that the test scores are normally distributed. From the graph, it can
7、 be seen that the skewness is -0.916 and kurtosis 0.728, according to the English language testing theory, values for skewness and kurtosis of between -2 and+2 indicate a reasonable normal distribution. That is to say, the data of this paper is reliable. And the standard deviation of this test paper
8、 is 14.915, which indicates that the discrimination of students scores is big. Therefore, we can say that the test results are reliable and can reflect test takers language ability and are also the basis for guarantee the whole test papers validity. However, the mean of the fourth part (Reading Comp
9、rehension) is 25point, whose total score is 30 point, and the highest score is 30 point, the lowest one is 6 point. It is because most of the students are familiar with the material of the selected articles in the fourth part. The kurtosis of RC is 2.720, it is a positive kurtosis and indicates that
10、 this part is a little easy, and the discrimination of students score is small. The detail analysis of this part will be analyzed later. 2.3. Validity Analysis Briefly, 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 i
11、s a short, informal classroom test or a public examination, should be as valid as the constructor can make it (Heaton, 2000). There are several measures examining validity of a test such as face validity, content validity, concurrent validity, predicative validity, response validity and construct va
12、lidity. The following graph shows the correlation coefficients among different parts as well as between each part and the total score of this test paper. Graph3 相关性 LC MCQ Cloze RC TBR Writing TS Pearson 相关性 1 .453* .291 .485* .644* .476* .829* 显著性（双侧） .003 .069 .002 .000 .002 .000 LC N 40 40 40 40
13、40 40 40 Pearson 相关性 .453* 1 .284 .346* .158 -.105 .497* 显著性（双侧） .003 .076 .029 .330 .521 .001 MCQ N 40 40 40 40 40 40 40 Pearson 相关性 .291 .284 1 .391* .181 -.020 .475* 显著性（双侧） .069 .076 .013 .265 .902 .002 Cloze N 40 40 40 40 40 40 40 Pearson 相关性 .485* .346* .391* 1 .248 .363* .813* 显著性（双侧） .002 .0
14、29 .013 .124 .021 .000 RC N 40 40 40 40 40 40 40 Pearson 相关性 .644* .158 .181 .248 1 .515* .656* 显著性（双侧） .000 .330 .265 .124 .001 .000 TBR N 40 40 40 40 40 40 40 Pearson 相关性 .476* -.105 -.020 .363* .515* 1 .600* 显著性（双侧） .002 .521 .902 .021 .001 .000 Writing N 40 40 40 40 40 40 40 Pearson 相关性 .829* .4
15、97* .475* .813* .656* .600* 1TS 显著性（双侧） .000 .001 .002 .000 .000 .000 N 40 40 40 40 40 40 40 *. 在 .01 水平（双侧）上显著相关。 *. 在 0.05 水平（双侧）上显著相关。 According to Ferguson (2006), the correlation coefficients between each part and total score should be higher than 0.7 and the correlation between parts should be
16、 0.30.6. The above graph shows that the part-total correlation coefficients of the test paper is not very satisfactory, only the correlation coefficients between reading comprehension or listening comprehension and the total score are above 0.7. What more, the correlation coefficients between parts
17、is not satisfied with the standard of 0.3-0.6, which indicates that the construct validity of this test paper is very low. 3. Detail Analysis about Reading Comprehension This part consists of three passages, the first passage is just an introduction about the American Museum of National History, the
18、 second one is a brief introduction about Tea, and the third one is about the celebrity Li Shizhen. The total score of this part is 30 point, which occupies 30% of the total score. 3.1. Data Collection In this paper, the author has selected 40 students as its object of study, which can guarantee the
19、 validity of the data in order that the further analysis can be carried out on the basis of these dates. Graph4 RC 频率 百分比 有效百分比 累积百分比 6 2 5.0 5.0 5.0 10 1 2.5 2.5 7.5 16 3 7.5 7.5 15.0 20 2 5.0 5.0 20.0 22 1 2.5 2.5 22.5 24 1 2.5 2.5 25.0 26 7 17.5 17.5 42.5 有效 28 14 35.0 35.0 77.5 30 9 22.5 22.5 10
20、0.0 合计 40 100.0 100.0 3.2. Reliability Analysis The following graph is the descriptive statistics of the results of the part of reading comprehension. Graph5 统计量 RC 有效 40N 缺失 0 均值 25.00 中值 28.00 众数 28 标准差 6.421 方差 41.231 偏度 -1.840 偏度的标准误 .374 峰度 2.720 峰度的标准误 .733 极小值 6 极大值 30 From the above graph, i
21、t can be easily seen that, the medium, mode and mean is 28.00, 28, and 25.00. According to the above introduction, it is known to us all that the total score of reading comprehension is 30 point, therefore, the medium, mode and mean is very high. The skewnes of this part is -1.840 and the kurtosis 2
22、.720. However, only if the values of skewness and kurtosis are between -2 and +2, which can indicate a normal distribution. Whats more, the standard deviation is 6.421 indicating that the discrimination of students score is low. And it can also be inferred from the following graph6- the normal distr
23、ibution of reading comprehension. Graph6 There are four variants of normal distribution, in terms of skewness, it can be classified into positively skewed distribution and negatively skewed distribution. The former one indicates that there are more scores below the mean and the test or the part is d
24、ifficult, the latter one indicates that there are more scores above the mean and the test or the part is easy. And another two types are classified in terms of kurtosis, it can be classified into negative kurtosis and positive kurtosis. The former one indicates that the discrimination of students sc
25、ore is big while the latter one indicates that the discrimination of students score is small. From the graph4, graph5 and graph6, it is clearly seen that 75% of the scores of RC part is above the mean 25.00, the skewness is -1.840, approximating to -2 and the normal distribution is a negatively skew
26、ed one. All the evidences indicate that the Reading Comprehension part is easy; the discrimination of students score in this part is small. 3.3. The Detail Analysis of Reading Comprehension This part consists of three passages. Lets take one of the passages for example: Passage A The American Museum
27、 of National History Address: Central Park Westside 70th St., New York Telephone number: 212222-3500 Traffic: You can get there by bus, subway or car. Parking: There is a small parking lot (停车场). You have to pay $7.75 per day, but it is often full. Parking in the neighborhood is also limited (有限的).
28、Open hours: 10:00am - 5:45pm (Sunday, Monday, and Tuesday) 10:00am - 9:00 pm (Wednesday, Thursday, Friday and Saturday) The museum is open every day of the year except Christmas Day and Thanksgiving Day. Ticket Price: $3.20 for grown-ups, one-fourth (四分之一) for children, and free on Friday and Saturd
29、ay evening from 5:00 pm to 9:00 pm. 46. The American Museum of National History is in_. A. Florida B. Washington C. Boston D. New York 47. You can reach the American Museum of National History_ A. by bus, train, or air B. by bus, subway, or car C. by bus, train, or ship D. by bus, subway, or ship It
30、 is clearly seen that this passage is easy, and the students can easily obtain some important and useful information directly from this passage. For example, the stem No.46 is just a question searching for the address of The American Museum of National History. It is easy for students to find the an
31、swer in the first line of this passage. Whats more, the design of the four options is very easy. It is the same case with the stem No. 47. And the following two passages in this part are also easy to understand. Thats the reason why the kurtosis of this part exceeded the normal range of kurtosis (-2
32、2).In order to raise the validity of this part and improve the quality of junior English test paper, some suggestions are proposed in this paper as follows: In terms of selecting article, this part should include diversified kinds of articles, such as some easy economic, political articles and liter
33、ary works. On one hand, it can reflect the requirements of the test and on the other hand, it can raise the discrimination degree of this part. Consequently, it can test most of the students real English proficiency. Whats more, when designing the options of every stem, first, some of the stems shou
34、ld not always be direct but can be technically designed; second, the options must include some distractions, and the distractions must be reasonable. Thus, it can raise the discrimination degree of this part and improve the validity of this part. 4. Conclusion This paper carries out a study on the v
35、alidity of the test paper for Grade8 junior English and also proposes some suggestions for improving the quality of junior English test paper. English test is independent to English teaching, and the reflection of the students English language proficiency. A test of good quality can not only reach t
36、he goal of evaluating the students achievements in English but also testing the effects of English teaching. Teachers can adjust their teaching strategies and methods according to the result of the tests. It is concluded that teachers should make every possible effort to ensure test quality so as to
37、 minimize the risk of making wrong decisions in junior English teaching. Reference Bachman, Lyle F. & Palmer, Adrian S. Language Testing in PracticeM. 上海：上海外语教育出版社, 1999. Ferguson, G. Handout: Lecture on the validity and validation. Sheffield: University of Sheffield, 2006. Heaton, J.B. Writing English Language Tests. Beijing: Foreign Language Teaching and Research Press, 2000. 桂诗春语言测试：新技术与新理论J. 外语教学与研究， 1989. 孔德惠英语测试学M. 长春：东北师人出版社， 2000. 李绍山语言测试的反拨作用与语言测试设计J. 外语界， 2005. 刘宝全，席仲恩SPSS 在英语试卷统计分析中的应用J. 外语电化教学，2004. 沈齐齐测试与英语教学J. 外语界，1994. 邹申语言测试M上海：上海外语出版社， 2005.