The Washback of IELTS Listening on Senior School English Listening Teaching and Learning

: IELTS test is one of the well-known international standard level of English language ability test for those who intend to use English as the language of communication in the country or regions to study or work. This study was conducted in Mainland China with a number of 12 senior high school students, aiming to explore the washback of IELTS listening and based on the findings to give some suggestions on English listening teaching and learning. Thirty percent of the original test paper was revised and quantitative data was collected. Statistics analysis was used for evaluating the obtained data. This finding highlighted the importance of listening strategies in listening test and and the washback on teachers’ teaching and students’ learning.


Introduction
The era of globalization has long since dawned. As the most widely used language, English is an indispensable tool for us to connect and explore the world (Liang, Zheng, & Wang, 2015). English plays an essential role in the intercommunication between different countries and different regions. For the students who want to study abroad, the TOEFL and IELTS scores are the keys to open the doors of the international prestigious universities. The status of English is also important in China, which not only has a high proportion of the middle school examination and college entrance examination, but also is a proficient skill for the graduate or doctoral students (Zhou, 2017). Therefore, as an English teacher, it is important to enhance students' interest in learning English and improve their English level.
In order to test the listening ability of senior students, we did an IELTS listening test for them. In this test, we revised 30% of the original test paper and chose 12 Mainland senior school students to participate in the IELTS listening test. After the test, we analyzed their test results. Thus, this paper aims to test the IELTS listening level of Mainland senior school students, and on this basis to explore the washback of IELTS listening and further provide some suggestions on English listening teaching and learning in the senior school according to the results of the test.

Literature Review
Since 1978, China began implementing the "reform and opening up"policy, especially after China joined the WTO in 2001, the demand for English has increased quickly. English always has a high position in Chinese education system. In recent years, more and more Chinese people are planning to study or immigrate to English-speaking countries (Deng, 2006). In this case, IELTS test has become the important way for those who want to go to English-speaking countries. According to the IELTS Annual Review in 2005 reported that in the decade from 1995 (when the test was last revised) to 2005, the number of worldwide IELTS candidate increased from about 50000 to 500000, the growth rate reached 1000%.
IELTS refers to the International English Language Testing System. It is one of the famous international English language proficiency standardization tests which is held for non-native speakers. It provides a credible and reliable test method on a global scale to test the English communication ability of candidates' in an English language environment. Currently, more than a million candidates worldwide participate in this test each year and it is recognized in 120 countries (Wilson, 2010).
The IELTS is jointly owned by three organizations, namely the University of Cambridge Local Examinations Syndicate (UCLES), the International Development Program of Australian Universities and Colleges (IDP), and the British Council (Green, 2007). According to Wilson (2010, p.219), "It not only serves as an entry requirement to demonstrate English proficiency in universities in Australia, the United Kingdom, Canada, New Zealand, and the United States, but also as a requirement for immigrants from Australia, New Zealand, and Canada".
IELTS comprehensively assesses English proficiency from four aspects (listening, speaking, reading, and writing) and is available in two different types: Academic purposes and General Training purposes (Green, 2007). The listening and speaking test are the same for all candidates, those who intend to study in an English-speaking country are required to take the reading and writing test in Academic type. While those candidates who want to immigrate or have vocational training are required to take the General Training reading and writing tests (Wilson, 2010).
There are 48 IELTS tests each year and candidates can adjust their time to take the test at any time. The IELTS test a nine-grade scoring system to measure the test results. Candidates have separate scores for listening, speaking, reading, and writing. Their total score will be scored based on the four parts' scores (Wilson, 2010). Different schools have different requirements, but universities usually require 6 or 7 points for unconditional admission on the 9 points of the IELTS scale (Green, 2007).
The IELTS listening consists of four sections. Section one and section two assess issues related to the social needs and evaluate comprehension of everyday conversation. Section 3 and section 4 mainly happen in an academic context, which tests the candidates' express and understanding ability in academic discourse. Section 1 is usually a daily conversation for transaction purposes. In this section, speakers may use phrasal verbs, fillers, and oral words, or they may leave unfinished sentences or interrupt each other (Deng, 2006). Section 2 is usually a monologue to explain things (Sabet & Babaei, 2017). Deng (2006, p.19) pointed out that "In section 2, the speaker can speak directly to the audience, for example, the guide's speech, and the language may be informal. Alternatively, this section can provide information in the form of recorded information or radio broadcasts for invisible audiences. In this case, languages may be more formal and longer sentences and more complicated grammars". Section 3 is an academic conversation. This dialogue may be similar in the form of section 1, but the vocabulary and structure are likely to be heard frequently in the context of academic terms. The scope of the vocabulary may be broader than the previous conversation. Section 4 is an academic monologue, which is probably well-prepared. This section is probably the closest to the written language, with intensive information and complex clauses and sentences (Deng, 2006).
A total of 40 tasks in the listening part, 10 tasks in each part. Participants need to understand the materials and complete all the tasks with 30 minutes. According to IELTS 10 (2015) listed, the main types of question are gap filling, multiple choice, sentence completion, short answer question, labeling a diagram/plan/map, Note/summary/flow chart/table completion items, matching items, and classification items. As Deng (2006, p.18), mentioned "The most important message that an IELTS listening test conveys to candidates is that the language is used for a variety of different functions. In each case, the context, purpose, and relationship between speakers will influence the language used".

Participants
Participants are from a key senior high school in Shanghai. We randomly selected 12 students from a class in Grade 2. According to their GPA results, the chosen participants have an advanced English level.

Test Purpose
This IELTS listening test was designed to measure participants' listening abilities. Instead of applying the same test paper from the IELTS 10 Test 1, we changed 30 percent of the questions, which means 12 questions were revised. The revised questions were distributed separately in question 7,9,11,12,14-15,18,22-23,27,35-36. One reason is that we want to decrease the difficult degrees for those participants, to make it more approachable for them to answer the questions. For example, question 14, the written answer should be "safety rules", the test point then changed to "the equipment". Participants thus can hear the answer directly from the listening tape. As for question 15, the original answer was "plan", and we made a change to "sixweek" so that participants were supposed to hear the information about time, which evaluated students' ability of capturing specific information.
Another reason is to adjust the order of the questions. The purpose was to make it more consistent with the listening materials, thus the students can follow the tape and find out the answers, such as the exchanging position of question 11 and question 12.

Implementation
The test was provided during an English class in the morning, almost the same hour that the real IELTS listening exam conducts. It cost 45 minutes to finish the test. During the examination, there was one English teacher invigilating the whole class to avoid cheating. The listening tape was broadcast through the computer equipment in the class.

Total Accuracy Rate
There are 40 questions in total in IELTS listening test. Based on participants' test results, we analyzed the data of total accuracy rate from three aspects ( Figure 1): accuracy rate equals to 0, accuracy rate over 50 percent and accuracy rate between the two.

Accuracy Rate of Each Section
The listening test has four sections altogether. Figure 2 counts the accuracy rate of each section.

Accuracy Rate of Question Type
Questions in the listening test are made up of three types.

Accuracy Rate of Constructs
The listening test consists of four kinds of constructs. Figure 4 shows the accuracy rate of each construct in these 40 questions.

Findings
According to Figure 1, we found that the majority of the participants fall into the area of zero to 50 percent. The accuracy rate of zero also takes a large proportion in the whole. The accuracy rate over 50 percent, however, is half less than the accuracy rate of zero and nearly 6 times less than the range of zero to 50 percent. Only 5 questions get an accuracy rate over 50 percent out of 40 questions, which are question 2, 6, 21, 25 and 32. Twelve questions, as in question 7, 13, 17, 18, 19, 26, 29, 34, 35, 37, 38 and 40, get zero scores.
In the structure of listening, question 11 to question 20 belong to section 2. Starting from question 13 to question 20, the question type is all about sentence completion. Question 31 to question 40 belong to section 4, and all the question types in section 4 are sentence completion. In the listening test, according to the above data, we found that participants perform relatively weak in section 2 and section 4. The data analysis in Figure 2 also confirms this. The errors that participants made in section 2 and section 4 are mainly focusing on one question type, which is sentence completion. As shown in Figure 3, we can also see clearly that the accuracy rate of sentence completion is much lower than the other two.
As for the accuracy rate of constructs, the construct on listening for specific information consists of 30 questions, taking a great proportion in the whole test. Ten questions get zero scores, which are largely in section 2 and section 4. Three questions get an accuracy rate over 50 percent, which are mainly in section 1. The rest fall into the range of 0 to 50 percent. As to distinguishing the main idea from supporting details or examples, there are 7 questions in total, which greatly gather in section 3. The question types in this area are mostly multiple choice, which help reduce the difficulty of the questions. As we can see in Figure 4, the vast majority of the high accuracy rate of constructs falls into the area of distinguishing the main idea from supporting details or examples. There are two questions in making inferences, which are question 36 and question 37. The accuracy rate in these two questions is quite low with an accuracy rate less than 10 percent. According to Figure 4, making inferences accounts for a much smaller proportion than the other two. The question type in deducing the meaning of unfamiliar words from context is made up of one question, which is question 38, getting an accuracy rate of zero.
As above mentioned, the problems largely show in section 2 and section 4, which are all about sentence completion. To complete these questions requires an advanced level in interpreting the listening material, where the levels students are exactly weak at.
The reasons behind these we found that for those high accuracy rate questions, the answers are easy for participants because they can directly hear the information. In addition, the instructions of these questions are helpful for them to get a correct answer. If the information in instruction is not clear, it will contribute to students' difficulty in finding out the correct answer.
For those low accuracy rate questions, however, we found that the serious problems are chiefly caused by synonym replacement, unfamiliar words and the design of the question. For example, some instructions are replaced by other synonyms of the original text. It takes time for participants to think so that their information processing speed becomes slower, but the time for thinking is rather short. If they spend too much time on thinking about the answers may lead to the following listening comprehension a big problem. Take question 37, which is to test the construct of making inferences, as an example, "Unrestricted fishing is affecting the salmon supply." The correct answer is the underlined word. According to the word "salmon" in the instruction, we can predict that the answer should be a noun, but the word "unrestricted" replaced the original phrase "no legal limit" which is followed by "fishing" in the listening material. Participants are not familiar with the word in the instruction, so no one can make the inference of the correct answer.
Some words are too difficult for students because the vocabulary is less used, causing students' spelling difficulties, such as question 36 and question 38. Take one more example in question 38, "The bears' existence is also threatened by Deducing the meaning of unfamiliar words from context their low rate of reproduction." No one can deduce the correct answer because the word is beyond their knowledge so that they cannot spell correctly.
In regard to the design of questions, question 36 might be rather challenging for students to make inferences. There is no wonder why students get very low accuracy rate in this question. The original listening material is "Over many years, logging companies have stripped the land by cutting down a large number of trees", but in the answer sheet, students are allowed to write only one word to replace "cutting down a large number of trees" as in the answer key "Habitat is being lost due to deforestation and construction of roads by logging companies."

Discussion
Test Validity is usually taken as a test evaluation interpretation, it includes content validity, construct validity and criteria validity. Content validity evaluates how well the content of instrument reflected the purpose it might measure. Construct validity questions whether the theoretical construct or trail is actually measured, while criteria validity has the standard to measure the outcomes of the study, and it can be divided into predictive and diagnostic type. The former measures the further outcomes that participants will behave, and the latter measures the performance of the current study (Ary et al.,1985).
When it is used to evaluate an IELTS listening test, content validity pays attention to how well the test reflected the students' learning. Predictive validity serves as an approach to predict students' future performance in the IELTS exam, while diagnostic validity is used to identify problems students have in IELTS listening. In the meantime, construct validity addresses how much the test takers understand this listening exam.
In the current study, the participants we chose did not receive special IELTS training programs, what they learnt is for the preparation of the College Entrance Examination. This exam therefore did not directly reflect on the students' learning procedures. From the perspective of content validity, the test did not reach its goal. It represented that students had some weakness in the listening ability, and they were not familiar with the question types of IELTS listening. Although IELTS listening is testing participants' English listening fluency and the students' results of this study did not show a good performance, it did not mean that these students do not achieve a fluency in English listening. They might be just lack of experience in dealing with the IELTS listening questions.
The data results also illustrated that students are lack of some listening strategies to deal with the IELTS test. Garbutt and Kerry provided some suggestions for IELTS listening in their book. They suggested that it is not necessary for test participants to understand every word clearly, because this will also not be practical in our native language test. Students should predict what words will come out in the listening materials according to the information given in the questions. There are many synonymous substitutions between the questions information and listening materials' information. They can estimate that the answer would be a noun through observation on the grammatical forms of the sentence in the question (Garbutt & Kerry,1996).
According to the data analysis of the questions 17 to 19 where students got zero accuracy rate, it is obvious that synonymous substitutions hinder students' thinking. For example, in question 17, the saying "entitles to" replaced "are given to" in the instruction. In the question 18, the word "preference" replaced "priority", as well as the word "times" took over "hours" from the question instruction. In the question 19, "be sent passes for" changed into "can bring" in the question information. If the students already know this strategy, they will get the correct answers easily. Therefore, if the students are more familiar with the problems sets, and apply appropriate strategies for those questions, they would do better in the listening part.
In addition, the results can indicate that students are unfamiliar with the IELTS listening materials. At the first and second sessions of the listening part, the materials are about daily life topics, and the conversations are more authentic than the listening material of Gaokao. It includes house renting, asking for travel routines or job hunting. The last two sessions are about the academic topic, the contents usually include research note-taking, discussion about projects or presentation of a program. The students did not have that academic experience before, so they lacked the background information which can be an obstacle for them to understand the listening materials in the exam.
If these students can receive sufficient training about IELTS listening test, such as practicing more similar questions types, increasing their understanding of the listening materials and how the question information is given. As well as the trainer exposes them more into the authentic English context and enhance their understanding of the background information. They will achieve great performance in the IELTS listening test in the future.

Limitations on Research
Generally speaking, the whole class are preparing for the college entrance examination for a long period of time and they have been exposed to IELTS test for three times. That is to say, students are not that familiar with IELTS test as we expected.
With respect to times of play, for the college entrance exam, the listening materials of the first five questions will be played once, then the other 15 questions will be played twice. While in IELTS listening test, all of the questions will be provided only once, so many students feel confused and anxious.
In terms of question type, the college entrance examination is all about multiple-choice questions while the major question type of IELTS is filling in the blanks. This set of paper involves 7 multiple-choice questions, and the rest 33 questions are all filling in the blanks. Therefore, another reason for low accuracy is that even though some students with high level of English may capture the information quickly, they are not able to spell it correctly. The difficulty of filling in the blanks is higher than multiple-choice questions.

Limitations on the Test Paper
Take question 36 for an example, students find it significantly hard to paraphrase or summarize the meaning of "cutting down a large number of trees" in only one word "deforestation", which itself is a brand-new word for the majority of students. Therefore, a small number of questions may not be organized reasonably and appropriately for the listening part to evaluate student's ability of capturing information and distinguishing the main idea from listening materials.

Implications for Learning
According to Kömür and Özdemir (2015), when phonological and lexical feedback is provided in interaction, it will be more likely for learners to perceive it correctly. In addition, based on the research of Hazrat and Hessamy (2013), setting oral pushed output as preparation activities before listening is of great efficiency because it inspires participants to be engaged in the active oral production and to concentrate on the pronunciation of words through the process of oral output. The following implications are not only for enhancing listening skills in IELTS test but also for improving English listening comprehensive ability. : feel-fuel-few Students will be able to discriminate one from another and distinguish accurate meaning from listening materials if they are becoming more sensitive to subtle phonetic differences. It is beneficial for them to record their own pronunciation and make a contrast with native speakers and other students, which assists to find out their own mistakes when using native speakers' recording as a mirror.
Besides the pronunciation, completing vocabulary knowledge requires knowing the spelling, word category, word stress, intonation, rhythm, connotative and denotative meanings and so on. The minimal pair practice is also appropriate for these.

2) Word Boundary Phenomena -Linking, Assimilation and Elision
For question 12, "we've expanded it to eight lanes" (audio scripts). There are three confusing word boundary phenomena occurring in this sentence: linking, assimilation and elision. Expanded ends with ed, which is pronounced as [ɪd]. Besides the voice assimilation phenomenon, expanded is also linking with it and pronounced as [ɪk'spendɪt]. Meanwhile, eight is read as [eɪ] and [t] is often omitted, which leads to students' confusion.
To conclude, it is significantly necessary for students to pay attention to popular word boundary phenomena in an authentic English-speaking environment. Analyzing the characters of word linking, assimilation and elision respectively according to various contexts, taking notes and making the comparison between them will decrease students' confusion to a large extent.

Input and Output
1) BBC News Video Based on our research findings, the most serious problem is that students are not good at paraphrasing with a synonym or similar phrases. However, the paraphrasing skills are based on a great amount of input and accumulation. Watching BBC news video is a wonderful way to immerse students into an authentic English context to activate "unfamiliar" words in their mind and accumulate idiomatic expressions.
Firstly, in the stage of extensive listening, our aim is to capture the main idea of the news and summarize in two or three sentences. In the next stage of intensive listening, our goal is to retell or paraphrase each sentence without reference of transcripts. Listening to the news without transcripts until understanding the meaning of each sentence needs patience and perseverance. Students gradually conquer the challenges and could retell the news using as many words or phrases in the transcripts as possible.
2) Simultaneous Interpretation Practice In a study on L2 vocabulary acquisition (Laufer, 1998), it was observed that learners' use of passive or receptive words developed to a higher level than did their active vocabulary. The main reason for that is the lack of elicitation practices.
One of the most effective methods is to translate Chinese dialogues or news to English at the same time with the sequence of increasing difficulty. If students could take more paraphrasing even simultaneous interpretation practice, the newly taught vocabulary will be activated and integrated into their language production flexibly.
Another way to do elicitation exercises and receive useful feedback is seeking any opportunities to communicate with native speakers choosing a wide range of topics involving weather, daily life, entertainment, politics, sports, pop stars and whatever learners are interested in. On the one hand, it will remind students of some active or newly taught vocabulary and provide chances to activate the words to express themselves, which facilitates the process of language production. On the other hand, native speakers will also provide idiomatic expressions. Something we could not ignore is that Chinese students should overcome their fear and embarrassment when making mistakes and also should embrace the culture shock with an open mind.

Implications for Teaching
The current English tests like Gao Kao, CET-4, CET-6, they are not only inhibiting students' enthusiasm and initiative in learning English, but also prompting students' bad learning habits. Hu (2013) pointed out that the objective test type occupies the most parts of the tests while the subjective test type only occupies a small amount in those tests. Therefore, teachers do not pay much attention to the function of real communication of English in their teaching. They put almost all their efforts on teaching the objective question types of the tests to students. According to the implementation of IELTS in the high school, what can teachers learn from this research?
The following are the implications on teaching: 1) Prepare authentic listening materials. Sun (2003) pointed out that IELTS listening test content is the use of authentic listening materials from life. Therefore, teachers should prepare those authentic listening materials for their students, such as UK Radios, BBC News, VOA Special English. Sun (2003) continued that authentic listening materials include formal and informal ones. The formal authentic listening materials such as news, speeches, academic lectures, etc. These materials are generally rigorously structured, with large information content and complete sentences. The informal ones include daily conversations, shopping, telephone calls, etc. These materials are usually loosely structured, have redundant information, and contain many unconventional forms of speech, such as section one and section two in the listening part. Take question 7 in section one for example, the answer is 525. However, the accuracy rate of this question is zero. Students get used to listen to the materials that are recorded in the written form. The information that happens in a natural way in the conversation was hard for them to get. Therefore, teachers must make their students understand that these real listening materials are different from written language recordings in order to understand the real communication contents in the natural environment.
Moreover, the data in the corpus is also an excellent authentic language material that teachers can use. Johns (1991) and Johns (1994) said that the use of corpora in language teaching provides learners with language data to meet their learning needs. Rutherford (1987) claimed that corpus is the study of special software program named concordances that helps to recognize common patterns in authentic language materials. According to Smirnova (2017, p. 308), "the use of corpora in the classroom makes the process of preparation for the IELTS test more efficient and interesting for the learners and helps to develop their general learning skills and language abilities, thus creating beneficial washback." Therefore, using corpus is one of the teaching approaches that teachers could implement in their teaching.
2) Input cultural knowledge in the teaching process. According to Feng (2007), students in China pay most their attention to the grammar, structure of the English sentences. Thus, their cultural knowledge of English is rather limited. Those authentic listening materials in IELTS contains the cultural background of the language. Without understanding the difference between western culture and eastern culture, students will be hard to truly understand the meaning behind the conversations. Thus, teachers should enlarge students' English knowledge, not only subject knowledge but also cultural knowledge.
For the cultural knowledge, Sun (2003) claimed that there are two types of cultural knowledge, one is surface culture and the other is deep culture. For the surface culture, it includes the culture of connotation of knowledge, communication and vocabulary. Take the culture of knowledge for an example, it includes target language country's history, geography, politics, economy, ethnicity, social system, religious beliefs, education law, literature and arts, science and technology, lifestyles, customs, relationships, life customs, social etiquette and so on. For the deep culture, it includes the culture of life outlook, world outlook and values as well as people's ways of thinking, philosophy and national psychology in target language countries. Deep culture influences the way language is expressed and how people behave. The Section four in this paper is related to the subject knowledge as well as cultural knowledge. Total ten questions in this part, half of the questions get accuracy rate of zero. For example, the word fragile in the listening content has another meaning that does not mean weak in mentally or spiritually. Thus, the lack of understanding the meaning behind the word hinder students' understanding of the listening material. Furthermore, students need to have some knowledge of white bear that help them understand the listening content better.
Therefore, in the daily teaching, it is better for teachers to cover both sides of cultural knowledge. Furthermore, teachers can recommend students to watch melodramas like Friends or documentaries like BBC documentaries as those documentaries go deep into the stories of real life, the world around us and our past.
3) Help students reduce their listening anxiety. Golchi (2012) confirmed that low anxiety students perform better in their listening tests. IELTS academic listening recording plays only once. Christenberry (2003) emphasized the listening test's nature and it may cause anxiety on test takers. Chastain (1979) pointed out that listening is a complex skill, therefore, students are afraid that whether they understand and interpret the information correctly or not. Furthermore, Gonen (2009) claimed that due to the authenticity of the listening text, comprehensibility of the listening materials as well as some other factors, such as noise and the quality of the headsets, learners may feel anxious during the listening period. Furthermore, many students confirmed that they had strong anxiety while listened to the IELTS listening test. Therefore, it is significant for teachers to help their students reduce listening anxiety, otherwise, students cannot perform their full potential while listening.
According to Golchi (2012), teachers can adopt three teaching strategies to help their students decrease anxiety: metacognitive strategy, cognitive strategy and socio-affective strategy. For metacognitive strategy, teachers can introduce plans of pre-listening, while-listening, post-listening to their students and create more opportunities for them to expose to IELTS listening test. For cognitive strategy, teachers can use problem-solving strategy during the class, facilitate students' ability of problem-solving skills. Take question 37 in section four for example, no one get the correct answer. Teachers can demonstrate to their students how to get the correct answer, the way of exclude information that hinder understanding. For socio-affective strategy, it is beneficial for teachers to use cooperative learning strategies in their classroom. Students can get help from peers to check their understanding or lower their anxiety.
Finally, besides the above three suggestions, there is also a SMART model for teachers' reference. The SMART model is concluded from Deng's (2006) dissertation. S means to select topics that related to students' daily life. M means to memorize and build a pool of vocabulary that is indispensable for learning a foreign language. Phonetics: vowels, consonants, stress, pause, intonation. A means to provide activities that remind students of the topics and related languages they have learned before. R means to provide extra listening opportunities for students in the form of non-test formats, such as news broadcasts, video, songs, real-life speakers-aiming at experiencing authentic English language and cultural background knowledge. T means to make sure that students are quite clear of the requirements in listening and what they have to do in each task type.
Despite the limitations, this research indicates many useful suggestions on learning and teaching. About implications for learning, students can concentrate on the words' pronunciation through doing minimal pair practice and focusing on word boundary phenomena. Besides, listening to BBC news and taking more paraphrasing even simultaneous interpretation practice assist to accumulate idiomatic expressions and activate "unfamiliar" words during the process of oral output. About implications for teaching, teachers should prepare authentic listening materials for their students, input cultural background knowledge in the teaching process and help students reduce their listening anxiety.